Rank understanding the code second, and say why it is not an afterthought

AGENTS.md

@@ -1,11 +1,27 @@
 # Rules
 
-- Go runs only through `docker compose run --rm <test|vet|fmt|build>`; the merge gate is `docker build .` plus CI's changelog check.
+- Go runs only through `docker compose run --rm <test|vet|fmt|build|cyclo>`; the merge gate is `docker build .` plus CI's changelog check.
 - Tests first, in their own commit; the implementation follows in the next. A re-pin of seeded output or of `testdata/shipped_shape.txt` is its own commit.
 - A change under `data/` or to the shape pin adds its `CHANGELOG.md` entry under `Unreleased` in the same PR, and so does a change to a flag, an exit code, an exported name, a fence, a builtin or the lowest Go; what is major is the README's Versioning table.
 - One-line commit messages: no ticket prefix, no repo name, no authorship trailers.
+- Merges are fast-forward only, which `tea pr merge` cannot spell — every style it
+  offers fails with "is it still open?", as does a merge whose required checks are
+  still pending, so read the checks before believing the style is the problem. Use
+  `tea api /repos/{owner}/{repo}/pulls/<n>/merge -f Do=fast-forward-only`.
+- Gitea queues Actions runs that `/repos/{owner}/{repo}/actions/tasks` does not
+  list, so an empty task list says nothing about whether CI ran; read
+  `/repos/{owner}/{repo}/commits/<sha>/status`.
+- README goal 2 is gated at 7.0 on the `comprehension-panel` skill, and below it
+  nothing else merges — no feature, no category, no data — bar a security fix and a
+  dependency bump. `todo.md` carries the round: the nine-seat findings run at depth 1,
+  a PR per item it names, then the run again, until the score passes. At or above 7.0
+  every pull request scores with the four-seat run and answers it in one run. No merge
+  lowers the last score.
 - Hard tabs. No comment by default; delete a restatement, a rationale, history, or a file preamble.
 - One spelling per result: reject the other at `New`, and let the error name the spelling to use.
 - A standing choice a reader would relitigate goes under Decisions in the README, not in a comment.
+- Prose naming a source states what that source states: read the register's own field
+  before paraphrasing it, and name the register you read before claiming none
+  publishes a spelling.
 - A README example is a `json` block that loads and renders as a category; `readme_test.go` runs every one.
-- Cyclomatic complexity is gated at 14: the table-shaped dispatches (`eachToken`, `calc.factor`, `walkPath`) sit at 13–14 and stay whole; anything else that reaches 14 is decomposed.
+- Cyclomatic complexity is gated at 14: the table-shaped dispatches (`linkParent`, `compileTemplate`, `renderEdges`) sit at it and stay whole; a function that would pass it is decomposed.

CHANGELOG.md

@@ -11,24 +11,82 @@ replacement, and each removed path, column or flag.
 - First release: the CLI, the library and the shipped data set.
 - Table categories: a category JSON naming a `rows` TSV beside it, with the options
   `key`, `name`, `weight` and `parent`; a path selects a row by key or name,
-  `misc.country[SE]`, and descends to a linked table by name; linked tables draw
+  `misc.territory[SE]`, and descends to a linked table by name; linked tables draw
   consistently within one render and draw group. `rows` is an option, so no
-  template may carry a field of that name. Refused at `New`: a `name` without a
+  template may carry a field of that name. A `name` without a `key` resolves
-  `key`, a name spelling another row's key, a table named like a column of any
+  inside the table's `parent`. Refused at `New`: a `name` without a `key` or a
+  `parent`, a name repeating inside one parent row, a name spelling another row's key, a table named like a column of any
   table above it, a table whose format or cell references its own family, and,
   within one render and draw group, a path drawing a table another path selects a
   row of, or two paths pinning different rows of one table.
 - `New` refuses a root choice of templates sharing one format and one set of string
   fields, naming the rows TSV to write instead.
-- `misc.country`, `misc.currency`, `misc.language`, `misc.httpstatus` and
+- `misc.territory`, `misc.currency`, `misc.language`, `misc.httpstatus` and
-  `misc.mimetype` are tables. `misc.country` is every ISO 3166 country that has a
+  `misc.mimetype` are tables. `misc.territory` is every ISO 3166-1 territory that has
-  capital, a currency and a TLD, with the columns `calling-code`, `capital`,
+  a capital, a currency and a TLD, with the columns `calling-code`, `capital`,
-  `currency`, `flag`, `languages`, `numeric` and `tld` added; `misc.currency` the
+  `country`, `currency`, `flag`, `languages`, `numeric` and `tld` added, `country`
+  holding the alpha-2 code of the sovereign state it belongs to and its own where it
+  is one; `misc.currency` the
   current ISO 4217 currencies with a minor unit, with `decimals` and `numeric`
-  added, and its symbols from CLDR. `DATA-LICENSES.md` lists each source.
+  added, and its symbols from CLDR; `misc.language` every ISO 639-2 entry carrying a
+  639-1 code, with a `code3` column holding its 639-2/T code; `misc.httpstatus` each
+  code the IANA registry lists with a plain reason phrase; and `misc.mimetype` each
+  type mime-db records as IANA-registered and gives a filename extension.
+  `DATA-LICENSES.md` lists each source.
+- `misc.timezone`, `misc.car` and `misc.useragent` are tables too. `misc.timezone`
+  is every zone tzdb `zone.tab` gives a shipped territory, with `offset` holding the
+  zone's standard UTC offset and `territory` linking to `misc.territory`, so
+  `misc.territory[SE].timezone` draws `Europe/Stockholm`; zones of a territory
+  `misc.territory` does not ship, Antarctica's among them, and the constant `UTC`
+  are gone; spell that one as the text `"UTC"`. A `weight` column carries the
+  population GeoNames records in each zone, so a draw favours the zones people live
+  in. `misc.useragent` is the top-user-agents desktop and mobile lists with `browser`,
+  `device` and `os` columns. `misc.car` keeps its makes and models in a `make` and a
+  `model` column.
+- `misc.httpmethod`, `misc.protocol` and `misc.port` are IANA's registries as tables.
+  `misc.httpmethod` is the eight methods RFC 9110 defines and PATCH, carrying
+  `safe` and `idempotent` as `true` or `false`; `misc.protocol` the keyword, its
+  `number` and its `name`, selectable by either; `misc.port` renders a TCP port number,
+  with the IANA `service` name beside it, keeping the first service the registry
+  describes for a port so a number selects one row.
+- `misc.tld` is the IANA root zone as a table: every delegated TLD keyed with its
+  leading dot, `.se` and `.xn--p1ai`, the register's `type` beside it, and a `unicode`
+  column holding the form the register displays, so `misc.tld[.рф]` selects the row
+  `.xn--p1ai` renders. A delegation the register marks "Not assigned" does not ship.
+- `misc.loglevel` is the eight syslog severities: the `keyword` a configuration writes,
+  `err` and `info`, keyed by the numerical `code` a PRI encodes and selectable by
+  either, with the `severity` RFC 5424 spells beside it.
 - `geo.SE` and `geo.US`: five linked tables per country, `region`, `municipality`,
   `locality`, `postal-code` and `street`, weighted by population and address counts
   and built from SCB, GeoNames, Trafikverket NVDB and the US Census Bureau, and an
   `address` record over one consistent draw of them. `sv_SE.address` and
   `en_US.address` read those records, so `en_US.address.street` no longer carries
   `name` and `suffix`, and a locale folder loads only beside `geo`.
+- `{date(from,to,'layout')}` and `{time('layout')}`: a second between two days, or
+  within one, in a single-quoted Go layout, drawn in UTC; `from` may equal `to`.
+  `sv_SE.date`, `en_US.date`, `sv_SE.time` and `en_US.time` render through them, so
+  `date.year`, `date.month`, `date.day`, `time.hour`, `time.minute`, `time.minute.t`,
+  `time.sec` and `time.ampm` are no longer paths — a part of a date is now its own
+  `{date(…,'2006')}`. `misc.datetime` is an RFC 3339 instant.
+- `sex`, `first-name` and `last-name` tables in `sv_SE` and `en_US`, weighted by
+  bearers from SCB, the SSA and the Census Bureau; `first-name` links to `sex`, and a
+  name both sexes carry is a row under each. `person` reads them, so its columns are
+  `first`, `last`, `prefix` and `sex`; `person.femalefirst` and `person.malefirst` are
+  no longer paths — draw `sex[f].first-name` and `sex[m].first-name` instead.
+  `en_US.title` links to `sex` as well, so `en_US.person.prefix` draws `Mr` or `Ms`
+  without contradicting the record's `sex`; `sv_SE.title` is an unsexed table of the
+  same shape.
+- `sv_SE.personnummer` and `sv_SE.samordningsnummer`, Skatteverket's test series
+  from a `sv_SE.birth-number` table under `sex`, in place of `sv_SE.ssn`, whose
+  `ssn.mmdd`, `ssn.mmdd.m` and `ssn.mmdd.d` go with it. `en_US.ssn` now draws the
+  ranges the SSA assigns and carries the columns `area`, `group` and `serial`, so
+  `--format csv en_US.ssn` writes a header where it used to fail; `en_US.itin` is new
+  and carries the same three. `person.prefix` is null where a person has no title,
+  where it used to be an empty string, so `--format sql` writes `NULL` and a `string`
+  struct field reading it becomes `*string`.
+- `ErrNoColumns` is exported, so a caller can tell the one record fence a path can
+  answer from the rest.
+- An error names a spelling that runs: a layout is named single-quoted and free of
+  its own quotes, a row of a table with no key is named as the path that selects it,
+  `sv_SE.sex[f].first-name[Kim]`, and a category with no columns names the record
+  that gives it one.

DATA-LICENSES.md

@@ -11,10 +11,21 @@ Every shipped dataset, its source, its licence and the attribution it asks for.
 | `geo/SE/street.tsv` | [Trafikverket NVDB](https://www.trafikverket.se/) Gatunamn, through the open API | CC0 1.0 | none required | `data-import/geo-se.py` |
 | `geo/US/region.tsv`, `municipality.tsv`, `locality.tsv` | [Census Bureau](https://www.census.gov/) Gazetteer 2026 and population estimates 2025 | [public domain](https://www.usa.gov/government-works) | none required | `data-import/geo-us.py` |
 | `geo/US/postal-code.tsv`, `street.tsv` | Census Bureau ZCTA to place relationships 2020 and TIGER/Line 2025 address ranges and feature names | public domain | none required | `data-import/geo-us.py` |
-| `misc/country.tsv` | [datasets/country-codes](https://github.com/datasets/country-codes) | [PDDL 1.0](https://opendatacommons.org/licenses/pddl/1-0/) | none required | `data-import/country.py` |
+| `sv_SE/first-name.tsv`, `last-name.tsv` | [SCB](https://www.scb.se/) names with at least two bearers, 31 December 2022 | CC0 1.0 | "Källa: SCB" | `data-import/names-se.py` |
+| `en_US/first-name.tsv` | [SSA](https://www.ssa.gov/oact/babynames/) baby names, births 1930 to 2020, through [hackerb9/ssa-baby-names](https://github.com/hackerb9/ssa-baby-names) | public domain | none required | `data-import/names-us.py` |
+| `en_US/last-name.tsv` | Census Bureau surnames occurring 100 or more times, 2010 | public domain | none required | `data-import/names-us.py` |
+| `sv_SE/sex.tsv`, `sv_SE/birth-number.tsv`, `sv_SE/title.tsv`, `en_US/sex.tsv`, `en_US/title.tsv`, `misc/car.tsv` | curated; Skatteverket's test birth numbers are facts, and `car.tsv` waits for an international source (goal 12) | — | — | — |
 | `misc/currency.tsv` | [datasets/currency-codes](https://github.com/datasets/currency-codes); symbols from [Unicode CLDR](https://github.com/unicode-org/cldr) `en.xml` and `root.xml` | PDDL 1.0; [Unicode License v3](https://www.unicode.org/license.txt) | CLDR: "Copyright © 1991-2025 Unicode, Inc. Unicode and the Unicode Logo are registered trademarks of Unicode, Inc. in the United States and other countries." | `data-import/currency.py` |
-| `misc/httpstatus.tsv` | curated (IANA HTTP status codes are facts) | — | — | — |
+| `misc/httpmethod.tsv` | [IANA HTTP Method Registry](https://www.iana.org/assignments/http-methods/) | [public domain](https://www.iana.org/help/licensing-terms) | none required | `data-import/httpmethod.py` |
-| `misc/language.tsv` | curated (ISO 639-1 codes are facts) | — | — | — |
+| `misc/httpstatus.tsv` | [IANA HTTP Status Code Registry](https://www.iana.org/assignments/http-status-codes/) | [public domain](https://www.iana.org/help/licensing-terms) | none required | `data-import/httpstatus.py` |
-| `misc/mimetype.tsv` | curated (IANA media types are facts) | — | — | — |
+| `misc/language.tsv` | [datasets/language-codes](https://github.com/datasets/language-codes), the [Library of Congress](https://www.loc.gov/standards/iso639-2/) ISO 639-2 register | [PDDL 1.0](https://opendatacommons.org/licenses/pddl/1-0/) | none required | `data-import/language.py` |
+| `misc/loglevel.tsv` | [RFC 5424](https://www.rfc-editor.org/rfc/rfc5424.txt) Table 2, the syslog message severities; keywords from the [POSIX](https://pubs.opengroup.org/onlinepubs/9799919799/basedefs/syslog.h.html) `<syslog.h>` severity macros | facts | none required | `data-import/loglevel.py` |
+| `misc/mimetype.tsv` | [mime-db](https://github.com/jshttp/mime-db), the IANA media type registry with filename extensions | [MIT](https://github.com/jshttp/mime-db/blob/master/LICENSE) | "Copyright (c) 2014 Jonathan Ong, Copyright (c) 2015-2022 Douglas Christopher Wilson" | `data-import/mimetype.py` |
+| `misc/port.tsv` | [IANA Service Name and Transport Protocol Port Number Registry](https://www.iana.org/assignments/service-names-port-numbers/) | [public domain](https://www.iana.org/help/licensing-terms) | none required | `data-import/port.py` |
+| `misc/protocol.tsv` | [IANA Protocol Numbers](https://www.iana.org/assignments/protocol-numbers/) | [public domain](https://www.iana.org/help/licensing-terms) | none required | `data-import/protocol.py` |
+| `misc/territory.tsv` | [datasets/country-codes](https://github.com/datasets/country-codes) | [PDDL 1.0](https://opendatacommons.org/licenses/pddl/1-0/) | none required | `data-import/territory.py` |
+| `misc/timezone.tsv` | [IANA tzdb](https://www.iana.org/time-zones) 2026d `zone.tab` and the standard offset of each zone; weights from [GeoNames](https://www.geonames.org/) `cities15000` populations | [public domain](https://data.iana.org/time-zones/tzdb/LICENSE); [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | "Populations from GeoNames, www.geonames.org" | `data-import/timezone.py` |
+| `misc/tld.tsv` | [IANA Root Zone Database](https://www.iana.org/domains/root/db), checked against the root zone's [TLD list](https://data.iana.org/TLD/tlds-alpha-by-domain.txt) | [public domain](https://www.iana.org/help/licensing-terms) | none required | `data-import/tld.py` |
+| `misc/useragent.tsv` | [top-user-agents](https://github.com/microlinkhq/top-user-agents) desktop and mobile lists | [MIT](https://github.com/microlinkhq/top-user-agents/blob/master/LICENSE.md) | "Copyright © 2020 Kiko Beats" | `data-import/useragent.py` |
 
-Every other category is hand-written JSON under [`data/`](data), MIT like the code.
+Every other category is hand-written under [`data/`](data), MIT like the code.

README.md

@@ -17,11 +17,12 @@ fejkdata sv_SE.person.last                     # Eriksson
 fejkdata --seed 42 sv_SE.address               # the same address every run
 fejkdata -n 3 --separator ', ' sv_SE.word      # nät, barn, sol
 fejkdata --list                                # every path the data offers
-fejkdata 'misc.country[SE].capital'            # Stockholm — a table's row, selected by key or name
+fejkdata 'misc.territory[SE].capital'          # Stockholm — a table's row, selected by key or name
 fejkdata 'geo.SE.locality[Lund].street'        # Fjelievägen — a linked table, drawn inside the row
 fejkdata --data-path ./mydata sv_SE.word       # layer a directory over the shipped data
 fejkdata --no-shipped-data -d ./mydata --list  # only your data
 fejkdata 'name: {/sv_SE.person.last}'          # name: <a surname> — an inline template
+fejkdata "{date(1990-01-01,2010-12-31,'2006-01-02')}"  # 2003-11-27 — the argument in "…", the layout in '…'
 fejkdata '{"format":"name: {x}","x":["bosse","lina"]}'  # name: bosse or name: lina
 ```
 
@@ -32,11 +33,9 @@ a JSON object, array or string, or that carries a `{` token, is instead an
 **inline template**: a format string or a JSON value compiled and rendered on the
 spot. Its tokens reach the data by reference from the root —
 `{/sv_SE.person.last}`, so shipped and `--data-path` categories are alike
-available. An inline template sits in no folder, so the folder-relative `{.name}`
+available. A path never contains a brace or a quote, and a bracket only as a selector
-and `{..name}` are rejected naming the root spelling, and one reference alone —
+after a name, so the two spellings cannot collide; which spellings an inline template
-`{/sv_SE.person}` — is the path written as a template, rejected naming the path, as is
+rejects, and what each names instead, is under [Decisions](#decisions).
-a path written `/sv_SE.person`. A path never contains a brace or a quote, and a
-bracket only as a selector after a name, so the two cannot collide (see [Decisions](#decisions)).
 
 | Flag | |
 |------|--|
@@ -114,6 +113,12 @@ the template itself, which composes the format into one string rather than
 projecting columns — ask for more records with `--repeat`. A `repeat` on a column
 is fine.
 
+Columns are written in name order, whatever order the fields appear in. A category
+whose fields are the parts of one value — `sv_SE.price`, `sv_SE.version`, `misc.uuid` —
+projects those parts rather than the value, so for one column holding what `Fake`
+renders, write `{"format":"","price":"{/sv_SE.price}"}` as a fieldless category asks
+for.
+
 A column carrying a newline keeps it inside the quoted CSV field or the SQL string
 literal, so a row can span physical lines: read the stream with a CSV or SQL
 parser rather than splitting it on newlines.
@@ -141,6 +146,119 @@ a value two fields share in its own category and reference that. A field hold, t
 operand ties fields together within one column as always (see
 [Correlated fields](#correlated-fields) and [Decisions](#decisions)).
 
+## Data
+
+The shipped set under [`data/`](data) — one folder per locale (`en_US`, `sv_SE`)
+plus a locale-neutral `misc` folder — is embedded, so the CLI and the library
+work with no data on disk. A directory is a namespace: each JSON file is a
+category named after the file, each subdirectory a dot-path segment, so
+`mydata/sv_SE/person.json` is `sv_SE.person` and replaces the shipped one.
+Sources merge in order; matching folders combine, any other clash is won by the
+last loaded. Names may not use `.`, `|`, `(`, `{`, `}`, `[`, `]`, `"` or `/`, nor be
+`-`, which a struct tag reserves; dot-prefixed entries are skipped, so a data directory can also be a checkout.
+
+Each locale carries `address`, `color`, `company`, `date`, `email`, `first-name`,
+`ip`, `last-name`, `person`, `phone`, `price`, `sentence`, `sex`, `time`, `url`,
+`username`, `version` and `word`, formatted per locale; `sv_SE` adds
+`personnummer` and `samordningsnummer`, `en_US` adds `ssn` and `itin`. `misc`
+carries `car`, `coordinate`, `creditcard` (Luhn-valid), `currency` (ISO 4217),
+`datetime` (RFC 3339), `emoji`, `httpmethod`, `httpstatus`, `language` (ISO 639-1,
+with its 639-2/T code), `loglevel` (syslog severity), `mac`, `mimetype`, `objectid`,
+`port`, `protocol`, `territory` (ISO 3166-1), `timezone` (IANA), `tld` (IANA root
+zone), `useragent` and `uuid` (v4). Many carry sub-fields — `misc.currency.symbol`,
+`misc.territory.alpha2`, `misc.httpstatus.code` — which `--list` shows. `car`,
+`currency`, `httpmethod`, `httpstatus`, `language`, `loglevel`, `mimetype`, `port`,
+`protocol`, `territory`, `timezone`, `tld` and `useragent` are [tables](#table), so
+`misc.territory[SE].capital` and `misc.currency[Euro].symbol` select a row; `car` and
+`useragent` carry no key or name, so they are drawn from rather than selected in.
+
+`misc.httpmethod`, `misc.protocol` and `misc.port` are IANA's registries.
+`misc.httpmethod` is the eight methods RFC 9110 defines and PATCH; the register's
+other entries, WebDAV and DeltaV among them, do not ship. Its `safe` and
+`idempotent` are `true` or `false`, so a Go `bool` reads them, though `--format json`
+and `sql` write them as text, a table column carrying no datatype. `misc.protocol`
+carries a `number` and a `name`, the keyword itself where the register spells none out,
+and selects by either spelling.
+`misc.port` is the TCP assignments, rendering the number a port field holds with the
+IANA `service` beside it, and selecting by number alone. A draw spans the whole
+register, so pin `misc.port[443]` where a fixture needs a port a reader recognises;
+it stops at 49150, IANA assigning nothing above, so an ephemeral source port is
+`{int(49152,65535)}`.
+`misc.tld` is the root zone's delegated TLDs, keyed with the leading dot
+`misc.territory.tld` already carries, `.se`. Its `type` is the register's —
+`country-code`, `generic`, `generic-restricted`, `infrastructure` or `sponsored` — and
+its `unicode` the form the register displays, `.рф` for `.xn--p1ai` and the key itself
+for an ASCII one, which selects the row too, so `misc.tld[.рф]` renders `.xn--p1ai`. A
+draw spans the whole zone, `.arpa` and `.zuerich` alike, so pin `misc.tld[.com]` where
+a fixture needs one a reader recognises.
+`misc.loglevel` is the eight syslog severities, rendering the keyword a configuration
+writes, `err` and `info`, keyed by the numerical code a PRI encodes and selectable by
+either, so `misc.loglevel[3]` and `misc.loglevel[err]` are one row.
+`misc.loglevel.severity` is the name RFC 5424 spells, `Error` beside `err`; it
+selects nothing, the code and the keyword do.
+[`DATA-LICENSES.md`](DATA-LICENSES.md) names each table's source and licence.
+
+`misc.timezone` is every zone tzdb gives a shipped territory, from one apiece for most
+of them to dozens for the largest, weighted by the population GeoNames records in each,
+so `misc.territory[US].timezone` draws `America/New_York` far more often than
+`America/Nome`. It links to `misc.territory`, so `misc.territory[SE].timezone` is
+`Europe/Stockholm` and a drawn territory and zone agree. Its `offset` is the zone's
+*standard* offset, so do not pair it with a drawn `misc.datetime`: in a zone that
+observes DST it is the wrong one half the year, which is what storing a zone name
+avoids. A zone tzdb named recently — `Europe/Kyiv`, `America/Ciudad_Juarez` — is
+rejected outright by a consumer resolving it against older tzdata, so where the
+consumer validates the zone, pin one rather than draw it. There is no `UTC` row, tzdb
+giving that name no territory — spell it as the text `"UTC"`. `misc.useragent` carries
+`browser`, `device` and `os` beside the string, and `misc.car` a `make` and a `model`,
+until an international source replaces them.
+
+ISO 3166-1 codes territories, not sovereign states, so that is what the table is
+called: Greenland and Åland have codes of their own, and `misc.territory.country`
+names the state each belongs to — `DK` for Greenland, `FI` for Åland, and its own
+code for a sovereign one, or for one the register names no state for.
+
+`sex`, `first-name` and `last-name` are tables weighted by bearers, from SCB, the
+SSA and the Census Bureau. `first-name` links to `sex`, so `sv_SE.sex[f].first-name`
+draws a woman's name, and a name both sexes carry is a row under each, so
+`en_US.sex[m].first-name[Taylor]` names the one a `first-name[Taylor]` alone cannot.
+`person` reads one draw of the three, so its `first` and `sex` columns agree, and so
+does a `personnummer` in the same render: its birth number, `sv_SE.birth-number`
+under `sex`, is Skatteverket's test series, 238 for a woman and 239 for a man, which no
+real person is ever given. `en_US.title` links to `sex` too, so a person's prefix
+never contradicts it.
+
+A person of a chosen sex is assembled from the tables — `sex[f].first-name` beside
+`last-name` — while a shipped `personnummer` agrees with the sex its own render
+*drew*, not with one a path selects. That test series is also small: a personnummer
+is one of about 70,000 values, a day in 1930–2025 against the two birth numbers, so a
+fixture past a few hundred rows repeats one and a `UNIQUE` column needs a category of
+your own. `sv_SE.date` and `en_US.date` are uniform over 1970-01-01 to 2029-12-31,
+`misc.datetime` over 2000-01-01 to 2029-12-31. What the two locales do not share:
+`en_US.address` carries a `region` column the Swedish one has no use for, and
+`sv_SE.title` has no `parent`, so it is selected as `sv_SE.title[dr]` rather than
+inside a sex.
+
+A `geo` folder holds one tree per country under its alpha-2 code: five
+[linked tables](#linked-tables) named alike, and an `address` record over one
+consistent draw of them, which the locale's `address` reads.
+
+| Table | `geo.SE` | `geo.US` | Weight |
+|-------|----------|----------|--------|
+| `region` | län, by code or name | state, by USPS abbreviation or name; `code` is the FIPS code | population |
+| `municipality` | kommun, by code or name | county, by FIPS code or name | population |
+| `locality` | postort, by name | incorporated place of 25,000 people or more with a postal code of its own, by GEOID or name; Hawaii has none | tätort population, the kommun's where the postort names it, else 200; place population |
+| `postal-code` | postnummer with street delivery, by code | ZCTA, by code | one; address ranges |
+| `street` | gatunamn, the ten with most road segments per postort | street name, the ten with most address ranges per place | segments; address ranges |
+
+`geo.SE.region[Skåne län].municipality` draws a kommun in Skåne,
+`geo.SE.locality[Lund].street` a street in Lund, and
+`geo.US.region[IL].locality[Springfield]` settles which Springfield. A region row
+carries its `timezone`, the state's predominant zone, and a locality its `lat` and
+`lon`. What ports across countries is the five table names, the `name` column,
+selection by name, and the `address` record's columns `street`, `street-number`,
+`postal-code` and `locality`; every other column is the country's own, `code` on a
+Swedish region but `abbr` on a US one.
+
 ## Library
 
 ```sh
@@ -152,7 +270,7 @@ f, err := fejkdata.New(fejkdata.WithSeed(42))
 if err != nil {
 	log.Fatal(err)
 }
-v, err := f.Fake("sv_SE.address") // "Kungsvägen 68\n379 17 Stockholm"
+v, err := f.Fake("sv_SE.address") // "Järvedsvägen 43\n891 77 Järved"
 paths := f.List()                  // every path Fake accepts, sorted
 v, err = f.FakeTemplate("name: {/sv_SE.person.last}")      // compile + render in one call
 t, err := f.NewTemplate(`{"format":"name: {x}","x":["bosse","lina"]}`) // compile once
@@ -206,49 +324,6 @@ A `*Generator` is safe for concurrent use; a seeded sequence is reproducible onl
 when drawn from one goroutine. Changing how a value is composed shifts the seeded
 stream for that value and everything drawn after it.
 
-## Data
-
-The shipped set under [`data/`](data) — one folder per locale (`en_US`, `sv_SE`)
-plus a locale-neutral `misc` folder — is embedded, so the CLI and the library
-work with no data on disk. A directory is a namespace: each JSON file is a
-category named after the file, each subdirectory a dot-path segment, so
-`mydata/sv_SE/person.json` is `sv_SE.person` and replaces the shipped one.
-Sources merge in order; matching folders combine, any other clash is won by the
-last loaded. Names may not use `.`, `|`, `(`, `{`, `}`, `[`, `]`, `"` or `/`, nor be
-`-`, which a struct tag reserves; dot-prefixed entries are skipped, so a data directory can also be a checkout.
-
-Each locale carries `address`, `color`, `company`, `date`, `email`, `ip`,
-`person`, `phone`, `price`, `sentence`, `ssn`, `time`, `url`, `username`,
-`version` and `word`, formatted per locale. `misc` carries `car`, `coordinate`,
-`country` (ISO 3166), `creditcard` (Luhn-valid), `currency` (ISO 4217), `emoji`,
-`httpstatus`, `language` (ISO 639), `mac`, `mimetype`, `objectid`, `timezone`
-(IANA), `useragent` and `uuid` (v4). Many carry sub-fields — `misc.currency.symbol`,
-`misc.country.alpha2`, `misc.httpstatus.code` — which `--list` shows. `country`,
-`currency`, `httpstatus`, `language` and `mimetype` are [tables](#table), so
-`misc.country[SE].capital` and `misc.currency[Euro].symbol` select a row;
-[`DATA-LICENSES.md`](DATA-LICENSES.md) names each table's source and licence.
-
-A `geo` folder holds one tree per country under its alpha-2 code: five
-[linked tables](#linked-tables) named alike, and an `address` record over one
-consistent draw of them, which the locale's `address` reads.
-
-| Table | `geo.SE` | `geo.US` | Weight |
-|-------|----------|----------|--------|
-| `region` | län, by code or name | state, by USPS abbreviation or name; `code` is the FIPS code | population |
-| `municipality` | kommun, by code or name | county, by FIPS code or name | population |
-| `locality` | postort, by name | incorporated place of 25,000 people or more, by GEOID or name; Hawaii has none | tätort population, the kommun's where the postort names it, else 200; place population |
-| `postal-code` | postnummer with street delivery, by code | ZCTA, by code | one; address ranges |
-| `street` | gatunamn, the ten with most road segments per postort | street name, the ten with most address ranges per place | segments; address ranges |
-
-`geo.SE.region[Skåne län].municipality` draws a kommun in Skåne,
-`geo.SE.locality[Lund].street` a street in Lund, and
-`geo.US.region[IL].locality[Springfield]` settles which Springfield. A region row
-carries its `timezone`, the state's predominant zone, and a locality its `lat` and
-`lon`. What ports across countries is the five table names, the `name` column,
-selection by name, and the `address` record's columns `street`, `street-number`,
-`postal-code` and `locality`; every other column is the country's own, `code` on a
-Swedish region but `abbr` on a US one.
-
 ## Data format
 
 Every value is a **node**, nestable without limit:
@@ -407,7 +482,9 @@ cell token, and refuses a TSV no category names, a key that is empty or repeats,
 weight that is not a positive number, and a key or name holding `[`, `]`, `{`, `}`,
 `"` or `|`, which a selector cannot spell; the rows are indexed on the first draw that
 selects one. A `name` needs a `key`, since a name naming several rows is reported by
-their keys, and a name spelling another row's key is refused, since the key would
+their keys, or a `parent`, inside whose row a name names one row, so `first-name[Kim]`
+is settled by the `sex` selected before it and a name repeating inside one parent row
+is refused; a name spelling another row's key is refused, since the key would
 select first and the name never. The table's options are its own — `rows`, `key`,
 `name`, `weight` and `parent` — so a column may be named `name`, as one usually is.
 
@@ -417,11 +494,11 @@ inline template has no file beside it, so there it stays a choice.
 
 ### Row selection
 
-`[key]` or `[name]` after a table's name selects one row: `misc.country[SE]` and
+`[key]` or `[name]` after a table's name selects one row: `misc.territory[SE]` and
-`misc.country[Sweden]` name one row, and `misc.country[SE].capital` reads its column.
+`misc.territory[Sweden]` name one row, and `misc.territory[SE].capital` reads its column.
 A name naming several rows is an error listing their keys, unless a row selected
 before it settles which ([Linked tables](#linked-tables)). A selector is part of the path, so it works
-wherever a path does: `Fake`, `FakeRecord`, a `{/misc.country[SE].capital}` reference
+wherever a path does: `Fake`, `FakeRecord`, a `{/misc.territory[SE].capital}` reference
 and a struct tag. A dot inside the brackets belongs to the key or name, so
 `city[St. Louis]` selects it. A path starts with a name, and `[` still opens a JSON
 array at the start of a CLI argument, so `'[SE]'` alone names nothing.
@@ -505,21 +582,34 @@ stays reproducible.
 | `{ulid()}` | sample | ULID, 26 Crockford base32 chars |
 | `{nanoid(n)}` | sample | URL-safe Nano ID, `n` chars |
 | `{iban(CC)}` | sample | length- and mod-97-valid IBAN for BE, DE, DK, ES, FI, NO or SE |
+| `{date(from,to,'layout')}` | sample | a second between two `YYYY-MM-DD` days, both included, in a quoted Go layout: `'2006-01-02'`, `'January 2, 2006'`, `'060102'`, `'2006-01-02T15:04:05Z'` |
+| `{time('layout')}` | sample | a second within a day: `'15:04'`, `'3:04 PM'` |
 | `{seq()}`, `{seq(name)}` | counter | next integer from 1 in this generator; `name` selects an independent counter |
 | `{calc(expr)}`, `{calc(expr,dp)}` | computation | an arithmetic expression over sibling fields ([Computation](#computation)) |
 | `{lowercase(x)}`, `{uppercase(x)}`, `{ascii(x)}` | transform | a field's value rewritten ([Transforms](#transforms)) |
 
 A derivation reads what is to its left, so place it after its payload; the
 buffer is per expansion, so a nested template keeps fixed parts out of the sum. A
-Swedish personnummer is a Luhn checksum over the nine digits before it:
+Swedish personnummer is a Luhn checksum over the nine digits before it, six of
+them a birthdate:
 
 ```json
-{ "format": "{century}{core}", "century": ["19", "20"],
+{ "format": "{date(1930-01-01,2010-12-31,'060102')}-{birth}{luhn()}", "birth": ["238", "239"] }
-  "core": { "format": "{digits(2)}{mmdd}-{digits(3)}{luhn()}", "mmdd": ["0115", "0704", "1218"] } }
 ```
 
-Renders e.g. `19811218-9876`. `{seq()}` spans `Fake` calls and `repeat`, resets
+Renders e.g. `811218-2389`. A layout is Go's: the reference time `Mon Jan 2
-with a new generator, and is the natural primary key for the SQL example above.
+15:04:05 MST 2006` spelled as the output should look, quoted, since a layout may
+carry the comma that separates arguments, with English names. Every second
+between the two days is reachable, so a layout with a clock draws the time too, and
+`from` may equal `to`, which is that one day. The instant is UTC, so a zone in the
+layout prints `UTC` or `Z`.
+The quotes delimit a layout outside a selector only, so `[O'Fallon]` in an
+argument stays a name. Rejected at `New`: a bound that is no calendar date, or not
+before the other; an unquoted layout, naming the single-quoted one; a layout naming no field, which is
+text, as is one day in a layout with no clock; for `date` a layout naming no date
+field, naming `time`; and for `time` a layout naming a date field, naming `date`. `{seq()}` spans `Fake`
+calls and `repeat`, resets with a new generator, and is the natural primary key for
+the SQL example above.
 
 ### Computation
 
@@ -573,9 +663,9 @@ without naming `sv_SE`:
 
 Renders e.g. `Hej, Pat Smith!`. A reference path into a category is held like a
 [correlated](#correlated-fields) path, but for the whole render — one `Fake`, or one
-record — rather than one format: `{.person.femalefirst} {.person.last}` name one
+record — rather than one format: `{.person.first} {.person.last}` name one
 person, as do the same two references in sibling fields or a nested template, and
-`{lowercase(.person.femalefirst)}` reads that same draw. Each `repeat` iteration is
+`{lowercase(.person.first)}` reads that same draw. Each `repeat` iteration is
 a render of its own, in no group, so it draws anew, and a [draw group](#draw-group) holds a
 draw apart. A bare reference names no field and makes its own picks each time —
 `{/misc.uuid} {/misc.uuid}` is two draws — while the reference paths inside what it
@@ -594,8 +684,8 @@ its groups by name; the unnamed group spans them all.
 
 ```json
 { "format": "{payer} pays {payee}; signed {signature}",
-  "payer": { "format": "{/sv_SE.person.femalefirst} {/sv_SE.person.last}", "drawGroup": "payer" },
+  "payer": { "format": "{/sv_SE.person.first} {/sv_SE.person.last}", "drawGroup": "payer" },
-  "payee": "{/sv_SE.person.femalefirst} {/sv_SE.person.last}",
+  "payee": "{/sv_SE.person.first} {/sv_SE.person.last}",
   "signature": { "format": "{/sv_SE.person.last}", "drawGroup": "payer" } }
 ```
 
@@ -672,7 +762,7 @@ a minor only adds, and a major is the only release that changes what exists.
 | Surface | Major | Minor |
 |---------|-------|-------|
 | Shipped data | remove or rename a path; change a category's format; remove a value, or change a weight or a repeat; add a reference from one shipped category into another; change a table's key, name, weight or parent column, or remove a row | a path outside a record's columns, a locale, a value in a list, a row |
-| Records | remove, rename, retype or add a column; let a column be null | a record, as a new category |
+| Records (a table is one) | remove, rename, retype or add a column; let a column be null | a record, as a new category |
 | Data format | a fence: a spelling `New` rejects that it accepted; a template option, since it reserves a field name | a builtin |
 | CLI | remove or rename a flag, or change its default; change what an exit code means; change the framing a `--format` writes (header, quoting, statement shape), the `--list` layout, or what an error names | a flag, a format |
 | Library | change or remove an exported name; raise the lowest supported Go | an exported name, a `With…` option |
@@ -697,30 +787,71 @@ datatype and nullability, and each table's key, name, weight and parent columns;
 changes it or `data/` adds its `CHANGELOG.md` entry, which CI checks. A removed,
 renamed or retyped line is a major.
 
+## Audience
+
+App developers writing tests and fixtures, in Go and at a shell:
+
+- a **bulk fixture author**, thousands of rows into CSV or SQL
+- a **Go test author**, filling a struct with `FakeStruct`
+- a **hand fixture author**, one value at a shell
+- a **validator-facing author**, who needs a value a real checker accepts
+
+and a **contributor**, who reads [`todo.md`](todo.md), [`AGENTS.md`](AGENTS.md) and
+the Development section below, and who ships a register the four above then draw from.
+
 ## Goals
 
-1. **Valid by construction** — every value passes the check its real consumer
+1. **The best fake-data tool available, and used because it is** — a developer
+   choosing a seed-data library or CLI, in any language, should find this the one
+   worth switching to, and every goal below it is what that costs. Merit is only
+   half the goal: once the grammar, the data and the CLI are complete and stable,
+   it ships where those developers install from and is announced where they read.
+2. **Understanding the code is part of the product** — it is a compiler with static
+   fences behind a small API, it will grow, and the people who extend it will not be
+   the people who wrote it. So reading cost is judged alongside behaviour: reaching
+   the unit behind a symptom without asking a person, changing one piece without
+   holding the rest, names that tell the truth, and a file that stands with no second
+   document open beside it. A simulated reader panel scores those four, the score is a
+   ratchet no release lowers, and [`AGENTS.md`](AGENTS.md) says what each pull request
+   owes it.
+3. **Valid by construction** — every value passes the check its real consumer
    applies; facts that belong together come from one draw, within a value and
    across categories.
-2. **Text means what it says** — a format renders as written; only `{…}` varies,
+4. **Text means what it says** — a format renders as written; only `{…}` varies,
    random characters included (`{digits(3)}`). One spelling per result; the wrong
    one is a load error naming the right one.
-3. **Every mistake is a load error** — `New` rejects the data and `NewTemplate`
+5. **Every mistake is a load error** — `New` rejects the data and `NewTemplate`
    the inline template; on a loaded generator `Fake` fails only for an unknown
    path, `FakeStruct` only for a non-struct argument or a type its tags do not
    describe, with the same error every call, and `Template.Fake` cannot fail at all.
-4. **Zero to a value in one command** — `go install`, then `fejkdata sv_SE.person`:
+6. **Zero to a value in one command** — `go install`, then `fejkdata sv_SE.person`:
    no checkout, no flag. Flags are GNU-form (`--seed 42`, `-n 3`) in any position;
    the first custom template needs no escape and no option.
-5. **Data lives in JSON** — a builtin only for what data can't express.
+7. **Data lives in JSON** — a builtin only for what data can't express.
-6. **Reproducible** — seed in, same stream out; no builtin reads a clock.
+8. **Reproducible** — seed in, same stream out; no builtin reads a clock.
-7. **Zero dependencies** — standard library only.
+9. **Zero dependencies** — standard library only.
-8. **Docs index the grammar** — every syntax feature is a heading; every example
+10. **Docs index the grammar** — every syntax feature is a heading; every example
     runs under test and shows its output; a rule is stated once.
-9. **Fast enough to be free** — a value renders in about a microsecond and `New`
+11. **Fast enough to be free** — a value renders in about a microsecond and `New`
     parses and validates the whole set once upfront, so generating fixtures stays
     noise against a test's own runtime.
-
+12. **Data is sourced, or on its way there** — a shipped fact, a name, place,
+    code, id or classification, is read from a register or open dataset by a
+    [`data-import/`](data-import) script wherever one exists to read; where none
+    does yet a small hand-written set ships and [`todo.md`](todo.md) carries the
+    step that replaces it. Only non-factual copy stays authored. A sourced table
+    holds the rows its source holds: none is added by hand, and one is dropped
+    only by a rule the script states.
+13. **Breadth follows what most systems store** — a category is added in proportion
+    to how many real schemas hold it: names, addresses, phones, ids, money and
+    timestamps before anything domain-specific, and a catalogue serving one niche
+    waits behind everything serving many.
+14. **Realism is the default, inertness is selectable** — where a value could reach
+    something real, a domain anyone may register or an account a bank could issue,
+    the realistic breadth ships *and* so does the subset that provably reaches
+    nothing, each on its own path. A fixture that looks nothing like production
+    tests nothing; the caller who needs a value that can touch nothing asks for it
+    by name.
 ## Decisions
 
 - **Options and fields share one namespace.** `format`, `weight`, `repeat`,
@@ -728,8 +859,7 @@ renamed or retyped line is a major.
   key, or prefixing options, would tax every template to guard against a
   misspelt option.
 - **`{a|b}` stays beside nested choices.** `[[…], […]]` picks the same way, but
-  its arms are anonymous; `{femalefirst|malefirst}` keeps `person.femalefirst`
+  its arms are anonymous; `{female|male}` keeps `person.female` addressable.
-  addressable.
 - **Flags follow getopt_long.** `--name value` and `--name=value` both work; a
   short flag's value attaches or follows (`-s42`, `-s 42`) and short flags bundle
   (`-hn 3`), as every shell user expects. A single-dash long flag is rejected
@@ -747,7 +877,9 @@ renamed or retyped line is a major.
   library's own advice reachable: the error for an object holding only a format
   names `"…"`, and that spelling has to work where it is printed. An argument or struct
   tag of one reference alone, `{/users}`, is refused naming the path `users`: both
-  render the same text, and only the path names a record. `IsTemplate` exports the
+  render the same text, and only the path names a record. A folder-relative `{.name}`
+  or `{..name}` is refused naming `{/name}`, since an inline template sits in no
+  folder. `IsTemplate` exports the
   rule, so the CLI, struct tags and any other caller read one.
 - **An inline template skips the cycle fence.** `New` proves the loaded tree
   acyclic, an inline node is a finite tree of its own, and nothing in the tree can
@@ -972,8 +1104,7 @@ renamed or retyped line is a major.
   A table is a category with a TSV beside its file, so only a root choice has the
   spelling the fence names; a nested choice of same-shaped templates and an inline
   one keep loading. Fields must all be strings because a cell is a string node: a
-  choice whose items carry a nested choice, as `misc.car` does, is not one table but
+  choice whose items carry a nested choice is not one table but two linked ones.
-  two linked ones, which a later conversion writes.
 - **A table is a record of string columns.** Its columns are the CSV header and the
   `INSERT` column list, fixed by the TSV header, so a table is a record by
   construction; every column is a string until a typed column option earns its place.
@@ -1035,14 +1166,164 @@ renamed or retyped line is a major.
   holding most of its land inside places.** `I- 55 Bus` and `US Hwy 1` carry the
   most address ranges in many places and would head every address, so the import
   drops names spelled as a route. A ZCTA goes to the place its largest in-place part
-  lies in, and ships only when that place does; counting the land outside every
+  lies in, census-designated places left out since they never ship, and ships only
-  place too would drop a quarter of the places, whose codes straddle unincorporated
+  when that place does, so a few dozen places whose every code lies mostly in a
-  land, for a postal city the USPS mostly names the same way.
+  bigger neighbour ship no address; counting the land outside every place too would
+  drop a quarter of the places, whose codes straddle unincorporated land, for a
+  postal city the USPS mostly names the same way.
+- **`--list` stays a plain list of paths.** It is what a script reads, so every line
+  has to be a path that `Fake` takes; a marker for the tables a `[selector]` follows,
+  or a legend above them, would make the output something to parse before use.
+  `--help` names the selector spelling instead, and the Table section teaches it.
+- **A layout is always quoted.** A layout may carry the comma that separates
+  arguments, `'January 2, 2006'`, and one spelling for every layout beats a rule
+  about which ones need the quotes, so the bare spelling is refused naming the
+  quoted one. The layout is Go's reference time because the library renders with
+  it and a Go caller already knows it; its names are English, and a locale's own
+  month and weekday names are data.
+- **A title is a table under `sex`.** A prefix drawn apart would put `Mr` on a record
+  whose `sex` column says `female`, which is the disagreement the record exists to
+  prevent; the tables this set already has are what a title needs, so `en_US.title`
+  links to `sex` as `first-name` does. Swedish has no everyday sexed honorific, so
+  `sv_SE.title` is a table as well but carries no `parent`. Its weight column is
+  `share`, not the `count` a name table carries, because the values are a curated
+  proportion rather than bearers anyone counted.
+- **A table owns the spelling of a selector on it.** A reference reaches a table by a
+  path that carries no selector — `sv_SE.person.first` reads `first-name` through
+  `sex` — so the walk that resolved a name cannot say where a reader would type one.
+  The table's own location can, which is why it keeps its path, and why an ambiguity
+  error names `sv_SE.sex[f].first-name[Kim]` rather than the table's own name.
+- **No builtin reads the clock, so a date is bounded by days, never by an age.**
+  An `age(min,max)` would make a seeded fixture change with the day it runs on,
+  which is what a seed exists to prevent; a birthdate for someone 20 to 60 is
+  `date(1966-01-01,2006-12-31,…)`, re-pinned as any fixture is.
+- **A name column without a key resolves inside its parent.** A given name both
+  sexes carry is a row under each, so `name` cannot be the key; the parent's row
+  tells the two apart, `sex[f].first-name[Kim]`, the ambiguity error spells each
+  row inside its parent, and a name repeating inside one parent row is refused at
+  load, since nothing could then select it.
+- **`misc` is what every locale shares.** A category whose facts differ by country
+  belongs in that country's locale, read from the register that country's own
+  records use; `misc` takes only sources that are international. NHTSA vPIC and
+  Mobility Sweden's registrations are national, so they build `en_US.car` and
+  `sv_SE.car`; `misc.car` waits for an international source rather than take one
+  of theirs.
+- **A register's canonical spelling loses to the one its domain writes.** Where a
+  source offers several spellings of one fact, the shipped one is what records in
+  that domain carry. `misc.timezone` reads `zone.tab` and not the `zone1970.tab`
+  that supersedes it, because the latter keeps one zone per set of countries that
+  have agreed since 1970: it spells Sweden `Europe/Berlin`, and no Swedish system
+  writes that. For the same reason `misc.language` takes the ISO 639-2 register's
+  first synonym over CLDR, which says "Chinese, Mandarin" for `zh`.
+- **`misc.territory` is the spine, and a `misc` table naming a territory links to
+  it.** Where every territory row has a child the column is a `parent`, and the
+  import drops the child rows whose territory the set does not ship — 17 of
+  `misc.timezone`'s, Antarctica's ten among them. Agreement across a record is
+  worth more than the last rows of a table. Where no such link can hold the fact
+  stays a column. Layer your own `misc.territory` over the shipped one and you
+  must layer `misc.timezone` too, or the link fails at load naming the row.
+- **A table whose register publishes no frequency draws evenly.** `misc.httpmethod`,
+  `misc.port`, `misc.httpstatus`, `misc.mimetype` and `misc.tld` weigh every row alike,
+  so GET is a ninth of the methods drawn. Goal 10 keeps an authored fact out of a
+  sourced table, and no register publishes how often a method, a port or a TLD is used,
+  so a weight here would be invented. Where one exists it is read, as `misc.timezone`
+  reads GeoNames populations and `sv_SE.first-name` SCB bearers.
+- **`misc.port` selects by number, and carries no `name`.** 29 of its services sit on
+  more than one port, `http-alt` on three, so `misc.port[http-alt]` could name no one
+  row. The number is what a port field holds anyway, and `misc.port[443].service`
+  reads the other way.
+- **`misc.car` is one flat table, not a make linked to its models.** A row is a make
+  and a model drawn together, so no render pairs a Volvo with a RAV4. Two linked
+  tables would reach the same pairs and add a selector nothing asks for; a make alone
+  is `misc.car.make`.
+- **`misc.territory` names its sovereign in a column, and there is no `misc.country`
+  table.** ISO 3166-1 codes territories, so `territory` is the honest name, and
+  `is_independent` in the register gives each one's state. A second table of the 195
+  sovereigns would hold a `DK` row beside the territory `DK` row, both carrying
+  Denmark's capital, currency, flag and TLD — two owners for one fact, drifting at
+  the next import. Splitting the columns to avoid that is worse: put `capital` on
+  the territory alone and a country row can no longer name Copenhagen. The column
+  cannot be a `parent`, since a table never reaches its own family; a test proves
+  every value names a row instead. A territory the register records no state for
+  stands alone, which is `EH` alone, and naming one for it would be a claim
+  fejkdata has no business making.
+- **`misc.loglevel` is a table keyed by the code, rendering POSIX's keyword.** A flat
+  list of names carries neither the code a PRI encodes nor a selector reaching it, and
+  goal 3 draws the two as one fact. The canonical spelling losing to the one its domain
+  writes, above, settles the rest: a configuration writes `info`, so RFC 5424's
+  `Informational` stays the `severity` column, and the code and the keyword hold the
+  two selector slots.
+- **`misc.tld` keys carry the leading dot, where other tables key on a bare code.**
+  The register spells a TLD `.se` and `misc.territory.tld` already ships it so, which a
+  bare key would make two spellings of one fact; `{/misc.tld}` also composes onto a
+  host with no separator. `misc.tld[se]` misses for it, which `todo.md` carries.
+- **`misc.tld` is a table of its own, and `misc.territory.tld` stays a column.** A
+  `parent` demands a child for every parent row, so linking them would drop every root
+  zone row naming no territory, which is most of them, and goal 12 holds a sourced
+  table whole. The loader refuses the link outright anyway: `tld` is a column of
+  `misc.territory`, and a table may not be named like a column of its ancestor.
+- **`misc.territory` carries a currency code, it does not link to `misc.currency`.**
+  A `parent` demands a child for every parent row, and ISO 4217 registers codes no
+  country's row can name: the funds codes (Mvdol, WIR Euro, US Dollar (Next day)),
+  and VED beside VES, both Venezuela's, of which a country row names one. Linking
+  would trade the register for the link, and `misc.territory.currency` already pairs a
+  country with its currency in one draw.
+- **An extension may name two media types.** `.xml`, `.rtf`, `.sub`, `.mpp` and `.ac`
+  each name two rows of `misc.mimetype`. Separating them would mean dropping a
+  registered type, or naming one by an extension that is not its own — `.mpt` is
+  Project's template, not its document. Selecting such a name is an error listing
+  both keys; select by type instead.
+- **The Swedish ids draw Skatteverket's test series.** A Luhn-valid personnummer
+  over a random birth number may be a living person's; 238 and 239 after any date
+  are blocked from assignment, so the shipped `personnummer` and
+  `samordningsnummer` use those. They sit in a `birth-number` table under `sex`
+  rather than as a column of it: the render's shared draw of the family is what
+  makes the number and the name agree on sex, and `sex` stays one shape across
+  locales instead of collecting every sex-keyed id fact. A samordningsnummer's
+  day, the birthday plus 60, is drawn from 61 to 88, valid in every month, rather
+  than computed from the date drawn.
+- **The US given names come from a mirror of the SSA file.** ssa.gov refuses a
+  client outside the US, so `names-us.py` reads a GitHub copy that ends at 2020,
+  which a count over the births since 1930 barely feels; `--names` takes the
+  official zip. The SSA's placeholder rows are top-1000 entries that name nobody, so
+  the import drops them by name rather than by a rank a regeneration would move.
 - **`List` advertises direct descents only.** `region.municipality.locality` is
   listed, and `region.locality` resolves too but is not: the set of every descent
   through a chain of five tables is every subsequence of it, and the direct chain is
   the one a reader can predict from the tables' parents.
 
+## Layout
+
+```
+fejkdata.go     Generator, New, options, the embedded data set, List
+node.go         the node model and JSON -> node compilation
+table.go        tables: the rows TSV, its options and links, row selection and draws
+path.go         the dotted-path walk with its selectors, and proving a path resolves
+render.go       Fake and the recursive renderer (choices, format strings, expansions)
+record.go       records: Record, the JSON/CSV/SQL serializers, and their entry points
+struct.go       structs: FakeStruct, fake tags, and a field's Go type as its column's datatype
+inline.go       inline templates: Template, NewTemplate, FakeTemplate, IsTemplate, and their compile and link
+template.go     the {token} grammar: scanning, tokens, operands, validation, compiling a format
+hold.go         the hold: one draw per expansion for paths and operands, and its fences
+draw.go         one reference draw per render and group: draw sets, the group option, and its fence
+family.go       a family of linked tables: the rows a render pins, and the fence over paths into one family
+reference.go    reference sigils, and binding references across the tree
+graph.go        the render graph: edges, cycles, the repeat bound, tree walks
+builtins.go     the {name()} function registry and its implementations
+layout.go       date and time layouts: the instants one is proved against, and the two samples
+checksum.go     the check characters a derivation appends, and the IBAN they sit inside
+transform.go    the builtins that rewrite an operand's value, and the ASCII folding
+calc.go         the {calc()} arithmetic evaluator: parser, eval, validation
+datatype.go     column datatypes: DataType, where datatype and null may sit, a column's datatype
+value.go        the value proof: what a typed column or calc operand holds, checked at load
+data.go         data loading: fs.FS folders/files -> namespace tree, multi-source merge
+cmd/fejkdata/   the fejkdata CLI
+data/           shipped data (JSON, and a TSV per table), embedded at build: locale folders, geo, misc
+data-import/    the scripts that rebuild each sourced table (see DATA-LICENSES.md)
+release-tooling/ the release CI publishes from the changelog heading
+testdata/       the pinned shipped shape (see Versioning)
+```
+
 ## Development
 
 Everything runs in Docker — **no local tooling beyond Docker is needed**.
@@ -1065,6 +1346,15 @@ docker compose run --rm --user "$(id -u):$(id -g)" fmt   # gofmt -w .
 docker compose run --rm --user "$(id -u):$(id -g)" tidy  # go mod tidy
 ```
 
+A `--user` command answering `permission denied` on `/cache` has met root-owned
+files in the `gocache` volume: Docker creates the volume root-owned, and every
+command here that omits `--user` writes into it as root. Hand it back, and again
+whenever it recurs:
+
+```sh
+docker compose run --rm --user root --entrypoint chown test -R "$(id -u):$(id -g)" /cache
+```
+
 Every pull request runs `docker build .` against both the latest and the lowest
 supported Go, and must pass before it can be merged — unless it changes none of
 the files the build and its tests read, nor the workflow itself, in which case
@@ -1088,16 +1378,32 @@ REPIN=1 docker compose run --rm --user "$(id -u):$(id -g)" test
 A shipped table built from a source is rebuilt by its script under
 [`data-import/`](data-import), one command per dataset, fetching the source named in
 [`DATA-LICENSES.md`](DATA-LICENSES.md). Downloads are cached under
-`data-import/cache/`, so delete it to fetch afresh; `geo-us.py` fetches two
+`data-import/cache/`, which is ignored by version control and grows past a
-TIGER/Line files per county it ships, a few hundred megabytes, and `geo-se.py` needs
+gigabyte, so pass `--cache DIR` to reuse a copy you already have and delete it to
+fetch afresh; `geo-us.py` fetches two
+TIGER/Line files per county it ships, a few hundred megabytes, `geo-se.py` needs
 a Trafikverket API key, free at [data.trafikverket.se](https://data.trafikverket.se/),
-in `TRAFIKVERKET_API_KEY` or a `--key-file`:
+in `TRAFIKVERKET_API_KEY` or a `--key-file`, and the Census host behind `geo-us.py`
+and `names-us.py` rejects a client for a while after a burst, so `--surnames` takes
+a copy of the surname file:
 
 ```sh
-docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/country.py
 docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/currency.py
-docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/geo-us.py
 docker compose run --rm --user "$(id -u):$(id -g)" -e TRAFIKVERKET_API_KEY data-import data-import/geo-se.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/geo-us.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/httpmethod.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/httpstatus.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/language.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/loglevel.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/mimetype.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/names-se.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/names-us.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/port.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/protocol.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/territory.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/timezone.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/tld.py
+docker compose run --rm --user "$(id -u):$(id -g)" data-import data-import/useragent.py
 ```
 
 To release, head `CHANGELOG.md` with the version's section in place of `Unreleased`
@@ -1105,35 +1411,6 @@ and merge: once `main` passes the gate, CI tags that commit `vX.Y.Z` and publish
 the Gitea release with the section as its body. A top heading of `[Unreleased]`
 publishes nothing.
 
-## Layout
-
-```
-fejkdata.go     Generator, New, options, the embedded data set, List
-node.go         the node model and JSON -> node compilation
-table.go        tables: the rows TSV, its options and links, row selection and draws
-path.go         the dotted-path walk with its selectors, and proving a path resolves
-render.go       Fake and the recursive renderer (choices, format strings, expansions)
-record.go       records: Record, the JSON/CSV/SQL serializers, and their entry points
-struct.go       structs: FakeStruct, fake tags, and a field's Go type as its column's datatype
-inline.go       inline templates: Template, NewTemplate, FakeTemplate, IsTemplate, and their compile and link
-template.go     the {token} grammar: scanning, tokens, operands, validation, compiling a format
-hold.go         the hold: one draw per expansion for paths and operands, and its fences
-draw.go         one reference draw per render and group: draw sets, the group option, and its fence
-family.go       a family of linked tables: the rows a render pins, and the fence over paths into one family
-reference.go    reference sigils, and binding references across the tree
-graph.go        the render graph: edges, cycles, the repeat bound, tree walks
-builtins.go     the {name()} function registry and its implementations
-calc.go         the {calc()} arithmetic evaluator: parser, eval, validation
-datatype.go     column datatypes: DataType, where datatype and null may sit, a column's datatype
-value.go        the value proof: what a typed column or calc operand holds, checked at load
-data.go         data loading: fs.FS folders/files -> namespace tree, multi-source merge
-cmd/fejkdata/   the fejkdata CLI
-data/           shipped data (JSON, and a TSV per table), embedded at build: locale folders, geo, misc
-data-import/    the scripts that rebuild each sourced table (see DATA-LICENSES.md)
-release-tooling/ the release CI publishes from the changelog heading
-testdata/       the pinned shipped shape (see Versioning)
-```
-
 ## License
 
 MIT — see [LICENSE](LICENSE). Forked from [github.com/Timewave-AB/fakes](https://github.com/Timewave-AB/fakes).

bench_test.go

@@ -30,7 +30,7 @@ func BenchmarkPerson(b *testing.B)     { benchPath(b, "data", "sv_SE.person") }
 func BenchmarkAddress(b *testing.B)      { benchPath(b, "data", "sv_SE.address") }
 func BenchmarkWord(b *testing.B)         { benchPath(b, "data", "sv_SE.word") }
 func BenchmarkCreditcard(b *testing.B)   { benchPath(b, "data", "misc.creditcard") }
-func BenchmarkSSN(b *testing.B)        { benchPath(b, "data", "sv_SE.ssn") }
+func BenchmarkPersonnummer(b *testing.B) { benchPath(b, "data", "sv_SE.personnummer") }
 func BenchmarkUUIDv7(b *testing.B)       { benchPath(b, "data", "misc.uuid") }
 
 func tmpData(b *testing.B, name, body string) string {

builtins.go

@@ -7,7 +7,6 @@ import (
 	"math"
 	"strconv"
 	"strings"
-	"unicode"
 )
 
 // maxLen caps sample output lengths (hex, nanoid, base64, digits, upper, lower)
@@ -60,6 +59,8 @@ var builtins = map[string]builtin{
 		cc := a[0]
 		return func(s *session, _ string, _ []string) string { return iban(s, cc) }
 	}},
+	"date":      {arity: -1, check: dateArgs, prep: datePrep},
+	"time":      {arity: -1, check: timeArg, prep: timePrep},
 	"calc":      {arity: -1, check: checkCalc, prep: calcPrep, operands: calcOperands},
 	"lowercase": {arity: 1, check: transformArg, prep: transformPrep(strings.ToLower), operands: transformOperand},
 	"uppercase": {arity: 1, check: transformArg, prep: transformPrep(strings.ToUpper), operands: transformOperand},
@@ -88,13 +89,11 @@ func derive(f func(emitted string) string) func([]string) callFn {
 		return func(_ *session, emitted string, _ []string) string { return f(emitted) }
 	}
 }
-
 func sample(f func(rng) string) func([]string) callFn {
 	return func([]string) callFn {
 		return func(s *session, _ string, _ []string) string { return f(s) }
 	}
 }
-
 func chars(alphabet string) func([]string) callFn {
 	return func(a []string) callFn {
 		n := atoi(a[0])
@@ -113,96 +112,6 @@ func formatFloat(v float64, dp int) string {
 	return s
 }
 
-// transforms are the builtins that rewrite one operand's value; they nest, so
-// {lowercase(ascii(x))} folds then lowers.
-var transforms = map[string]func(string) string{
-	"ascii":     asciiFold,
-	"lowercase": strings.ToLower,
-	"uppercase": strings.ToUpper,
-}
-
-// unwrapTransform peels nested transform calls off an operand arg, returning the
-// field it finally names and the transforms to apply, innermost last.
-func unwrapTransform(arg string) (leaf string, chain []func(string) string, err error) {
-	for {
-		name, args, isCall := funcCall(arg)
-		if !isCall {
-			return arg, chain, nil
-		}
-		fn, isTransform := transforms[name]
-		if !isTransform {
-			return "", nil, fmt.Errorf("%s(%s) is not a transform, so it cannot be an operand", name, strings.Join(args, ","))
-		}
-		if len(args) != 1 {
-			return "", nil, fmt.Errorf("%s takes 1 arg, got %d", name, len(args))
-		}
-		chain = append(chain, fn)
-		arg = args[0]
-	}
-}
-
-func transformArg(fields map[string]node, a []string) error {
-	leaf, _, err := unwrapTransform(a[0])
-	if err != nil {
-		return err
-	}
-	if isRef(leaf) {
-		_, _, err := refShape(leaf)
-		return err
-	}
-	return checkArm(leaf, fields, false)
-}
-
-func transformOperand(a []string) []string {
-	leaf, _, err := unwrapTransform(a[0])
-	if err != nil {
-		return nil
-	}
-	return []string{leaf}
-}
-
-func transformPrep(outer func(string) string) func([]string) callFn {
-	return func(a []string) callFn {
-		_, chain, err := unwrapTransform(a[0])
-		if err != nil {
-			panic(fmt.Sprintf("fejkdata: transform arg %q reached prep unvalidated: %v", a[0], err))
-		}
-		return func(_ *session, _ string, operands []string) string {
-			v := operands[0]
-			for i := len(chain) - 1; i >= 0; i-- {
-				v = chain[i](v)
-			}
-			return outer(v)
-		}
-	}
-}
-
-// asciiFolds maps the Latin letters with diacritics or ligatures to ASCII.
-var asciiFolds = map[rune]string{
-	'À': "A", 'Á': "A", 'Â': "A", 'Ã': "A", 'Ä': "A", 'Å': "A", 'Æ': "AE", 'Ç': "C",
-	'È': "E", 'É': "E", 'Ê': "E", 'Ë': "E", 'Ì': "I", 'Í': "I", 'Î': "I", 'Ï': "I",
-	'Ð': "D", 'Ñ': "N", 'Ò': "O", 'Ó': "O", 'Ô': "O", 'Õ': "O", 'Ö': "O", 'Ø': "O",
-	'Ù': "U", 'Ú': "U", 'Û': "U", 'Ü': "U", 'Ý': "Y", 'Þ': "Th", 'ß': "ss", 'Œ': "OE",
-	'à': "a", 'á': "a", 'â': "a", 'ã': "a", 'ä': "a", 'å': "a", 'æ': "ae", 'ç': "c",
-	'è': "e", 'é': "e", 'ê': "e", 'ë': "e", 'ì': "i", 'í': "i", 'î': "i", 'ï': "i",
-	'ð': "d", 'ñ': "n", 'ò': "o", 'ó': "o", 'ô': "o", 'õ': "o", 'ö': "o", 'ø': "o",
-	'ù': "u", 'ú': "u", 'û': "u", 'ü': "u", 'ý': "y", 'þ': "th", 'ÿ': "y", 'œ': "oe",
-}
-
-// asciiFold rewrites s to ASCII: folded Latin letters stay, any other non-ASCII
-// rune is dropped.
-func asciiFold(s string) string {
-	var b strings.Builder
-	for _, r := range s {
-		if r <= unicode.MaxASCII {
-			b.WriteRune(r)
-		} else {
-			b.WriteString(asciiFolds[r])
-		}
-	}
-	return b.String()
-}
-
 // atoi parses an arg a builtin's check already validated. It panics rather than
 // returning zero, so a check that stops covering its own args is a stack trace and
 // not a silently wrong length, range or decimal count.
@@ -222,7 +131,6 @@ func atof(s string) float64 {
 	}
 	return f
 }
-
 func randBytes(r rng, n int) []byte {
 	b := make([]byte, n)
 	for i := range b {
@@ -230,7 +138,6 @@ func randBytes(r rng, n int) []byte {
 	}
 	return b
 }
-
 func randChars(r rng, n int, alphabet string) string {
 	b := make([]byte, n)
 	for i := range b {
@@ -253,7 +160,6 @@ func plainInt(s string) (int, error) {
 	}
 	return n, nil
 }
-
 func posIntArg(_ map[string]node, a []string) error {
 	n, err := plainInt(a[0])
 	if errors.Is(err, strconv.ErrRange) {
@@ -270,7 +176,6 @@ func posIntArg(_ map[string]node, a []string) error {
 	}
 	return nil
 }
-
 func intRangeArgs(_ map[string]node, a []string) error {
 	lo, err := plainInt(a[0])
 	if err != nil {
@@ -291,7 +196,6 @@ func intRangeArgs(_ map[string]node, a []string) error {
 	}
 	return nil
 }
-
 func floatArgs(_ map[string]node, a []string) error {
 	lo, e1 := strconv.ParseFloat(a[0], 64)
 	hi, e2 := strconv.ParseFloat(a[1], 64)
@@ -319,10 +223,9 @@ func floatArgs(_ map[string]node, a []string) error {
 	}
 	return nil
 }
-
 func seqArg(_ map[string]node, a []string) error {
 	if len(a) > 1 {
-		return fmt.Errorf("seq takes at most one name, got %d args", len(a))
+		return fmt.Errorf("seq takes at most one name, got %d", len(a))
 	}
 	if len(a) == 1 && a[0] == "" {
 		return fmt.Errorf("seq name must not be empty")
@@ -373,96 +276,3 @@ func ulid(r rng) string {
 	}
 	return string(out)
 }
-
-// luhnCheck returns the Luhn check digit (0-9) over the digits of s; non-digit
-// runes are skipped. Doubling runs from the rightmost digit, so the result is
-// correct whatever the payload length.
-func luhnCheck(s string) int {
-	sum, double := 0, true
-	for i := len(s) - 1; i >= 0; i-- {
-		c := s[i]
-		if c < '0' || c > '9' {
-			continue
-		}
-		d := int(c - '0')
-		if double {
-			if d *= 2; d > 9 {
-				d -= 9
-			}
-		}
-		double = !double
-		sum += d
-	}
-	return (10 - sum%10) % 10
-}
-
-// mod11Check returns the weighted mod-11 check character over the digits of s
-// (weights 2..7 cycling from the right). A would-be value of 10 emits 'X', as in
-// ISBN-10 / ISO 7064; non-digits are skipped.
-func mod11Check(s string) string {
-	sum, w := 0, 2
-	for i := len(s) - 1; i >= 0; i-- {
-		c := s[i]
-		if c < '0' || c > '9' {
-			continue
-		}
-		sum += int(c-'0') * w
-		if w++; w > 7 {
-			w = 2
-		}
-	}
-	if chk := (11 - sum%11) % 11; chk != 10 {
-		return string(rune('0' + chk))
-	}
-	return "X"
-}
-
-// eanCheck returns the EAN-13 / UPC-A / ISBN-13 / GTIN check digit over the
-// digits of s: weights 3 and 1 alternating from the rightmost digit, mod 10.
-func eanCheck(s string) string {
-	sum, w := 0, 3
-	for i := len(s) - 1; i >= 0; i-- {
-		c := s[i]
-		if c < '0' || c > '9' {
-			continue
-		}
-		sum += int(c-'0') * w
-		w = 4 - w // 3 <-> 1
-	}
-	return string(rune('0' + (10-sum%10)%10))
-}
-
-// ibanLen maps a supported country code to the full IBAN length. The check digits
-// sit between the country code and the BBAN, so — unlike luhn/ean — iban can't be
-// a left-to-right derivation; it generates the whole value instead.
-var ibanLen = map[string]int{"BE": 16, "DE": 22, "DK": 18, "ES": 24, "FI": 18, "NO": 15, "SE": 24}
-
-func ibanArg(_ map[string]node, a []string) error {
-	if _, ok := ibanLen[a[0]]; !ok {
-		return fmt.Errorf("iban(%q): unsupported country code", a[0])
-	}
-	return nil
-}
-
-// iban generates a structurally valid IBAN for cc: a numeric BBAN of the right
-// length, then mod-97 check digits. Real bank/branch structure isn't modelled —
-// the result passes length and checksum validation, which is what fake data needs.
-func iban(r rng, cc string) string {
-	bban := make([]byte, ibanLen[cc]-4)
-	for i := range bban {
-		bban[i] = byte('0' + r.IntN(10))
-	}
-	rem := 0
-	feed := func(d int) { rem = (rem*10 + d) % 97 }
-	for _, c := range bban {
-		feed(int(c - '0'))
-	}
-	for i := 0; i < len(cc); i++ { // letters A-Z -> 10..35, fed as two digits
-		v := int(cc[i]-'A') + 10
-		feed(v / 10)
-		feed(v % 10)
-	}
-	feed(0)
-	feed(0)
-	return fmt.Sprintf("%s%02d%s", cc, 98-rem, bban)
-}

builtins_test.go

@@ -4,7 +4,9 @@ import (
 	"encoding/base64"
 	"regexp"
 	"strconv"
+	"strings"
 	"testing"
+	"time"
 )
 
 func TestBuiltinIDGenerators(t *testing.T) {
@@ -41,6 +43,7 @@ func TestBuiltinSamplesReproducible(t *testing.T) {
 		`"{uuid()}"`, `"{ulid()}"`,
 		`"{nanoid(12)}"`, `"{int(1,1000000)}"`,
 		`"{float(0,1,6)}"`, `"{base64(12)}"`, `"{iban(SE)}"`,
+		`"{date(2000-01-01,2020-12-31,'2006-01-02 15:04:05')}"`, `"{time('15:04:05')}"`,
 	} {
 		if a, b := mustRender(t, engine(7), tmpl), mustRender(t, engine(7), tmpl); a != b {
 			t.Fatalf("%s not reproducible: %q != %q", tmpl, a, b)
@@ -233,3 +236,135 @@ func TestClassBuiltinArgs(t *testing.T) {
 		}
 	}
 }
+
+// TestBuiltinDateAndTime pins the two clock-free samples: date draws a second in
+// [from 00:00:00, to 23:59:59], both days reachable, and renders it in the quoted Go
+// layout, commas and English names included; time draws a second within one day.
+func TestBuiltinDateAndTime(t *testing.T) {
+	f := engine(1)
+	seen := map[string]bool{}
+	for i := 0; i < 500; i++ {
+		got := mustRender(t, f, `"{date(1990-01-01,1990-12-31,'2006-01-02')}"`)
+		if d, err := time.Parse("2006-01-02", got); err != nil || d.Year() != 1990 {
+			t.Fatalf("date = %q, want a 1990 calendar date (err %v)", got, err)
+		}
+		seen[got] = true
+	}
+	if len(seen) < 200 {
+		t.Fatalf("date drew %d distinct days of 365 in 500, want a uniform spread", len(seen))
+	}
+	lo, hi := false, false
+	for i := 0; i < 200; i++ {
+		got := mustRender(t, f, `"{date(2020-02-28,2020-02-29,'2006-01-02')}"`)
+		if got != "2020-02-28" && got != "2020-02-29" {
+			t.Fatalf("date(2020-02-28,2020-02-29) = %q, out of range", got)
+		}
+		lo, hi = lo || got == "2020-02-28", hi || got == "2020-02-29"
+	}
+	if !lo || !hi {
+		t.Fatalf("date never hit a bound: lo=%v hi=%v (bounds must be inclusive)", lo, hi)
+	}
+	if got := mustRender(t, f, `"{date(2020-07-04,2020-07-05,'January 2, 2006')}"`); got != "July 4, 2020" && got != "July 5, 2020" {
+		t.Fatalf("date with a comma in its layout = %q", got)
+	}
+	rfc := regexp.MustCompile(`^2021-\d\d-\d\dT\d\d:\d\d:\d\dZ$`)
+	clock := regexp.MustCompile(`^([01]\d|2[0-3]):[0-5]\d$`)
+	ampm := regexp.MustCompile(`^(1[0-2]|[1-9]):[0-5]\d (AM|PM)$`)
+	seconds := map[string]bool{}
+	for i := 0; i < 300; i++ {
+		got := mustRender(t, f, `"{date(2021-01-01,2021-12-31,'2006-01-02T15:04:05Z07:00')}"`)
+		if !rfc.MatchString(got) {
+			t.Fatalf("date in an RFC 3339 layout = %q, want %s", got, rfc)
+		}
+		seconds[got[17:19]] = true
+		if got := mustRender(t, f, `"{time('15:04')}"`); !clock.MatchString(got) {
+			t.Fatalf("time('15:04') = %q, want %s", got, clock)
+		}
+		if got := mustRender(t, f, `"{time('3:04 PM')}"`); !ampm.MatchString(got) {
+			t.Fatalf("time('3:04 PM') = %q, want %s", got, ampm)
+		}
+		got = mustRender(t, f, `"{date(1950-01-01,2000-12-31,'060102')}-238{luhn()}"`)
+		if d := digitsOnly(got); len(d) != 10 || !luhnValid(d) {
+			t.Fatalf("a personnummer over date() = %q, want ten Luhn-valid digits", got)
+		}
+	}
+	if len(seconds) < 30 {
+		t.Fatalf("date drew %d distinct seconds in 300, want the whole day, not midnight", len(seconds))
+	}
+}
+
+// TestBuiltinDateArgs pins the New-time checks: bounds are calendar dates in order,
+// the layout is quoted, names a field, and for time names no date field.
+func TestBuiltinDateArgs(t *testing.T) {
+	for tmpl, want := range map[string]string{
+		`"{date(1990-13-01,1990-12-31,'2006-01-02')}"`:        "1990-13-01",
+		`"{date(1990-12-31,1990-01-01,'2006-01-02')}"`:        "is after",
+		`"{date(1990-01-01,1990-01-01,'2006-01-02')}"`:        "write it as text",
+		`"{date(1990-01-01,1990-12-31,2006-01-02)}"`:          "'2006-01-02'",
+		`"{date(1990-01-01,1990-12-31,'January 2, 2006)}"`:    "'",
+		`"{date(1990-01-01,1990-12-31,\"January 2, 2006\")}"`: "quoted: 'January 2, 2006'",
+		`"{date(1990-01-01,1990-12-31,'x')}"`:                 "text",
+		`"{date(1990-01-01,1990-12-31,'')}"`:                  "text",
+		`"{date(1990-01-01,1990-12-31)}"`:                     "3 arguments",
+		`"{date(1990-01-01,1990-12-31,January 2, 2006)}"`:     "'January 2, 2006'",
+		`"{time(3:04 PM, Mon)}"`:                              "'3:04 PM, Mon'",
+		`"{time(15:04)}"`:                                     "'15:04'",
+		`"{time('2006-01-02 15:04')}"`:                        "date(",
+		`"{time('x')}"`:                                       "text",
+		`"{date(1990-01-01,1990-12-31,'15:04')}"`:             "time('15:04')",
+	} {
+		_, err := compile(parse(t, tmpl))
+		if err == nil || !strings.Contains(err.Error(), want) {
+			t.Errorf("compile(%s) = %v, want an error mentioning %q", tmpl, err, want)
+		}
+	}
+}
+
+// TestBuiltinLayoutErrorsNameARunnableSpelling pins the two layout errors a user
+// can follow: a double-quoted layout is named single-quoted, without its own
+// quotes carried into the suggestion, whether it split on a comma or not.
+func TestBuiltinLayoutErrorsNameARunnableSpelling(t *testing.T) {
+	_, err := compile(parse(t, `"{date(1990-01-01,1990-12-31,\"2006-01-02\")}"`))
+	if err == nil || !strings.Contains(err.Error(), "write '2006-01-02'") {
+		t.Fatalf("a double-quoted layout = %v, want it named single-quoted", err)
+	}
+	if strings.Contains(err.Error(), `'"`) {
+		t.Errorf("%v names a layout that renders its own quotes", err)
+	}
+	if !strings.Contains(err.Error(), "is double-quoted") {
+		t.Errorf("%v does not say which quotes were wrong", err)
+	}
+	_, err = compile(parse(t, `"{time(15:04)}"`))
+	if err == nil || !strings.Contains(err.Error(), "write '15:04'") {
+		t.Fatalf("a bare layout = %v, want it named quoted", err)
+	}
+	// The comma hint belongs to a layout that split, not to a call given extra args.
+	_, err = compile(parse(t, `"{time(0,12,'15:04')}"`))
+	if err == nil || !strings.Contains(err.Error(), "takes 1 argument, got 3") {
+		t.Fatalf("time with three args = %v, want the count named in the singular", err)
+	}
+	if strings.Contains(err.Error(), "holding a comma") {
+		t.Errorf("%v offers the comma hint though the layout is already quoted", err)
+	}
+}
+
+// TestBuiltinDateSpansOneDay pins that from == to is a day: with a clock layout it
+// draws every second of it, and without one it could only emit one value.
+func TestBuiltinDateSpansOneDay(t *testing.T) {
+	f := engine(1)
+	seen := map[string]bool{}
+	for i := 0; i < 500; i++ {
+		got := mustRender(t, f, `"{date(2026-01-01,2026-01-01,'2006-01-02 15:04:05')}"`)
+		if !strings.HasPrefix(got, "2026-01-01 ") {
+			t.Fatalf("date over one day = %q, out of range", got)
+		}
+		seen[got] = true
+	}
+	if len(seen) < 400 {
+		t.Fatalf("date over one day drew %d distinct seconds in 500, want the whole day", len(seen))
+	}
+	_, err := compile(parse(t, `"{date(2026-01-01,2026-01-01,'2006-01-02')}"`))
+	if err == nil || !strings.Contains(err.Error(), "write it as text") {
+		t.Fatalf("one day in a date-only layout = %v, want it named a constant", err)
+	}
+}

checksum.go

@@ -0,0 +1,96 @@
+package fejkdata
+
+import "fmt"
+
+// luhnCheck returns the Luhn check digit (0-9) over the digits of s; non-digit
+// runes are skipped. Doubling runs from the rightmost digit, so the result is
+// correct whatever the payload length.
+func luhnCheck(s string) int {
+	sum, double := 0, true
+	for i := len(s) - 1; i >= 0; i-- {
+		c := s[i]
+		if c < '0' || c > '9' {
+			continue
+		}
+		d := int(c - '0')
+		if double {
+			if d *= 2; d > 9 {
+				d -= 9
+			}
+		}
+		double = !double
+		sum += d
+	}
+	return (10 - sum%10) % 10
+}
+
+// mod11Check returns the weighted mod-11 check character over the digits of s
+// (weights 2..7 cycling from the right). A would-be value of 10 emits 'X', as in
+// ISBN-10 / ISO 7064; non-digits are skipped.
+func mod11Check(s string) string {
+	sum, w := 0, 2
+	for i := len(s) - 1; i >= 0; i-- {
+		c := s[i]
+		if c < '0' || c > '9' {
+			continue
+		}
+		sum += int(c-'0') * w
+		if w++; w > 7 {
+			w = 2
+		}
+	}
+	if chk := (11 - sum%11) % 11; chk != 10 {
+		return string(rune('0' + chk))
+	}
+	return "X"
+}
+
+// eanCheck returns the EAN-13 / UPC-A / ISBN-13 / GTIN check digit over the
+// digits of s: weights 3 and 1 alternating from the rightmost digit, mod 10.
+func eanCheck(s string) string {
+	sum, w := 0, 3
+	for i := len(s) - 1; i >= 0; i-- {
+		c := s[i]
+		if c < '0' || c > '9' {
+			continue
+		}
+		sum += int(c-'0') * w
+		w = 4 - w // 3 <-> 1
+	}
+	return string(rune('0' + (10-sum%10)%10))
+}
+
+// ibanLen maps a supported country code to the full IBAN length. The check digits
+// sit between the country code and the BBAN, so — unlike luhn/ean — iban can't be
+// a left-to-right derivation; it generates the whole value instead.
+var ibanLen = map[string]int{"BE": 16, "DE": 22, "DK": 18, "ES": 24, "FI": 18, "NO": 15, "SE": 24}
+
+func ibanArg(_ map[string]node, a []string) error {
+	if _, ok := ibanLen[a[0]]; !ok {
+		return fmt.Errorf("iban(%q): unsupported country code", a[0])
+	}
+	return nil
+}
+
+// iban generates a structurally valid IBAN for cc: a numeric BBAN of the right
+// length, then mod-97 check digits. Real bank/branch structure isn't modelled —
+// the result passes length and checksum validation, which is what fake data needs.
+func iban(r rng, cc string) string {
+	bban := make([]byte, ibanLen[cc]-4)
+	for i := range bban {
+		bban[i] = byte('0' + r.IntN(10))
+	}
+	rem := 0
+	feed := func(d int) { rem = (rem*10 + d) % 97 }
+	for _, c := range bban {
+		feed(int(c - '0'))
+	}
+	for i := 0; i < len(cc); i++ { // letters A-Z -> 10..35, fed as two digits
+		v := int(cc[i]-'A') + 10
+		feed(v / 10)
+		feed(v % 10)
+	}
+	feed(0)
+	feed(0)
+	return fmt.Sprintf("%s%02d%s", cc, 98-rem, bban)
+}

cmd/fejkdata/main.go

@@ -24,10 +24,13 @@ import (
 const usage = `Usage: fejkdata [flags] <path|template>
 
   <path>                 a category, or a dotted path into one (person, person.last);
-                         a table's row by key or name: 'misc.country[SE]', 'misc.country[Sweden].tld'
+                         a table's row by key or name: 'misc.territory[SE]', 'misc.territory[Sweden].tld'
   <template>             a format string or JSON value to render inline, e.g.
                          'name: {/sv_SE.person.last}' or '{"format":"{x}","x":["bosse","lina"]}'
 
+A layout inside a template is single-quoted, so quote the whole argument with " to
+keep it: "{date(1990-01-01,2010-12-31,'2006-01-02')}".
+
 An argument containing a { token, or a JSON object, array or string, is a
 template; any other argument is a path (a path never contains a brace or a quote,
 and a bracket only as a [selector] after a table's name). Templates reach the
@@ -279,7 +282,7 @@ func (in invocation) check() (argKind, error) {
 		return argPath, nil
 	}
 	if len(in.paths) != 1 {
-		return argPath, fmt.Errorf("expected one path or template, got %d", len(in.paths))
+		return argPath, fmt.Errorf("expected one path or template, got %d%s", len(in.paths), templateSplit(in.paths))
 	}
 	return classify(in.paths[0])
 }
@@ -475,11 +478,25 @@ func run(args []string, stdout, stderr io.Writer) int {
 			return misuse(stderr, te.error)
 		}
 		fmt.Fprintln(stderr, err)
+		if errors.Is(err, fejkdata.ErrNoColumns) {
+			fmt.Fprintln(stderr, "wrap that JSON in single quotes, which keep a shell from expanding its braces")
+		}
 		return 1
 	}
 	return 0
 }
 
+// templateSplit names the cure for a template a shell split on its spaces, which is
+// what several arguments carrying a token mean.
+func templateSplit(paths []string) string {
+	for _, p := range paths {
+		if strings.Contains(p, "{") {
+			return `; a template's spaces split the argument, so wrap the whole argument in "…", or in '…' where it carries double quotes`
+		}
+	}
+	return ""
+}
+
 func misuse(stderr io.Writer, err error) int {
 	// A library error already names the program, so the prefix is not doubled.
 	fmt.Fprintf(stderr, "fejkdata: %s\ntry 'fejkdata --help'\n", strings.TrimPrefix(err.Error(), "fejkdata: "))

cmd/fejkdata/main_test.go

@@ -223,6 +223,14 @@ func TestRunMisuse(t *testing.T) {
 			t.Errorf("run(%v) stdout = %q, want nothing", args, out)
 		}
 	}
+	// A template a shell split names both quotings, since which one cures it turns
+	// on whether the argument carries a double-quoted JSON.
+	_, _, errb := runOut("{date(1990-01-01,2010-12-31,", "January", "2,", "2006)}")
+	for _, want := range []string{`wrap the whole argument in "…"`, `or in '…'`} {
+		if !strings.Contains(errb, want) {
+			t.Errorf("a split template stderr = %q, want it to name %s", errb, want)
+		}
+	}
 }
 
 func TestRunMultipleDataPaths(t *testing.T) {
@@ -344,8 +352,8 @@ func TestRunSelectsATableRow(t *testing.T) {
 	if code, _, errb := runOut("--no-shipped-data", "-d", dir, "[Skåne län]"); code != 2 || !strings.Contains(errb, "JSON") {
 		t.Fatalf("[Skåne län] = %d, stderr=%q, want misuse: a leading bracket that is no JSON names nothing", code, errb)
 	}
-	if code, out, _ := runOut("--seed", "1", "--format", "csv", "misc.country[SE]"); code != 0 || !strings.HasPrefix(out, "alpha2,") || !strings.Contains(out, "\nSE,SWE,") {
+	if code, out, _ := runOut("--seed", "1", "--format", "csv", "misc.territory[SE]"); code != 0 || !strings.HasPrefix(out, "alpha2,") || !strings.Contains(out, "\nSE,SWE,") {
-		t.Fatalf("--format csv misc.country[SE] = %d, %q", code, out)
+		t.Fatalf("--format csv misc.territory[SE] = %d, %q", code, out)
 	}
 }
 
@@ -613,3 +621,21 @@ func TestRunNumbersFollowTheShell(t *testing.T) {
 		t.Errorf("run(--seed 007 --repeat +3) = %d, %q, stderr %q; want the same as --seed 7 --repeat 3 %q", code, got, errb, want)
 	}
 }
+
+// TestRunRecordOfAFieldlessPath pins the way out of asking a category of one value
+// for columns: the record to write, and the quoting a shell needs to take it. The
+// argument is well-formed and only the data cannot serve it, so the code is 1.
+func TestRunRecordOfAFieldlessPath(t *testing.T) {
+	code, out, errb := runOut("--format", "csv", "sv_SE.personnummer")
+	if code != 1 {
+		t.Fatalf("run(--format csv sv_SE.personnummer) = %d, want 1", code)
+	}
+	for _, want := range []string{`{"format":"","personnummer":"{/sv_SE.personnummer}"}`, "single quotes"} {
+		if !strings.Contains(errb, want) {
+			t.Errorf("stderr = %q, want it to name %s", errb, want)
+		}
+	}
+	if out != "" {
+		t.Errorf("stdout = %q, want nothing", out)
+	}
+}

data-import/currency.py

@@ -11,6 +11,7 @@ import io
 import xml.etree.ElementTree as ET
 from pathlib import Path
 
+import source
 import tsv
 
 SOURCE = "https://raw.githubusercontent.com/datasets/currency-codes/main/data/codes-all.csv"
@@ -58,8 +59,8 @@ def main():
     p.add_argument("--symbols", nargs="+", default=SYMBOLS)
     p.add_argument("--out", default=str(OUT))
     a = p.parse_args()
-    symbol = symbols(tsv.fetch(s, a.cache, Path(s).name).decode("utf-8") for s in a.symbols)
+    symbol = symbols(source.fetch(s, a.cache, Path(s).name).decode("utf-8") for s in a.symbols)
-    table = rows(tsv.fetch(a.source, a.cache, "codes-all.csv").decode("utf-8"), symbol)
+    table = rows(source.fetch(a.source, a.cache, "codes-all.csv").decode("utf-8"), symbol)
     tsv.write(a.out, COLUMNS, sorted(table, key=lambda r: r["code"]))
 
 

data-import/geo-se.py

@@ -13,11 +13,12 @@ import os
 import re
 import sys
 import urllib.request
-import xml.etree.ElementTree as ET
 import zipfile
 from pathlib import Path
 
+import source
 import tsv
+import xlsx
 
 CODES = "https://www.scb.se/contentassets/7a89e48960f741e08918e489ea36354a/kommunlankod-2026.xlsx"
 POPULATION = "https://api.scb.se/OV0104/v1/doris/sv/ssd/START/BE/BE0101/BE0101A/BefolkningNy"
@@ -38,24 +39,11 @@ CACHE = Path(__file__).resolve().parent / "cache"
 TIMEZONE = "Europe/Stockholm"
 ONE_POSITION = {"Stockholm", "Göteborg", "Malmö"}
 UNMATCHED_POPULATION = 200
-XLSX_NS = {"m": "http://schemas.openxmlformats.org/spreadsheetml/2006/main"}
-
-
-def xlsx_rows(data):
-    z = zipfile.ZipFile(io.BytesIO(data))
-    strings = ["".join(t.text or "" for t in si.iter("{%s}t" % XLSX_NS["m"])) for si in ET.fromstring(z.read("xl/sharedStrings.xml")).findall("m:si", XLSX_NS)]
-    sheet = ET.fromstring(z.read("xl/worksheets/sheet1.xml"))
-    for row in sheet.findall(".//m:row", XLSX_NS):
-        cells = []
-        for c in row.findall("m:c", XLSX_NS):
-            v = c.find("m:v", XLSX_NS)
-            cells.append("" if v is None else strings[int(v.text)] if c.get("t") == "s" else v.text)
-        yield cells
 
 
 def scb_codes(cache):
     regions, municipalities = {}, {}
-    for cells in xlsx_rows(tsv.fetch(CODES, cache, "kommunlankod.xlsx", magic=b"PK")):
+    for cells in xlsx.rows(source.fetch(CODES, cache, "kommunlankod.xlsx", magic=b"PK")):
         if len(cells) < 2 or not re.fullmatch(r"\d{2}|\d{4}", cells[0]):
             continue
         (regions if len(cells[0]) == 2 else municipalities)[cells[0]] = cells[1].strip()
@@ -64,12 +52,12 @@ def scb_codes(cache):
 
 def scb_population(cache):
     body = json.dumps(POPULATION_QUERY).encode()
-    data = tsv.fetch(POPULATION, cache, "befolkning.json", data=body, headers={"Content-Type": "application/json"})
+    data = source.fetch(POPULATION, cache, "befolkning.json", data=body, headers={"Content-Type": "application/json"})
     return {row["key"][0]: row["values"][0] for row in json.loads(data.decode("utf-8-sig"))["data"]}
 
 
 def scb_tatorter(cache):
-    text = tsv.fetch(TATORTER, cache, "tatorter.csv").decode("utf-8")
+    text = source.fetch(TATORTER, cache, "tatorter.csv").decode("utf-8")
     by_name = collections.defaultdict(list)
     for r in csv.DictReader(io.StringIO(text)):
         by_name[r["tatort"]].append((r["kommun"], int(r["bef"])))
@@ -77,7 +65,7 @@ def scb_tatorter(cache):
 
 
 def geonames(cache):
-    z = zipfile.ZipFile(io.BytesIO(tsv.fetch(POSTAL_CODES, cache, "SE.zip", magic=b"PK")))
+    z = zipfile.ZipFile(io.BytesIO(source.fetch(POSTAL_CODES, cache, "SE.zip", magic=b"PK")))
     rows = []
     for line in z.read("SE.txt").decode("utf-8").splitlines():
         f = line.split("\t")

data-import/geo-us.py

@@ -14,6 +14,7 @@ import sys
 import zipfile
 from pathlib import Path
 
+import source
 import tsv
 
 GAZETTEER = "https://www2.census.gov/geo/docs/maps-data/data/gazetteer/2026_Gazetteer/2026_Gaz_{}_national.zip"
@@ -27,7 +28,7 @@ OUT = Path(__file__).resolve().parent.parent / "data" / "geo" / "US"
 CACHE = Path(__file__).resolve().parent / "cache"
 ESTIMATE = "POPESTIMATE2025"
 CDP = "57"
-HIGHWAY = re.compile(r"\b(I- |Hwy |Rte |Route |Rd )\d")
+HIGHWAY = re.compile(r"\b(I- |Hwy |Highway |Loop |Rte |Route |Rd )\d")
 SUFFIX = re.compile(r" (city and borough|city|town|village|borough|municipality|comunidad|zona urbana|metropolitan government|metro government|consolidated government|unified government|urban county|corporation|plantation)( \(balance\))?$")
 # Places whose Census name is a merged government's; the postal city is what an address carries.
 NAMES = {"1303440": "Athens", "1304204": "Augusta", "1349008": "Macon", "2148006": "Louisville", "3011397": "Butte", "4732742": "Hartsville", "4752006": "Nashville"}
@@ -56,14 +57,14 @@ def text(data):
 
 
 def gazetteer(cache, kind):
-    z = zipfile.ZipFile(io.BytesIO(tsv.fetch(GAZETTEER.format(kind), cache, f"gaz_{kind}.zip", magic=b"PK")))
+    z = zipfile.ZipFile(io.BytesIO(source.fetch(GAZETTEER.format(kind), cache, f"gaz_{kind}.zip", magic=b"PK")))
     rows = text(z.read(z.namelist()[0])).splitlines()
     header = [h.strip() for h in rows[0].split("|")]
     return [dict(zip(header, (c.strip() for c in row.split("|")))) for row in rows[1:]]
 
 
 def csv_rows(cache, url, name):
-    return list(csv.DictReader(io.StringIO(text(tsv.fetch(url, cache, name)))))
+    return list(csv.DictReader(io.StringIO(text(source.fetch(url, cache, name)))))
 
 
 def dbf_rows(data, wanted):
@@ -89,7 +90,7 @@ def dbf_rows(data, wanted):
 
 
 def tiger_zip(cache, kind, county):
-    return tsv.fetch(TIGER.format(kind.upper(), county, kind), cache, f"tl_{county}_{kind}.zip", magic=b"PK")
+    return source.fetch(TIGER.format(kind.upper(), county, kind), cache, f"tl_{county}_{kind}.zip", magic=b"PK")
 
 
 def tiger(cache, kind, county, wanted):
@@ -128,10 +129,10 @@ def localities(cache, min_population, counties):
 
 
 def postal_codes(cache, localities):
-    """Each ZCTA whose largest part inside any place lies in a shipped place."""
+    """Each ZCTA whose largest part inside an incorporated place lies in a shipped place."""
     parts = {}
-    for r in csv.DictReader(io.StringIO(text(tsv.fetch(ZCTA_PLACE, cache, "zcta-place.txt"))), delimiter="|"):
+    for r in csv.DictReader(io.StringIO(text(source.fetch(ZCTA_PLACE, cache, "zcta-place.txt"))), delimiter="|"):
-        if r["GEOID_ZCTA5_20"] and r["GEOID_PLACE_20"]:
+        if r["GEOID_ZCTA5_20"] and r["GEOID_PLACE_20"] and not r["NAMELSAD_PLACE_20"].endswith(" CDP"):
             parts.setdefault(r["GEOID_ZCTA5_20"], []).append((int(r["AREALAND_PART"]), r["GEOID_PLACE_20"]))
     largest = {zcta: max(p)[1] for zcta, p in parts.items()}
     return {zcta: place for zcta, place in largest.items() if place in localities}

data-import/httpmethod.py

@@ -0,0 +1,59 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/httpmethod.tsv from the IANA HTTP Method Registry.
+
+    data-import/httpmethod.py [--source URL_OR_FILE] [--cache DIR] [--out FILE]
+
+A method ships when the register cites RFC 9110 section 9.3 or RFC 5789 for it, which
+is the eight HTTP defines and PATCH; the register's other entries do not. The
+register's `yes` and `no` ship as `true` and `false`, which a consumer's boolean reads.
+"""
+import argparse
+import csv
+import io
+import re
+import sys
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://www.iana.org/assignments/http-methods/methods.csv"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "httpmethod.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["idempotent", "method", "safe"]
+CORE = re.compile(r"\[RFC9110, Section 9\.3\.|\[RFC5789, Section 2\]")
+BOOLEAN = {"yes": "true", "no": "false"}
+SELECTOR = '[]{}"|'  # mirrors inSelector in table.go
+EXPECTED = 9
+
+
+def rows(text):
+    for r in csv.DictReader(io.StringIO(text)):
+        method = (r["Method Name"] or "").strip()
+        if not CORE.search(r["Reference"] or ""):
+            continue
+        if set(method) & set(SELECTOR):
+            sys.exit(f"{method}: a method holding one of {SELECTOR} is one a selector cannot spell")
+        answers = {}
+        for column in ("Safe", "Idempotent"):
+            answer = (r[column] or "").strip()
+            if answer not in BOOLEAN:
+                sys.exit(f"{method}: {column} {answer!r} is neither yes nor no")
+            answers[column] = BOOLEAN[answer]
+        yield {"idempotent": answers["Idempotent"], "method": method, "safe": answers["Safe"]}
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--out", default=str(OUT))
+    a = p.parse_args()
+    table = sorted(rows(source.fetch(a.source, a.cache, "http-methods.csv").decode("utf-8")), key=lambda r: r["method"])
+    if len(table) != EXPECTED:
+        sys.exit(f"{len(table)} methods cite RFC 9110 section 9.3 or RFC 5789, not {EXPECTED}; the register's Reference column has moved")
+    tsv.write(a.out, COLUMNS, table)
+
+
+if __name__ == "__main__":
+    main()

data-import/httpstatus.py

@@ -0,0 +1,40 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/httpstatus.tsv from the IANA HTTP Status Code Registry.
+
+    data-import/httpstatus.py [--source URL_OR_FILE] [--cache DIR] [--out FILE]
+
+What ships is the codes in use: a range, an unassigned code and a qualified reason are all skipped.
+"""
+import argparse
+import csv
+import io
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://www.iana.org/assignments/http-status-codes/http-status-codes-1.csv"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "httpstatus.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["code", "reason"]
+
+
+def rows(text):
+    for r in csv.DictReader(io.StringIO(text)):
+        code, reason = r["Value"].strip(), r["Description"].strip()
+        if code.isdigit() and reason and "(" not in reason and reason != "Unassigned":
+            yield {"code": code, "reason": reason}
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--out", default=str(OUT))
+    a = p.parse_args()
+    table = rows(source.fetch(a.source, a.cache, "http-status-codes-1.csv").decode("utf-8"))
+    tsv.write(a.out, COLUMNS, sorted(table, key=lambda r: int(r["code"])))
+
+
+if __name__ == "__main__":
+    main()

data-import/language.py

@@ -0,0 +1,45 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/language.tsv from datasets/language-codes (PDDL), the LoC ISO 639-2 register.
+
+    data-import/language.py [--source URL_OR_FILE] [--cache DIR] [--out FILE]
+
+Only the 639-2 entries carrying a 639-1 code ship, since the alpha-2 code is the key.
+"""
+import argparse
+import csv
+import io
+import re
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://raw.githubusercontent.com/datasets/language-codes/main/data/language-codes-full.csv"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "language.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["code", "code3", "name"]
+
+
+def name(english):
+    """The register lists every synonym; the first, without its qualifier, is the one a selector spells."""
+    return re.sub(r"\s*\([^)]*\)", "", english.split(";")[0]).strip()
+
+
+def rows(text):
+    for r in csv.DictReader(io.StringIO(text)):
+        if r["alpha2"]:
+            yield {"code": r["alpha2"], "code3": r["alpha3-t"] or r["alpha3-b"], "name": name(r["English"])}
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--out", default=str(OUT))
+    a = p.parse_args()
+    table = rows(source.fetch(a.source, a.cache, "language-codes-full.csv").decode("utf-8"))
+    tsv.write(a.out, COLUMNS, sorted(table, key=lambda r: r["code"]))
+
+
+if __name__ == "__main__":
+    main()

data-import/loglevel.py

@@ -0,0 +1,68 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/loglevel.tsv from RFC 5424 and POSIX <syslog.h>.
+
+    data-import/loglevel.py [--source URL_OR_FILE] [--posix URL_OR_FILE] [--cache DIR] [--out FILE]
+
+Every severity in Table 2 ships, paired to POSIX's keywords by the order each source
+lists them, which holds only because a keyword prefixes the severity RFC 5424 numbers
+alike.
+"""
+import argparse
+import html
+import re
+import sys
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://www.rfc-editor.org/rfc/rfc5424.txt"
+POSIX = "https://pubs.opengroup.org/onlinepubs/9799919799/basedefs/syslog.h.html"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "loglevel.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["code", "keyword", "severity"]
+TABLE = re.compile(r"Table 1\.\s+Syslog Message Facilities(.*?)Table 2\.\s+Syslog Message Severities", re.S)
+ROW = re.compile(r"^ +(\d) +([A-Z][a-z]+): ", re.M)
+SECTION = re.compile(r"severity level portion of the.*?macros:(.*?)The following shall be declared as functions", re.S)
+MACRO = re.compile(r"\bLOG_([A-Z]+)\b")
+EXPECTED = 8
+
+
+def severities(text):
+    table = TABLE.search(text)
+    if not table:
+        sys.exit("the RFC holds no text between the facility and the severity caption; its tables have moved")
+    return ROW.findall(table.group(1))
+
+
+def keywords(page):
+    section = SECTION.search(html.unescape(re.sub(r"<[^>]+>", "", page)))
+    if not section:
+        sys.exit("POSIX spells no severity list between its own caption and the function declarations; the page has moved")
+    return [m.lower() for m in MACRO.findall(section.group(1))]
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--posix", default=POSIX)
+    p.add_argument("--out", default=str(OUT))
+    a = p.parse_args()
+    table = sorted(severities(source.fetch(a.source, a.cache, "rfc5424.txt").decode("utf-8")))
+    codes = [code for code, _ in table]
+    if codes != [str(i) for i in range(EXPECTED)]:
+        sys.exit(f"the severity table numbers {codes}, not 0 through {EXPECTED - 1}")
+    words = keywords(source.fetch(a.posix, a.cache, "posix-syslog.h.html").decode("utf-8"))
+    if len(words) != EXPECTED:
+        sys.exit(f"POSIX lists {len(words)} severity macros, not {EXPECTED}: {words}")
+    rows = []
+    for (code, severity), keyword in zip(table, words):
+        if not severity.lower().startswith(keyword):
+            sys.exit(f"severity {code} is {severity} in the RFC and {keyword} in POSIX, which is no prefix of it")
+        rows.append({"code": code, "keyword": keyword, "severity": severity})
+    tsv.write(a.out, COLUMNS, rows)
+
+
+if __name__ == "__main__":
+    main()

data-import/mimetype.py

@@ -0,0 +1,39 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/mimetype.tsv from mime-db (MIT), the IANA media type registry with extensions.
+
+    data-import/mimetype.py [--source URL_OR_FILE] [--cache DIR] [--out FILE]
+
+Only types mime-db records as IANA-registered, with a filename extension, ship.
+"""
+import argparse
+import json
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://raw.githubusercontent.com/jshttp/mime-db/master/db.json"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "mimetype.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["ext", "type"]
+FIXUPS = {"application/mp4": "mp4s", "audio/mpeg": "mp3", "audio/ogg": "ogg", "video/quicktime": "mov"}
+
+
+def rows(db):
+    for mimetype, entry in db.items():
+        if entry.get("source") == "iana" and entry.get("extensions"):
+            yield {"ext": "." + FIXUPS.get(mimetype, entry["extensions"][0]), "type": mimetype}
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--out", default=str(OUT))
+    a = p.parse_args()
+    table = rows(json.loads(source.fetch(a.source, a.cache, "mime-db.json").decode("utf-8")))
+    tsv.write(a.out, COLUMNS, sorted(table, key=lambda r: r["type"]))
+
+
+if __name__ == "__main__":
+    main()

data-import/names-se.py

@@ -0,0 +1,54 @@
+#!/usr/bin/env python3
+"""Rebuild data/sv_SE/first-name.tsv and last-name.tsv from SCB's whole-population name counts (CC0, "Källa: SCB").
+
+    data-import/names-se.py [--source URL_OR_FILE] [--first N] [--last N] [--cache DIR] [--out DIR]
+
+A tilltalsnamn goes under each sex that carries it, so a name both sexes carry has two rows.
+"""
+import argparse
+import re
+from pathlib import Path
+
+import source
+import tsv
+import xlsx
+
+SOURCE = "https://www.scb.se/contentassets/9fe7dbb460994c72b835163dbc491ef9/namn-med-minst-tva-barare-31-december-2022.xlsx"
+OUT = Path(__file__).resolve().parent.parent / "data" / "sv_SE"
+CACHE = Path(__file__).resolve().parent / "cache"
+SHEETS = {"f": "Tilltalsnamn kvinnor", "m": "Tilltalsnamn män"}
+SURNAMES = "Efternamn"
+NAME = re.compile(r"^[^\W\d_]{2,}([ '-][^\W\d_]{2,})*$")
+PARTICLES = {"af", "av", "da", "de", "del", "den", "der", "di", "dos", "du", "la", "le", "van", "von"}
+
+
+def cased(name):
+    """SCB's uppercase name as it is written: each part capitalised, a particle before another part lowercased."""
+    parts = name.lower().split(" ")
+    return " ".join(w if w in PARTICLES and len(parts) > 1 else w.title() for w in parts)
+
+
+def counted(rows, top):
+    """The top names of a sheet by bearers, initials and single letters dropped."""
+    names = [(cells[0], int(cells[1])) for cells in rows if len(cells) >= 2 and cells[1].isdigit() and NAME.match(cells[0])]
+    return sorted(names, key=lambda n: (-n[1], n[0]))[:top]
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--first", type=int, default=2000, help="names per sex")
+    p.add_argument("--last", type=int, default=5000)
+    p.add_argument("--out", default=str(OUT))
+    p.add_argument("--source", default=SOURCE)
+    a = p.parse_args()
+    data = source.fetch(a.source, a.cache, "scb-namn-2022.xlsx", magic=b"PK")
+    first = [{"name": cased(n), "sex": sex, "count": c} for sex, sheet in SHEETS.items() for n, c in counted(xlsx.rows(data, sheet), a.first)]
+    last = [{"name": cased(n), "count": c} for n, c in counted(xlsx.rows(data, SURNAMES), a.last)]
+    out = Path(a.out)
+    tsv.write(out / "first-name.tsv", ["name", "sex", "count"], sorted(first, key=lambda r: (r["name"], r["sex"])))
+    tsv.write(out / "last-name.tsv", ["name", "count"], sorted(last, key=lambda r: r["name"]))
+
+
+if __name__ == "__main__":
+    main()

data-import/names-us.py

@@ -0,0 +1,79 @@
+#!/usr/bin/env python3
+"""Rebuild data/en_US/first-name.tsv and last-name.tsv from SSA baby names and the Census 2010 surnames (public domain).
+
+    data-import/names-us.py [--names URL_OR_FILE] [--surnames URL_OR_FILE] [--from-year YEAR] [--first N] [--last N] [--cache DIR] [--out DIR]
+
+A given name is counted over the births of --from-year and later, under each sex it was given to, so a
+name both sexes carry has two rows. --names takes SSA's names.zip or a merged name,sex,count,year file;
+--surnames takes the Census names.zip or its Names_2010Census.csv.
+"""
+import argparse
+import collections
+import csv
+import io
+import re
+import zipfile
+from pathlib import Path
+
+import source
+import tsv
+
+NAMES = "https://raw.githubusercontent.com/hackerb9/ssa-baby-names/master/alldata.txt"
+SURNAMES = "https://www2.census.gov/topics/genealogy/2010surnames/names.zip"
+OUT = Path(__file__).resolve().parent.parent / "data" / "en_US"
+CACHE = Path(__file__).resolve().parent / "cache"
+MC = re.compile(r"^Mc([a-z])")
+PLACEHOLDERS = {"Baby", "Female", "Infant", "Male", "Notnamed", "Unknown", "Unnamed"}
+
+
+def csv_or_zip_rows(data, member):
+    """The rows of a CSV, or of every member of a zip named like member, name,sex,count[,year]."""
+    if data.startswith(b"PK"):
+        z = zipfile.ZipFile(io.BytesIO(data))
+        for n in sorted(z.namelist()):
+            if re.fullmatch(member, n):
+                year = re.sub(r"\D", "", n)
+                for r in csv.reader(io.StringIO(z.read(n).decode("utf-8-sig"))):
+                    yield r + [year] if year else r
+        return
+    yield from csv.reader(io.StringIO(data.decode("utf-8-sig")))
+
+
+def given(data, from_year):
+    counts = collections.Counter()
+    for r in csv_or_zip_rows(data, r"yob\d{4}\.txt"):
+        if len(r) >= 4 and r[3].isdigit() and int(r[3]) >= from_year and r[0] not in PLACEHOLDERS:
+            counts[(r[0], r[1].lower())] += int(r[2])
+    return counts
+
+
+def surname(name):
+    """A Census uppercase surname as it is written: capitalised, and Mc before a capital."""
+    return MC.sub(lambda m: "Mc" + m.group(1).upper(), name.title())
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--first", type=int, default=2000, help="names per sex")
+    p.add_argument("--from-year", type=int, default=1930)
+    p.add_argument("--last", type=int, default=5000)
+    p.add_argument("--names", default=NAMES)
+    p.add_argument("--out", default=str(OUT))
+    p.add_argument("--surnames", default=SURNAMES)
+    a = p.parse_args()
+    counts = given(source.fetch(a.names, a.cache, "ssa-names.txt"), a.from_year)
+    first = []
+    for sex in ("f", "m"):
+        top = sorted(((n, c) for (n, s), c in counts.items() if s == sex), key=lambda n: (-n[1], n[0]))[:a.first]
+        first += [{"name": n, "sex": sex, "count": c} for n, c in top]
+    rows = csv_or_zip_rows(source.fetch(a.surnames, a.cache, "census-surnames-2010.zip"), r"Names_2010Census\.csv")
+    last = [{"name": surname(r[0]), "count": int(r[2])} for r in rows if len(r) >= 3 and r[2].isdigit() and r[0].isalpha()]
+    last = sorted(last, key=lambda r: (-r["count"], r["name"]))[:a.last]
+    out = Path(a.out)
+    tsv.write(out / "first-name.tsv", ["name", "sex", "count"], sorted(first, key=lambda r: (r["name"], r["sex"])))
+    tsv.write(out / "last-name.tsv", ["name", "count"], sorted(last, key=lambda r: r["name"]))
+
+
+if __name__ == "__main__":
+    main()

data-import/port.py

@@ -0,0 +1,55 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/port.tsv from the IANA Service Name and Transport Protocol Port Number Registry.
+
+    data-import/port.py [--source URL_OR_FILE] [--cache DIR] [--out FILE]
+
+A row needs a service name and a numeric TCP port. A port the registry lists more than
+once keeps the first service it describes, or the first of them where it describes none,
+so a number selects one row.
+"""
+import argparse
+import csv
+import io
+import sys
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://www.iana.org/assignments/service-names-port-numbers/service-names-port-numbers.csv"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "port.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["number", "service"]
+FLOOR = 4000
+
+
+def rows(text):
+    best, described = {}, set()
+    for r in csv.DictReader(io.StringIO(text)):
+        service, number = (r["Service Name"] or "").strip(), (r["Port Number"] or "").strip()
+        transport, description = (r["Transport Protocol"] or "").strip(), (r["Description"] or "").strip()
+        if transport != "tcp" or not service or not number.isdigit():
+            continue
+        if not 1 <= int(number) <= 65535:
+            sys.exit(f"{service}: port {number} is outside 1-65535")
+        if number not in best or (description and number not in described):
+            best[number] = {"number": number, "service": service}
+        if description:
+            described.add(number)
+    return sorted(best.values(), key=lambda r: int(r["number"]))
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--out", default=str(OUT))
+    a = p.parse_args()
+    table = rows(source.fetch(a.source, a.cache, "service-names-port-numbers.csv").decode("utf-8"))
+    if len(table) < FLOOR:
+        sys.exit(f"only {len(table)} ports named a TCP service; the registry's columns have moved")
+    tsv.write(a.out, COLUMNS, table)
+
+
+if __name__ == "__main__":
+    main()

data-import/protocol.py

@@ -0,0 +1,67 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/protocol.tsv from the IANA Protocol Numbers registry.
+
+    data-import/protocol.py [--source URL_OR_FILE] [--cache DIR] [--out FILE]
+
+A row needs a number and a keyword, which drops the seven numbers assigned to a class of
+protocols rather than to one; a keyword the registry reserves or marks deprecated goes
+too. Where the registry names no protocol beside the keyword, the keyword is the name.
+"""
+import argparse
+import csv
+import io
+import sys
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://www.iana.org/assignments/protocol-numbers/protocol-numbers-1.csv"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "protocol.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["keyword", "name", "number"]
+SKIP = ("Reserved", "deprecated")
+SELECTOR = '[]{}"|'  # mirrors inSelector in table.go
+
+
+def rows(text):
+    for r in csv.DictReader(io.StringIO(text)):
+        number, keyword = (r["Decimal"] or "").strip(), (r["Keyword"] or "").strip()
+        if not number.isdigit() or not keyword or any(s in keyword for s in SKIP):
+            continue
+        if not 0 <= int(number) <= 255:
+            sys.exit(f"{keyword}: protocol number {number} is outside 0-255")
+        yield {"keyword": keyword, "name": " ".join((r["Protocol"] or "").split()) or keyword, "number": number}
+
+
+def refuse_an_unselectable_row(table):
+    """What the loader refuses at New, and a name that names two rows rather than one."""
+    keywords, names = {}, {}
+    for r in table:
+        for column in ("keyword", "name"):
+            if set(r[column]) & set(SELECTOR):
+                sys.exit(f"{r['keyword']}: {column} {r[column]!r} holds one of {SELECTOR}, which a selector cannot spell")
+        if r["keyword"] in keywords:
+            sys.exit(f"{r['keyword']}: repeats the keyword of protocol {keywords[r['keyword']]}")
+        keywords[r["keyword"]] = r["number"]
+    for r in table:
+        if r["name"] != r["keyword"] and r["name"] in keywords:
+            sys.exit(f"{r['keyword']}: name {r['name']!r} is another protocol's keyword")
+        if r["name"] in names:
+            sys.exit(f"{r['keyword']}: name {r['name']!r} repeats {names[r['name']]}'s, so it selects neither")
+        names[r["name"]] = r["keyword"]
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--out", default=str(OUT))
+    a = p.parse_args()
+    table = sorted(rows(source.fetch(a.source, a.cache, "protocol-numbers-1.csv").decode("utf-8")), key=lambda r: int(r["number"]))
+    refuse_an_unselectable_row(table)
+    tsv.write(a.out, COLUMNS, table)
+
+
+if __name__ == "__main__":
+    main()

data-import/source.py

@@ -0,0 +1,29 @@
+"""A source fetched once into the cache."""
+import re
+import sys
+import time
+import urllib.request
+from pathlib import Path
+
+
+def fetch(source, cache, name, magic=b"", data=None, headers=None):
+    """The bytes of a URL, downloaded into cache/name once, or of a local file."""
+    if not re.match(r"^https?://", source):
+        return Path(source).read_bytes()
+    path = Path(cache) / name
+    if path.exists():
+        return path.read_bytes()
+    for attempt in range(1, 6):
+        req = urllib.request.Request(source, data=data, headers={"User-Agent": "fejkdata data-import", **(headers or {})})
+        try:
+            with urllib.request.urlopen(req, timeout=600) as r:
+                body = r.read()
+        except OSError:
+            body = b""
+        if body and body.startswith(magic) and b"Request Rejected" not in body[:512]:
+            path.parent.mkdir(parents=True, exist_ok=True)
+            path.write_bytes(body)
+            return body
+        if attempt < 5:
+            time.sleep(10 * attempt)
+    sys.exit(f"{source}: no valid download in 5 attempts")

data-import/territory.py

@@ -1,20 +1,24 @@
 #!/usr/bin/env python3
-"""Rebuild data/misc/country.tsv from datasets/country-codes (PDDL).
+"""Rebuild data/misc/territory.tsv from datasets/country-codes (PDDL).
 
-    data-import/country.py [--source URL_OR_FILE] [--cache DIR] [--out FILE]
+    data-import/territory.py [--source URL_OR_FILE] [--cache DIR] [--out FILE]
 """
 import argparse
 import csv
 import io
 import re
+import sys
 from pathlib import Path
 
+import source
 import tsv
 
 SOURCE = "https://raw.githubusercontent.com/datasets/country-codes/main/data/country-codes.csv"
-OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "country.tsv"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "territory.tsv"
 CACHE = Path(__file__).resolve().parent / "cache"
-COLUMNS = ["alpha2", "alpha3", "calling-code", "capital", "currency", "flag", "languages", "name", "numeric", "tld"]
+COLUMNS = ["alpha2", "alpha3", "calling-code", "capital", "country", "currency", "flag", "languages", "name", "numeric", "tld"]
+SOVEREIGN = re.compile(r"^(?:Part of|Territor(?:y|ies) of|Crown dependency of|Commonwealth of|Associated with) ([A-Z]{2})$")
+STANDALONE = {"In contention", "International"}
 # Gaps in the source, keyed by alpha2.
 FIXUPS = {"TR": {"currency": "TRY"}}
 
@@ -32,6 +36,16 @@ def languages(field):
     return ",".join(p for p in field.split(",") if p)
 
 
+def country(alpha2, independent):
+    """The sovereign state the register records; a territory it records none for stands alone."""
+    m = SOVEREIGN.match(independent)
+    if m:
+        return m.group(1)
+    if independent == "Yes" or independent in STANDALONE:
+        return alpha2
+    sys.exit(f"{alpha2}: is_independent {independent!r} names no sovereign this script can read")
+
+
 def rows(text):
     for r in csv.DictReader(io.StringIO(text)):
         alpha2 = r["ISO3166-1-Alpha-2"]
@@ -40,6 +54,7 @@ def rows(text):
             "alpha3": r["ISO3166-1-Alpha-3"],
             "calling-code": calling_code(r["Dial"]),
             "capital": r["Capital"],
+            "country": country(alpha2, r["is_independent"]),
             "currency": r["ISO4217-currency_alphabetic_code"].split(",")[0],
             "flag": flag(alpha2),
             "languages": languages(r["Languages"]),
@@ -58,8 +73,12 @@ def main():
     p.add_argument("--source", default=SOURCE)
     p.add_argument("--out", default=str(OUT))
     a = p.parse_args()
-    table = rows(tsv.fetch(a.source, a.cache, "country-codes.csv").decode("utf-8"))
+    table = sorted(rows(source.fetch(a.source, a.cache, "country-codes.csv").decode("utf-8")), key=lambda r: r["alpha2"])
-    tsv.write(a.out, COLUMNS, sorted(table, key=lambda r: r["alpha2"]))
+    shipped = {r["alpha2"] for r in table}
+    for r in table:
+        if r["country"] not in shipped:
+            sys.exit(f"{r['alpha2']}: country {r['country']} is no shipped territory; a sovereign the set drops leaves its territories unreadable")
+    tsv.write(a.out, COLUMNS, table)
 
 
 if __name__ == "__main__":

data-import/timezone.py

@@ -0,0 +1,138 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/timezone.tsv from the IANA tzdb tarball (public domain) and GeoNames (CC BY 4.0).
+
+    data-import/timezone.py [--source URL_OR_FILE] [--cities URL_OR_FILE] [--cache DIR] [--out FILE] [--territories FILE]
+
+The offset is the zone's standard offset, the first field of its Zone rule's last
+continuation line, with a Link resolved to its target. The weight is the population
+GeoNames records in the zone's cities of 15,000 or more, never below that threshold,
+so a drawn zone is one people live in.
+"""
+import argparse
+import csv
+import io
+import re
+import sys
+import tarfile
+import zipfile
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://data.iana.org/time-zones/tzdata-latest.tar.gz"
+CITIES = "https://download.geonames.org/export/dump/cities15000.zip"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "timezone.tsv"
+TERRITORIES = Path(__file__).resolve().parent.parent / "data" / "misc" / "territory.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["offset", "territory", "weight", "zone"]
+REGIONS = ["africa", "antarctica", "asia", "australasia", "backward", "etcetera", "europe", "northamerica", "southamerica"]
+# GeoNames' own cutoff, so a zone it records no city in weighs less than one would.
+FLOOR = 15000
+
+
+def member(tar, name):
+    try:
+        f = tar.extractfile(name)
+    except KeyError:
+        f = None
+    if f is None:
+        sys.exit(f"{name}: the tarball no longer holds it as a file; the tzdb layout has moved")
+    return f.read().decode("utf-8")
+
+
+def offsets(tar):
+    """Every zone's standard offset, and every link's target."""
+    std, links = {}, {}
+    for name in REGIONS:
+        zone = None
+        for raw in member(tar, name).splitlines():
+            line = raw.split("#")[0].rstrip()
+            if not line.strip():
+                continue
+            if line.startswith("Zone"):
+                f = line.split()
+                zone, std[f[1]] = f[1], f[2]
+            elif line.startswith("Link"):
+                f = line.split()
+                links[f[2]], zone = f[1], None
+            elif line[0] in " \t" and zone:
+                std[zone] = line.split()[0]
+            else:
+                zone = None
+    return std, links
+
+
+def resolve(zone, std, links):
+    for _ in range(10):
+        if zone in std:
+            return std[zone]
+        if zone not in links:
+            return None
+        zone = links[zone]
+    return None
+
+
+def utc_offset(raw):
+    """±HH:MM from a tzdb STDOFF field, which writes the minutes and seconds only when it has them."""
+    m = re.match(r"^(-)?(\d{1,2})(?::(\d{2})(?::(\d{2}))?)?$", raw)
+    if not m or (m.group(4) or "00") != "00":
+        return None
+    return f"{'-' if m.group(1) else '+'}{int(m.group(2)):02d}:{m.group(3) or '00'}"
+
+
+def population(body):
+    """The population GeoNames records per timezone, summed over its cities."""
+    with zipfile.ZipFile(io.BytesIO(body)) as z:
+        text = z.read("cities15000.txt").decode("utf-8")
+    per = {}
+    for city in csv.reader(io.StringIO(text), delimiter="\t", quoting=csv.QUOTE_NONE):
+        if len(city) > 17 and city[17] and city[14].isdigit():
+            per[city[17]] = per.get(city[17], 0) + int(city[14])
+    return per
+
+
+def rows(tar, territories, pop):
+    std, links = offsets(tar)
+    for line in member(tar, "zone.tab").splitlines():
+        if line.startswith("#") or not line.strip():
+            continue
+        fields = line.split("\t")
+        if len(fields) < 3:
+            sys.exit(f"zone.tab: {line!r} has {len(fields)} fields; a row names a territory, a location and a zone")
+        territory, zone = fields[0], fields[2]
+        if territory not in territories:
+            continue
+        raw = resolve(zone, std, links)
+        if raw is None:
+            sys.exit(f"{zone}: the tarball gives it no Zone rule and no Link to one")
+        offset = utc_offset(raw)
+        if offset is None:
+            sys.exit(f"{zone}: standard offset {raw!r} is not a ±HH:MM the table can spell")
+        yield {"offset": offset, "territory": territory, "weight": max(pop.get(zone, 0), FLOOR), "zone": zone}
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--out", default=str(OUT))
+    p.add_argument("--cities", default=CITIES)
+    p.add_argument("--territories", default=str(TERRITORIES))
+    a = p.parse_args()
+    shipped = {r["alpha2"] for r in csv.DictReader(io.StringIO(Path(a.territories).read_text(encoding="utf-8")), delimiter="\t")}
+    pop = population(source.fetch(a.cities, a.cache, "cities15000.zip", magic=b"PK"))
+    body = source.fetch(a.source, a.cache, "tzdata-latest.tar.gz", magic=b"\x1f\x8b")
+    with tarfile.open(fileobj=io.BytesIO(body)) as tar:
+        table = sorted(rows(tar, shipped, pop), key=lambda r: r["zone"])
+        print(f"tzdb {tar.extractfile('version').read().decode('utf-8').strip()}", file=sys.stderr)
+    linked = {r["territory"] for r in table}
+    if missing := shipped - linked:
+        sys.exit(f"no zone for {sorted(missing)}; a parent row without a child is a load error")
+    if (floored := sum(1 for r in table if r["weight"] == FLOOR)) > len(table) // 2:
+        sys.exit(f"{floored} of {len(table)} zones fell to the floor; GeoNames' population or timezone column has moved")
+    tsv.write(a.out, COLUMNS, table)
+
+
+if __name__ == "__main__":
+    main()

data-import/tld.py

@@ -0,0 +1,68 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/tld.tsv from the IANA Root Zone Database.
+
+    data-import/tld.py [--source URL_OR_FILE] [--zone URL_OR_FILE] [--cache DIR] [--out FILE]
+
+A TLD ships when the database names a manager for it; one it marks "Not assigned" is a
+delegation the root zone no longer holds, and the zone file is fetched to prove the two
+agree. The key is the A-label the root zone holds; `unicode` is the form the database
+displays, less the bidi marks it wraps a right-to-left label in, and proved to punycode
+back to the key.
+"""
+import argparse
+import html
+import re
+import sys
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://www.iana.org/domains/root/db"
+ZONE = "https://data.iana.org/TLD/tlds-alpha-by-domain.txt"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "tld.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["tld", "type", "unicode"]
+TYPES = ("country-code", "generic", "generic-restricted", "infrastructure", "sponsored")
+BIDI = {0x200E: None, 0x200F: None}
+ROW = re.compile(r'<a href="/domains/root/db/([^"]+)\.html">([^<]+)</a>.*?</td>\s*<td>([^<]*)</td>\s*<td>(.*?)</td>', re.S)
+
+
+def rows(page):
+    for a_label, shown, kind, manager in ROW.findall(page):
+        if "Not assigned" in manager:
+            continue
+        kind = kind.strip()
+        if kind not in TYPES:
+            sys.exit(f"{a_label}: the database types it {kind!r}, which is none of {TYPES}")
+        yield {"tld": "." + a_label, "type": kind, "unicode": html.unescape(shown).strip().translate(BIDI)}
+
+
+def prove_each_unicode_form_is_its_key(table):
+    """The displayed form is the key's own spelling, which punycode encodes back to."""
+    for r in table:
+        label = r["unicode"][1:]
+        encoded = label if label.isascii() else "xn--" + label.encode("punycode").decode("ascii")
+        if "." + encoded != r["tld"]:
+            sys.exit(f"{r['tld']}: the database displays {r['unicode']!r}, which is no spelling of it")
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--zone", default=ZONE)
+    p.add_argument("--out", default=str(OUT))
+    a = p.parse_args()
+    table = sorted(rows(source.fetch(a.source, a.cache, "root-db.html").decode("utf-8")), key=lambda r: r["tld"])
+    prove_each_unicode_form_is_its_key(table)
+    listing = source.fetch(a.zone, a.cache, "tlds-alpha-by-domain.txt").decode("utf-8")
+    delegated = {"." + line.strip().lower() for line in listing.splitlines() if line.strip() and not line.startswith("#")}
+    shipped = {r["tld"] for r in table}
+    if len(table) != len(delegated) or shipped != delegated:
+        sys.exit(f"the database assigns {len(table)} rows where the root zone holds {len(delegated)} TLDs: {sorted(shipped ^ delegated)[:10]}")
+    tsv.write(a.out, COLUMNS, table)
+
+
+if __name__ == "__main__":
+    main()

data-import/tsv.py

@@ -1,40 +1,18 @@
-"""A source fetched once into the cache, and a table written as the loader admits it."""
+"""A table written as the loader admits it."""
 import re
 import sys
-import time
-import urllib.request
 from pathlib import Path
 
 
-def fetch(source, cache, name, magic=b"", data=None, headers=None):
-    """The bytes of a URL, downloaded into cache/name once, or of a local file."""
-    if not re.match(r"^https?://", source):
-        return Path(source).read_bytes()
-    path = Path(cache) / name
-    for attempt in range(1, 6):
-        if path.exists():
-            return path.read_bytes()
-        req = urllib.request.Request(source, data=data, headers={"User-Agent": "fejkdata data-import", **(headers or {})})
-        try:
-            with urllib.request.urlopen(req, timeout=600) as r:
-                body = r.read()
-        except OSError:
-            body = b""
-        if body and body.startswith(magic) and b"Request Rejected" not in body[:512]:
-            path.parent.mkdir(parents=True, exist_ok=True)
-            path.write_bytes(body)
-        elif attempt < 5:
-            time.sleep(10 * attempt)
-    sys.exit(f"{source}: no valid download in 5 attempts")
-
-
 def write(path, columns, rows):
-    """Write the rows as a TSV; every cell must be non-empty and free of tabs, newlines and braces."""
+    """Write the rows as a TSV; the table needs two rows, and every cell must be non-empty and free of tabs, newlines and braces."""
     lines = ["\t".join(columns)]
     for row in rows:
         cells = [str(row[c]) for c in columns]
         if not all(cells) or any(re.search(r"[\t\n{}]", c) for c in cells):
             raise ValueError(f"{path}: a cell is empty or holds a tab, newline or brace: {row}")
         lines.append("\t".join(cells))
+    if len(lines) < 3:
+        raise ValueError(f"{path}: has {len(lines) - 1} rows below its header; a table is at least two rows")
     Path(path).write_text("\n".join(lines) + "\n", encoding="utf-8")
     print(f"{path}: {len(lines) - 1} rows", file=sys.stderr)

data-import/useragent.py

@@ -0,0 +1,64 @@
+#!/usr/bin/env python3
+"""Rebuild data/misc/useragent.tsv from microlinkhq/top-user-agents (MIT).
+
+    data-import/useragent.py [--source URL] [--cache DIR] [--out FILE]
+
+The desktop and mobile lists partition the whole, so `device` is read rather than
+parsed. A row ships when the string names a browser and an operating system both.
+"""
+import argparse
+import json
+import re
+import sys
+from pathlib import Path
+
+import source
+import tsv
+
+SOURCE = "https://raw.githubusercontent.com/microlinkhq/top-user-agents/master/src"
+OUT = Path(__file__).resolve().parent.parent / "data" / "misc" / "useragent.tsv"
+CACHE = Path(__file__).resolve().parent / "cache"
+COLUMNS = ["browser", "device", "os", "ua"]
+# First match wins: a Chromium fork this list does not name ships as Chrome.
+BROWSERS = [("Edge", r"Edg(A|iOS)?/"), ("Opera", r"OPR/"), ("Samsung Internet", r"SamsungBrowser/"),
+            ("Chrome", r"(Chrome|CriOS)/"), ("Firefox", r"(Firefox|FxiOS)/"), ("Safari", r"Version/[\d.]+ .*Safari")]
+SYSTEMS = [("iOS", r"iPhone|iPad|CPU OS "), ("Android", r"Android"), ("ChromeOS", r"CrOS"),
+           ("Windows", r"Windows NT"), ("macOS", r"Macintosh|Mac OS X"), ("Linux", r"X11.*Linux|Ubuntu")]
+
+
+def named(table, ua):
+    for name, pattern in table:
+        if re.search(pattern, ua):
+            return name
+    return ""
+
+
+def rows(desktop, mobile):
+    seen, kept = set(), []
+    for device, uas in (("desktop", desktop), ("mobile", mobile)):
+        for ua in uas:
+            browser, os = named(BROWSERS, ua), named(SYSTEMS, ua)
+            if not browser or not os:
+                print(f"dropped, naming no {'browser' if not browser else 'operating system'}: {ua}", file=sys.stderr)
+                continue
+            if ua in seen:
+                sys.exit(f"{ua}: listed twice, which would draw it twice as often")
+            seen.add(ua)
+            kept.append({"browser": browser, "device": device, "os": os, "ua": ua})
+    if len(kept) < 0.8 * (len(desktop) + len(mobile)):
+        sys.exit(f"only {len(kept)} of {len(desktop) + len(mobile)} strings named both; the tokens have moved")
+    return kept
+
+
+def main():
+    p = argparse.ArgumentParser(description=__doc__.splitlines()[0])
+    p.add_argument("--cache", default=str(CACHE))
+    p.add_argument("--source", default=SOURCE)
+    p.add_argument("--out", default=str(OUT))
+    a = p.parse_args()
+    lists = [json.loads(source.fetch(f"{a.source}/{n}.json", a.cache, f"top-user-agents-{n}.json")) for n in ("desktop", "mobile")]
+    tsv.write(a.out, COLUMNS, sorted(rows(*lists), key=lambda r: (r["browser"], r["os"], r["ua"])))
+
+
+if __name__ == "__main__":
+    main()

data-import/xlsx.py

@@ -0,0 +1,21 @@
+"""The rows of an xlsx sheet."""
+import io
+import xml.etree.ElementTree as ET
+import zipfile
+
+NS = {"m": "http://schemas.openxmlformats.org/spreadsheetml/2006/main", "r": "http://schemas.openxmlformats.org/officeDocument/2006/relationships"}
+
+
+def rows(data, sheet=None):
+    """The rows of an xlsx sheet named sheet, the first sheet by default, each a list of cell texts."""
+    z = zipfile.ZipFile(io.BytesIO(data))
+    strings = ["".join(t.text or "" for t in si.iter("{%s}t" % NS["m"])) for si in ET.fromstring(z.read("xl/sharedStrings.xml")).findall("m:si", NS)]
+    rels = {r.get("Id"): r.get("Target") for r in ET.fromstring(z.read("xl/_rels/workbook.xml.rels"))}
+    sheets = {s.get("name"): rels[s.get("{%s}id" % NS["r"])] for s in ET.fromstring(z.read("xl/workbook.xml")).iter("{%s}sheet" % NS["m"])}
+    target = sheets[sheet] if sheet else next(iter(sheets.values()))
+    for row in ET.fromstring(z.read("xl/" + target)).findall(".//m:row", NS):
+        cells = []
+        for c in row.findall("m:c", NS):
+            v = c.find("m:v", NS)
+            cells.append("" if v is None else strings[int(v.text)] if c.get("t") == "s" else v.text)
+        yield cells

data.go

@@ -43,7 +43,7 @@ func loadData(sources []dataSource) (map[string]node, error) {
 			}
 			info, err := os.Stat(src.path)
 			if err != nil {
-				return nil, err
+				return nil, fmt.Errorf("data path %s: %w", src.path, err)
 			}
 			if !info.IsDir() {
 				return nil, fmt.Errorf("%s is not a directory", src.path)
@@ -62,6 +62,7 @@ func loadData(sources []dataSource) (map[string]node, error) {
 	if len(root) == 0 {
 		return nil, fmt.Errorf("no .json data found")
 	}
+	setTablePaths(root)
 	if err := linkTables(root); err != nil {
 		return nil, err
 	}

data/en_US/date.json

@@ -1,13 +1 @@
-{
+"{date(1970-01-01,2029-12-31,'01/02/2006')}"
-  "format": "{month}/{day}/{year}",
-  "month": ["01", "02", "03", "04", "05", "06", "07", "08", "09", "10", "11", "12"],
-  "day": ["01", "02", "03", "04", "05", "06", "07", "08", "09", "10", "11", "12", "13", "14", "15", "16", "17", "18", "19", "20", "21", "22", "23", "24", "25", "26", "27", "28"],
-  "year": [
-    "197{digits(1)}",
-    "198{digits(1)}",
-    { "format": "199{digits(1)}", "weight": 2 },
-    { "format": "200{digits(1)}", "weight": 3 },
-    { "format": "201{digits(1)}", "weight": 3 },
-    { "format": "202{digits(1)}", "weight": 2 }
-  ]
-}

data/en_US/first-name.json

@@ -0,0 +1 @@
+{ "format": "{name}", "rows": "first-name.tsv", "name": "name", "parent": "sex", "weight": "count" }

data/en_US/itin.json

@@ -0,0 +1,16 @@
+{
+  "format": "{area}-{group}-{serial}",
+  "area": "9{digits(2)}",
+  "group": [
+    { "format": "{int(50,65)}", "weight": 16 },
+    { "format": "{int(70,88)}", "weight": 19 },
+    { "format": "{int(90,92)}", "weight": 3 },
+    { "format": "{int(94,99)}", "weight": 6 }
+  ],
+  "serial": [
+    { "format": "000{int(1,9)}", "weight": 9 },
+    { "format": "00{int(10,99)}", "weight": 90 },
+    { "format": "0{int(100,999)}", "weight": 900 },
+    { "format": "{int(1000,9999)}", "weight": 9000 }
+  ]
+}

data/en_US/last-name.json

@@ -0,0 +1 @@
+{ "format": "{name}", "rows": "last-name.tsv", "key": "name", "weight": "count" }

data/en_US/person.json

@@ -1,7 +1,7 @@
 {
-  "format": "{prefix}{femalefirst|malefirst} {last}",
+  "format": "{prefix}{first} {last}",
-  "femalefirst": ["Abigail", "Addison", "Amelia", "Aria", "Aubrey", "Audrey", "Aurora", "Ava", "Bella", "Brooklyn", "Camila", "Caroline", "Charlotte", "Chloe", "Claire", "Eleanor", "Elizabeth", "Ella", "Ellie", "Emily", "Emma", "Evelyn", "Gianna", "Grace", "Hannah", "Harper", "Hazel", "Isabella", "Layla", "Leah", "Lillian", "Lily", "Lucy", "Luna", "Madison", "Maya", "Mia", "Mila", "Naomi", "Natalie", "Nora", "Olivia", "Paisley", "Penelope", "Riley", "Savannah", "Scarlett", "Sofia", "Sophia", "Stella", "Victoria", "Violet", "Zoe"],
+  "first": "{.first-name.name}",
-  "malefirst": ["Aiden", "Alexander", "Andrew", "Anthony", "Asher", "Benjamin", "Caleb", "Carter", "Charles", "Christopher", "Daniel", "David", "Dylan", "Elijah", "Ethan", "Ezra", "Gabriel", "Grayson", "Henry", "Isaac", "Jack", "Jackson", "Jacob", "James", "Jayden", "John", "Joseph", "Joshua", "Julian", "Levi", "Liam", "Lincoln", "Logan", "Lucas", "Luke", "Mason", "Mateo", "Matthew", "Michael", "Nathan", "Noah", "Oliver", "Owen", "Samuel", "Sebastian", "Theodore", "Thomas", "William", "Wyatt"],
+  "last": "{.last-name.name}",
-  "last": ["Adams", "Allen", "Anderson", "Bailey", "Baker", "Bell", "Bennett", "Brooks", "Brown", "Campbell", "Carter", "Clark", "Collins", "Cook", "Cooper", "Cox", "Davis", "Edwards", "Evans", "Flores", "Foster", "Garcia", "Gonzalez", "Gray", "Green", "Hall", "Harris", "Hayes", "Hernandez", "Hill", "Howard", "Hughes", "Jackson", "James", "Jenkins", "Johnson", "Jones", "Kelly", "King", "Lee", "Lewis", "Long", "Lopez", "Martin", "Martinez", "Miller", "Mitchell", "Moore", "Morgan", "Morris", "Murphy", "Nelson", "Parker", "Perez", "Perry", "Peterson", "Phillips", "Powell", "Price", "Ramirez", "Reed", "Richardson", "Rivera", "Roberts", "Robinson", "Rodriguez", "Rogers", "Ross", "Russell", "Sanchez", "Sanders", "Scott", "Smith", "Stewart", "Sullivan", "Taylor", "Thomas", "Thompson", "Torres", "Turner", "Walker", "Ward", "Watson", "White", "Williams", "Wilson", "Wood", "Wright", "Young"],
+  "prefix": [null, { "format": "{.title.name} ", "weight": 0.1 }],
-  "prefix": ["", { "format": "{title} ", "title": ["Dr", "Miss", "Mr", "Mrs", "Ms", "Mx", "Prof"], "weight": 0.1 }]
+  "sex": "{.sex.name}"
 }

data/en_US/sex.json

@@ -0,0 +1 @@
+{ "format": "{name}", "rows": "sex.tsv", "key": "code", "name": "name" }

data/en_US/sex.tsv

@@ -0,0 +1,3 @@
+code	name
+f	female
+m	male

data/en_US/ssn.json

@@ -1 +1,19 @@
-"{int(100,999)}-{digits(2)}-{digits(4)}"
+{
+  "format": "{area}-{group}-{serial}",
+  "area": [
+    { "format": "00{int(1,9)}", "weight": 9 },
+    { "format": "0{int(10,99)}", "weight": 90 },
+    { "format": "{int(100,665)}", "weight": 566 },
+    { "format": "{int(667,899)}", "weight": 233 }
+  ],
+  "group": [
+    { "format": "0{int(1,9)}", "weight": 9 },
+    { "format": "{int(10,99)}", "weight": 90 }
+  ],
+  "serial": [
+    { "format": "000{int(1,9)}", "weight": 9 },
+    { "format": "00{int(10,99)}", "weight": 90 },
+    { "format": "0{int(100,999)}", "weight": 900 },
+    { "format": "{int(1000,9999)}", "weight": 9000 }
+  ]
+}

data/en_US/time.json

@@ -1,6 +1 @@
-{
+"{time('3:04 PM')}"
-  "format": "{hour}:{minute} {ampm}",
-  "hour": ["1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12"],
-  "minute": { "format": "{t}{digits(1)}", "t": ["0", "1", "2", "3", "4", "5"] },
-  "ampm": ["AM", "PM"]
-}

data/en_US/title.json

@@ -0,0 +1 @@
+{ "format": "{name}", "rows": "title.tsv", "name": "name", "parent": "sex", "weight": "share" }

data/en_US/title.tsv

@@ -0,0 +1,11 @@
+name	sex	share
+Dr	f	6
+Dr	m	6
+Miss	f	4
+Mr	m	60
+Mrs	f	22
+Ms	f	30
+Mx	f	1
+Mx	m	1
+Prof	f	3
+Prof	m	3

data/geo/US/locality.tsv

@@ -224,7 +224,7 @@ code	name	municipality	population	lat	lon
 0648914	Monterey Park	06037	57457	34.04838	-118.133165
 0649138	Moorpark	06111	35023	34.285942	-118.87704
 0649270	Moreno Valley	06065	214263	33.923253	-117.205685
-0649582	Mountain House	06077	25315	37.779716	-121.537195
+0649278	Morgan Hill	06085	45625	37.132495	-121.641891
 0649670	Mountain View	06085	85438	37.399686	-122.079269
 0650076	Murrieta	06065	114124	33.573214	-117.191007
 0650258	Napa	06055	76562	38.297118	-122.301008
@@ -287,6 +287,7 @@ code	name	municipality	population	lat	lon
 0666070	San Dimas	06037	32978	34.107957	-117.808529
 0667000	San Francisco	06075	826079	37.727239	-123.032229
 0667042	San Gabriel	06037	38186	34.094636	-118.098469
+0667056	Sanger	06019	27228	36.700864	-119.55773
 0667112	San Jacinto	06065	56014	33.798864	-116.995914
 0668000	San Jose	06085	989814	37.296011	-121.814552
 0668028	San Juan Capistrano	06059	35119	33.500889	-117.654388
@@ -495,6 +496,7 @@ code	name	municipality	population	lat	lon
 1276500	West Melbourne	12009	31285	28.050398	-80.658327
 1276582	Weston	12011	69588	26.100076	-80.401266
 1276600	West Palm Beach	12099	127189	26.745114	-80.127038
+1278250	Winter Garden	12095	48063	28.541217	-81.597043
 1278275	Winter Haven	12105	62313	28.045836	-81.732743
 1278300	Winter Park	12095	30444	28.59663	-81.345748
 1278325	Winter Springs	12117	39585	28.695481	-81.272086
@@ -521,7 +523,6 @@ code	name	municipality	population	lat	lon
 1343192	Kennesaw	13067	37970	34.031195	-84.621287
 1344340	LaGrange	13285	34730	33.02698	-85.038261
 1345488	Lawrenceville	13135	32822	33.952875	-83.992231
-1348288	Mableton	13067	78820	33.812998	-84.576946
 1348624	McDonough	13151	33395	33.439611	-84.148899
 1349008	Macon	13021	157556	32.808995	-83.694051
 1349756	Marietta	13067	63574	33.953561	-84.542463
@@ -757,6 +758,7 @@ code	name	municipality	population	lat	lon
 2338740	Lewiston	23001	38866	44.089325	-70.172051
 2360545	Portland	23005	69911	43.633157	-70.185305
 2371990	South Portland	23005	27007	43.631402	-70.285989
+2401600	Annapolis	24003	40634	38.97216	-76.505334
 2404000	Baltimore	24510	569997	39.300061	-76.610518
 2408775	Bowie	24033	58383	38.954703	-76.738834
 2418750	College Park	24033	35418	38.996692	-76.936407
@@ -1037,6 +1039,7 @@ code	name	municipality	population	lat	lon
 3643335	Long Beach	36059	34918	40.585835	-73.665403
 3647042	Middletown	36071	30098	41.444922	-74.423573
 3649121	Mount Vernon	36119	72304	40.913105	-73.829317
+3650034	Newburgh	36071	28059	41.503249	-74.01964
 3650617	New Rochelle	36119	85864	40.923243	-73.779344
 3651000	New York	36047	8584629	40.662712	-73.938677
 3651055	Niagara Falls	36063	47255	43.092841	-79.014743
@@ -1472,6 +1475,7 @@ code	name	municipality	population	lat	lon
 5114968	Charlottesville	51540	44388	38.037658	-78.485381
 5116000	Chesapeake	51550	255332	36.679376	-76.301788
 5121344	Danville	51590	41647	36.581754	-79.408258
+5126496	Fairfax	51600	26772	38.853183	-77.299025
 5129744	Fredericksburg	51630	30393	38.299272	-77.486658
 5135000	Hampton	51650	137315	37.047961	-76.297293
 5135624	Harrisonburg	51660	50839	38.436255	-78.873303
@@ -1510,6 +1514,7 @@ code	name	municipality	population	lat	lon
 5338038	Lakewood	53053	62755	47.164092	-122.528106
 5340245	Longview	53015	38362	46.145707	-122.964299
 5340840	Lynnwood	53061	42132	47.828979	-122.300672
+5343150	Maple Valley	53033	28482	47.367147	-122.034815
 5343955	Marysville	53061	77686	48.080214	-122.154312
 5345005	Mercer Island	53033	25422	47.564004	-122.231214
 5347245	Moses Lake	53025	27469	47.12785	-119.276384

data/geo/US/municipality.tsv

@@ -255,6 +255,7 @@ code	name	region	population
 23001	Androscoggin County	ME	116487
 23005	Cumberland County	ME	317222
 23019	Penobscot County	ME	157967
+24003	Anne Arundel County	MD	603380
 24021	Frederick County	MD	302883
 24031	Montgomery County	MD	1074582
 24033	Prince George's County	MD	970374
@@ -601,6 +602,7 @@ code	name	region	population
 51540	Charlottesville city	VA	44388
 51550	Chesapeake city	VA	255332
 51590	Danville city	VA	41647
+51600	Fairfax city	VA	26772
 51630	Fredericksburg city	VA	30393
 51650	Hampton city	VA	137315
 51660	Harrisonburg city	VA	50839

data/misc/car.json

@@ -1,10 +1 @@
-[
+{ "format": "{make} {model}", "rows": "car.tsv" }
-  { "format": "{maker} {model}", "maker": "BMW", "model": ["3 Series", "5 Series", "X3", "X5"] },
-  { "format": "{maker} {model}", "maker": "Ford", "model": ["Fiesta", "Focus", "Mustang", "Explorer"] },
-  { "format": "{maker} {model}", "maker": "Honda", "model": ["Civic", "Accord", "CR-V", "Jazz"] },
-  { "format": "{maker} {model}", "maker": "Mercedes-Benz", "model": ["A-Class", "C-Class", "E-Class", "GLC"] },
-  { "format": "{maker} {model}", "maker": "Tesla", "model": ["Model 3", "Model S", "Model X", "Model Y"] },
-  { "format": "{maker} {model}", "maker": "Toyota", "model": ["Corolla", "Camry", "RAV4", "Yaris"] },
-  { "format": "{maker} {model}", "maker": "Volkswagen", "model": ["Golf", "Passat", "Polo", "Tiguan"] },
-  { "format": "{maker} {model}", "maker": "Volvo", "model": ["XC40", "XC60", "XC90", "V60"] }
-]

data/misc/car.tsv

@@ -0,0 +1,33 @@
+make	model
+BMW	3 Series
+BMW	5 Series
+BMW	X3
+BMW	X5
+Ford	Explorer
+Ford	Fiesta
+Ford	Focus
+Ford	Mustang
+Honda	Accord
+Honda	CR-V
+Honda	Civic
+Honda	Jazz
+Mercedes-Benz	A-Class
+Mercedes-Benz	C-Class
+Mercedes-Benz	E-Class
+Mercedes-Benz	GLC
+Tesla	Model 3
+Tesla	Model S
+Tesla	Model X
+Tesla	Model Y
+Toyota	Camry
+Toyota	Corolla
+Toyota	RAV4
+Toyota	Yaris
+Volkswagen	Golf
+Volkswagen	Passat
+Volkswagen	Polo
+Volkswagen	Tiguan
+Volvo	V60
+Volvo	XC40
+Volvo	XC60
+Volvo	XC90

data/misc/country.json

@@ -1 +0,0 @@
-{ "format": "{name}", "rows": "country.tsv", "key": "alpha2", "name": "name" }

data/misc/country.tsv

@@ -1,242 +0,0 @@
-alpha2	alpha3	calling-code	capital	currency	flag	languages	name	numeric	tld
-AD	AND	376	Andorra la Vella	EUR	🇦🇩	ca	Andorra	020	.ad
-AE	ARE	971	Abu Dhabi	AED	🇦🇪	ar-AE,fa,en,hi,ur	United Arab Emirates	784	.ae
-AF	AFG	93	Kabul	AFN	🇦🇫	fa-AF,ps,uz-AF,tk	Afghanistan	004	.af
-AG	ATG	1-268	St. John's	XCD	🇦🇬	en-AG	Antigua & Barbuda	028	.ag
-AI	AIA	1-264	The Valley	XCD	🇦🇮	en-AI	Anguilla	660	.ai
-AL	ALB	355	Tirana	ALL	🇦🇱	sq,el	Albania	008	.al
-AM	ARM	374	Yerevan	AMD	🇦🇲	hy	Armenia	051	.am
-AO	AGO	244	Luanda	AOA	🇦🇴	pt-AO	Angola	024	.ao
-AR	ARG	54	Buenos Aires	ARS	🇦🇷	es-AR,en,it,de,fr,gn	Argentina	032	.ar
-AS	ASM	1-684	Pago Pago	USD	🇦🇸	en-AS,sm,to	American Samoa	016	.as
-AT	AUT	43	Vienna	EUR	🇦🇹	de-AT,hr,hu,sl	Austria	040	.at
-AU	AUS	61	Canberra	AUD	🇦🇺	en-AU	Australia	036	.au
-AW	ABW	297	Oranjestad	AWG	🇦🇼	nl-AW,pap,es,en	Aruba	533	.aw
-AX	ALA	358	Mariehamn	EUR	🇦🇽	sv-AX	Åland Islands	248	.ax
-AZ	AZE	994	Baku	AZN	🇦🇿	az,ru,hy	Azerbaijan	031	.az
-BA	BIH	387	Sarajevo	BAM	🇧🇦	bs,hr-BA,sr-BA	Bosnia	070	.ba
-BB	BRB	1-246	Bridgetown	BBD	🇧🇧	en-BB	Barbados	052	.bb
-BD	BGD	880	Dhaka	BDT	🇧🇩	bn-BD,en	Bangladesh	050	.bd
-BE	BEL	32	Brussels	EUR	🇧🇪	nl-BE,fr-BE,de-BE	Belgium	056	.be
-BF	BFA	226	Ouagadougou	XOF	🇧🇫	fr-BF,mos	Burkina Faso	854	.bf
-BG	BGR	359	Sofia	EUR	🇧🇬	bg,tr-BG,rom	Bulgaria	100	.bg
-BH	BHR	973	Manama	BHD	🇧🇭	ar-BH,en,fa,ur	Bahrain	048	.bh
-BI	BDI	257	Gitega	BIF	🇧🇮	fr-BI,rn	Burundi	108	.bi
-BJ	BEN	229	Porto-Novo	XOF	🇧🇯	fr-BJ	Benin	204	.bj
-BL	BLM	590	Gustavia	EUR	🇧🇱	fr	St. Barthélemy	652	.gp
-BM	BMU	1-441	Hamilton	BMD	🇧🇲	en-BM,pt	Bermuda	060	.bm
-BN	BRN	673	Bandar Seri Begawan	BND	🇧🇳	ms-BN,en-BN	Brunei	096	.bn
-BO	BOL	591	Sucre	BOB	🇧🇴	es-BO,qu,ay	Bolivia	068	.bo
-BR	BRA	55	Brasilia	BRL	🇧🇷	pt-BR,es,en,fr	Brazil	076	.br
-BS	BHS	1-242	Nassau	BSD	🇧🇸	en-BS	Bahamas	044	.bs
-BT	BTN	975	Thimphu	INR	🇧🇹	dz	Bhutan	064	.bt
-BW	BWA	267	Gaborone	BWP	🇧🇼	en-BW,tn-BW	Botswana	072	.bw
-BY	BLR	375	Minsk	BYN	🇧🇾	be,ru	Belarus	112	.by
-BZ	BLZ	501	Belmopan	BZD	🇧🇿	en-BZ,es	Belize	084	.bz
-CA	CAN	1	Ottawa	CAD	🇨🇦	en-CA,fr-CA,iu	Canada	124	.ca
-CC	CCK	61	West Island	AUD	🇨🇨	ms-CC,en	Cocos Islands	166	.cc
-CD	COD	243	Kinshasa	CDF	🇨🇩	fr-CD,ln,ktu,kg,sw,lua	Congo - Kinshasa	180	.cd
-CF	CAF	236	Bangui	XAF	🇨🇫	fr-CF,sg,ln,kg	Central African Republic	140	.cf
-CG	COG	242	Brazzaville	XAF	🇨🇬	fr-CG,kg,ln-CG	Congo - Brazzaville	178	.cg
-CH	CHE	41	Bern	CHF	🇨🇭	de-CH,fr-CH,it-CH,rm	Switzerland	756	.ch
-CI	CIV	225	Yamoussoukro	XOF	🇨🇮	fr-CI	Côte d’Ivoire	384	.ci
-CK	COK	682	Avarua	NZD	🇨🇰	en-CK,mi	Cook Islands	184	.ck
-CL	CHL	56	Santiago	CLP	🇨🇱	es-CL	Chile	152	.cl
-CM	CMR	237	Yaounde	XAF	🇨🇲	en-CM,fr-CM	Cameroon	120	.cm
-CN	CHN	86	Beijing	CNY	🇨🇳	zh-CN,yue,wuu,dta,ug,za	China	156	.cn
-CO	COL	57	Bogota	COP	🇨🇴	es-CO	Colombia	170	.co
-CR	CRI	506	San Jose	CRC	🇨🇷	es-CR,en	Costa Rica	188	.cr
-CU	CUB	53	Havana	CUP	🇨🇺	es-CU,pap	Cuba	192	.cu
-CV	CPV	238	Praia	CVE	🇨🇻	pt-CV	Cape Verde	132	.cv
-CW	CUW	599	 Willemstad	XCG	🇨🇼	nl,pap	Curaçao	531	.cw
-CX	CXR	61	Flying Fish Cove	AUD	🇨🇽	en,zh,ms-CX	Christmas Island	162	.cx
-CY	CYP	357	Nicosia	EUR	🇨🇾	el-CY,tr-CY,en	Cyprus	196	.cy
-CZ	CZE	420	Prague	CZK	🇨🇿	cs,sk	Czechia	203	.cz
-DE	DEU	49	Berlin	EUR	🇩🇪	de	Germany	276	.de
-DJ	DJI	253	Djibouti	DJF	🇩🇯	fr-DJ,ar,so-DJ,aa	Djibouti	262	.dj
-DK	DNK	45	Copenhagen	DKK	🇩🇰	da-DK,en,fo,de-DK	Denmark	208	.dk
-DM	DMA	1-767	Roseau	XCD	🇩🇲	en-DM	Dominica	212	.dm
-DO	DOM	1-809	Santo Domingo	DOP	🇩🇴	es-DO	Dominican Republic	214	.do
-DZ	DZA	213	Algiers	DZD	🇩🇿	ar-DZ	Algeria	012	.dz
-EC	ECU	593	Quito	USD	🇪🇨	es-EC	Ecuador	218	.ec
-EE	EST	372	Tallinn	EUR	🇪🇪	et,ru	Estonia	233	.ee
-EG	EGY	20	Cairo	EGP	🇪🇬	ar-EG,en,fr	Egypt	818	.eg
-EH	ESH	212	El-Aaiun	MAD	🇪🇭	ar,mey	Western Sahara	732	.eh
-ER	ERI	291	Asmara	ERN	🇪🇷	aa-ER,ar,tig,kun,ti-ER	Eritrea	232	.er
-ES	ESP	34	Madrid	EUR	🇪🇸	es-ES,ca,gl,eu,oc	Spain	724	.es
-ET	ETH	251	Addis Ababa	ETB	🇪🇹	am,en-ET,om-ET,ti-ET,so-ET,sid	Ethiopia	231	.et
-FI	FIN	358	Helsinki	EUR	🇫🇮	fi-FI,sv-FI,smn	Finland	246	.fi
-FJ	FJI	679	Suva	FJD	🇫🇯	en-FJ,fj	Fiji	242	.fj
-FK	FLK	500	Stanley	FKP	🇫🇰	en-FK	Falkland Islands	238	.fk
-FM	FSM	691	Palikir	USD	🇫🇲	en-FM,chk,pon,yap,kos,uli,woe,nkr,kpg	Micronesia	583	.fm
-FO	FRO	298	Torshavn	DKK	🇫🇴	fo,da-FO	Faroe Islands	234	.fo
-FR	FRA	33	Paris	EUR	🇫🇷	fr-FR,frp,br,co,ca,eu,oc	France	250	.fr
-GA	GAB	241	Libreville	XAF	🇬🇦	fr-GA	Gabon	266	.ga
-GB	GBR	44	London	GBP	🇬🇧	en-GB,cy-GB,gd	UK	826	.uk
-GD	GRD	1-473	St. George's	XCD	🇬🇩	en-GD	Grenada	308	.gd
-GE	GEO	995	Tbilisi	GEL	🇬🇪	ka,ru,hy,az	Georgia	268	.ge
-GF	GUF	594	Cayenne	EUR	🇬🇫	fr-GF	French Guiana	254	.gf
-GG	GGY	44	St Peter Port	GBP	🇬🇬	en,nrf	Guernsey	831	.gg
-GH	GHA	233	Accra	GHS	🇬🇭	en-GH,ak,ee,tw	Ghana	288	.gh
-GI	GIB	350	Gibraltar	GIP	🇬🇮	en-GI,es,it,pt	Gibraltar	292	.gi
-GL	GRL	299	Nuuk	DKK	🇬🇱	kl,da-GL,en	Greenland	304	.gl
-GM	GMB	220	Banjul	GMD	🇬🇲	en-GM,mnk,wof,wo,ff	Gambia	270	.gm
-GN	GIN	224	Conakry	GNF	🇬🇳	fr-GN	Guinea	324	.gn
-GP	GLP	590	Basse-Terre	EUR	🇬🇵	fr-GP	Guadeloupe	312	.gp
-GQ	GNQ	240	Ciudad de la Paz	XAF	🇬🇶	es-GQ,fr,pt	Equatorial Guinea	226	.gq
-GR	GRC	30	Athens	EUR	🇬🇷	el-GR,en,fr	Greece	300	.gr
-GT	GTM	502	Guatemala City	GTQ	🇬🇹	es-GT	Guatemala	320	.gt
-GU	GUM	1-671	Hagatna	USD	🇬🇺	en-GU,ch-GU	Guam	316	.gu
-GW	GNB	245	Bissau	XOF	🇬🇼	pt-GW,pov	Guinea-Bissau	624	.gw
-GY	GUY	592	Georgetown	GYD	🇬🇾	en-GY	Guyana	328	.gy
-HK	HKG	852	Hong Kong	HKD	🇭🇰	zh-HK,yue,zh,en	Hong Kong	344	.hk
-HN	HND	504	Tegucigalpa	HNL	🇭🇳	es-HN,cab,miq	Honduras	340	.hn
-HR	HRV	385	Zagreb	EUR	🇭🇷	hr-HR,sr	Croatia	191	.hr
-HT	HTI	509	Port-au-Prince	HTG	🇭🇹	ht,fr-HT	Haiti	332	.ht
-HU	HUN	36	Budapest	HUF	🇭🇺	hu-HU	Hungary	348	.hu
-ID	IDN	62	Jakarta	IDR	🇮🇩	id,en,nl,jv	Indonesia	360	.id
-IE	IRL	353	Dublin	EUR	🇮🇪	en-IE,ga-IE	Ireland	372	.ie
-IL	ISR	972	Jerusalem	ILS	🇮🇱	he,ar-IL,en-IL	Israel	376	.il
-IM	IMN	44	Douglas	GBP	🇮🇲	en,gv	Isle of Man	833	.im
-IN	IND	91	New Delhi	INR	🇮🇳	en-IN,hi,bn,te,mr,ta,ur,gu,kn,ml,or,pa,as,bh,sat,ks,ne,sd,kok,doi,mni,sit,sa,fr,lus,inc	India	356	.in
-IO	IOT	246	Diego Garcia	USD	🇮🇴	en-IO	British Indian Ocean Territory	086	.io
-IQ	IRQ	964	Baghdad	IQD	🇮🇶	ar-IQ,ku,hy	Iraq	368	.iq
-IR	IRN	98	Tehran	IRR	🇮🇷	fa-IR,ku	Iran	364	.ir
-IS	ISL	354	Reykjavik	ISK	🇮🇸	is,en,de,da,sv,no	Iceland	352	.is
-IT	ITA	39	Rome	EUR	🇮🇹	it-IT,de-IT,fr-IT,sc,ca,co,sl	Italy	380	.it
-JE	JEY	44	Saint Helier	GBP	🇯🇪	en,fr,nrf	Jersey	832	.je
-JM	JAM	1-876	Kingston	JMD	🇯🇲	en-JM	Jamaica	388	.jm
-JO	JOR	962	Amman	JOD	🇯🇴	ar-JO,en	Jordan	400	.jo
-JP	JPN	81	Tokyo	JPY	🇯🇵	ja	Japan	392	.jp
-KE	KEN	254	Nairobi	KES	🇰🇪	en-KE,sw-KE	Kenya	404	.ke
-KG	KGZ	996	Bishkek	KGS	🇰🇬	ky,uz,ru	Kyrgyzstan	417	.kg
-KH	KHM	855	Phnom Penh	KHR	🇰🇭	km,fr,en	Cambodia	116	.kh
-KI	KIR	686	Tarawa	AUD	🇰🇮	en-KI,gil	Kiribati	296	.ki
-KM	COM	269	Moroni	KMF	🇰🇲	ar,fr-KM	Comoros	174	.km
-KN	KNA	1-869	Basseterre	XCD	🇰🇳	en-KN	St. Kitts & Nevis	659	.kn
-KP	PRK	850	Pyongyang	KPW	🇰🇵	ko-KP	North Korea	408	.kp
-KR	KOR	82	Seoul	KRW	🇰🇷	ko-KR,en	South Korea	410	.kr
-KW	KWT	965	Kuwait City	KWD	🇰🇼	ar-KW,en	Kuwait	414	.kw
-KY	CYM	1-345	George Town	KYD	🇰🇾	en-KY	Cayman Islands	136	.ky
-KZ	KAZ	7	Nur-Sultan	KZT	🇰🇿	kk,ru	Kazakhstan	398	.kz
-LA	LAO	856	Vientiane	LAK	🇱🇦	lo,fr,en	Laos	418	.la
-LB	LBN	961	Beirut	LBP	🇱🇧	ar-LB,fr-LB,en,hy	Lebanon	422	.lb
-LC	LCA	1-758	Castries	XCD	🇱🇨	en-LC	St. Lucia	662	.lc
-LI	LIE	423	Vaduz	CHF	🇱🇮	de-LI	Liechtenstein	438	.li
-LK	LKA	94	Colombo	LKR	🇱🇰	si,ta,en	Sri Lanka	144	.lk
-LR	LBR	231	Monrovia	LRD	🇱🇷	en-LR	Liberia	430	.lr
-LS	LSO	266	Maseru	LSL	🇱🇸	en-LS,st,zu,xh	Lesotho	426	.ls
-LT	LTU	370	Vilnius	EUR	🇱🇹	lt,ru,pl	Lithuania	440	.lt
-LU	LUX	352	Luxembourg	EUR	🇱🇺	lb,de-LU,fr-LU	Luxembourg	442	.lu
-LV	LVA	371	Riga	EUR	🇱🇻	lv,ru,lt	Latvia	428	.lv
-LY	LBY	218	Tripoli	LYD	🇱🇾	ar-LY,it,en	Libya	434	.ly
-MA	MAR	212	Rabat	MAD	🇲🇦	ar-MA,ber,fr	Morocco	504	.ma
-MC	MCO	377	Monaco	EUR	🇲🇨	fr-MC,en,it	Monaco	492	.mc
-MD	MDA	373	Chisinau	MDL	🇲🇩	ro,ru,gag,tr	Moldova	498	.md
-ME	MNE	382	Podgorica	EUR	🇲🇪	sr,hu,bs,sq,hr,rom	Montenegro	499	.me
-MF	MAF	590	Marigot	EUR	🇲🇫	fr	St. Martin	663	.gp
-MG	MDG	261	Antananarivo	MGA	🇲🇬	fr-MG,mg	Madagascar	450	.mg
-MH	MHL	692	Majuro	USD	🇲🇭	mh,en-MH	Marshall Islands	584	.mh
-MK	MKD	389	Skopje	MKD	🇲🇰	mk,sq,tr,rmm,sr	North Macedonia	807	.mk
-ML	MLI	223	Bamako	XOF	🇲🇱	fr-ML,bm	Mali	466	.ml
-MM	MMR	95	Nay Pyi Taw	MMK	🇲🇲	my	Myanmar	104	.mm
-MN	MNG	976	Ulaanbaatar	MNT	🇲🇳	mn,ru	Mongolia	496	.mn
-MO	MAC	853	Macao	MOP	🇲🇴	zh,zh-MO,pt	Macao	446	.mo
-MP	MNP	1-670	Saipan	USD	🇲🇵	fil,tl,zh,ch-MP,en-MP	Northern Mariana Islands	580	.mp
-MQ	MTQ	596	Fort-de-France	EUR	🇲🇶	fr-MQ	Martinique	474	.mq
-MR	MRT	222	Nouakchott	MRU	🇲🇷	ar-MR,fuc,snk,fr,mey,wo	Mauritania	478	.mr
-MS	MSR	1-664	Plymouth	XCD	🇲🇸	en-MS	Montserrat	500	.ms
-MT	MLT	356	Valletta	EUR	🇲🇹	mt,en-MT	Malta	470	.mt
-MU	MUS	230	Port Louis	MUR	🇲🇺	en-MU,bho,fr	Mauritius	480	.mu
-MV	MDV	960	Male	MVR	🇲🇻	dv,en	Maldives	462	.mv
-MW	MWI	265	Lilongwe	MWK	🇲🇼	ny,yao,tum,swk	Malawi	454	.mw
-MX	MEX	52	Mexico City	MXN	🇲🇽	es-MX	Mexico	484	.mx
-MY	MYS	60	Kuala Lumpur	MYR	🇲🇾	ms-MY,en,zh,ta,te,ml,pa,th	Malaysia	458	.my
-MZ	MOZ	258	Maputo	MZN	🇲🇿	pt-MZ,vmw	Mozambique	508	.mz
-NA	NAM	264	Windhoek	NAD	🇳🇦	en-NA,af,de,hz,naq	Namibia	516	.na
-NC	NCL	687	Noumea	XPF	🇳🇨	fr-NC	New Caledonia	540	.nc
-NE	NER	227	Niamey	XOF	🇳🇪	fr-NE,ha,kr,dje	Niger	562	.ne
-NF	NFK	672	Kingston	AUD	🇳🇫	en-NF	Norfolk Island	574	.nf
-NG	NGA	234	Abuja	NGN	🇳🇬	en-NG,ha,yo,ig,ff	Nigeria	566	.ng
-NI	NIC	505	Managua	NIO	🇳🇮	es-NI,en	Nicaragua	558	.ni
-NL	NLD	31	Amsterdam	EUR	🇳🇱	nl-NL,fy-NL	Netherlands	528	.nl
-NO	NOR	47	Oslo	NOK	🇳🇴	no,nb,nn,se,fi	Norway	578	.no
-NP	NPL	977	Kathmandu	NPR	🇳🇵	ne,en	Nepal	524	.np
-NR	NRU	674	Yaren	AUD	🇳🇷	na,en-NR	Nauru	520	.nr
-NU	NIU	683	Alofi	NZD	🇳🇺	niu,en-NU	Niue	570	.nu
-NZ	NZL	64	Wellington	NZD	🇳🇿	en-NZ,mi	New Zealand	554	.nz
-OM	OMN	968	Muscat	OMR	🇴🇲	ar-OM,en,bal,ur	Oman	512	.om
-PA	PAN	507	Panama City	PAB	🇵🇦	es-PA,en	Panama	591	.pa
-PE	PER	51	Lima	PEN	🇵🇪	es-PE,qu,ay	Peru	604	.pe
-PF	PYF	689	Papeete	XPF	🇵🇫	fr-PF,ty	French Polynesia	258	.pf
-PG	PNG	675	Port Moresby	PGK	🇵🇬	en-PG,ho,meu,tpi	Papua New Guinea	598	.pg
-PH	PHL	63	Manila	PHP	🇵🇭	tl,en-PH,fil,ceb,ilo,hil,war,pam,bik,bcl,pag,mrw,tsg,mdh,cbk,krj,sgd,msb,akl,ibg,yka,mta,abx	Philippines	608	.ph
-PK	PAK	92	Islamabad	PKR	🇵🇰	ur-PK,en-PK,pa,sd,ps,brh	Pakistan	586	.pk
-PL	POL	48	Warsaw	PLN	🇵🇱	pl	Poland	616	.pl
-PM	SPM	508	Saint-Pierre	EUR	🇵🇲	fr-PM	St. Pierre & Miquelon	666	.pm
-PN	PCN	870	Adamstown	NZD	🇵🇳	en-PN	Pitcairn	612	.pn
-PR	PRI	1	San Juan	USD	🇵🇷	en-PR,es-PR	Puerto Rico	630	.pr
-PT	PRT	351	Lisbon	EUR	🇵🇹	pt-PT,mwl	Portugal	620	.pt
-PW	PLW	680	Melekeok	USD	🇵🇼	pau,sov,en-PW,tox,ja,fil,zh	Palau	585	.pw
-PY	PRY	595	Asuncion	PYG	🇵🇾	es-PY,gn	Paraguay	600	.py
-QA	QAT	974	Doha	QAR	🇶🇦	ar-QA,es	Qatar	634	.qa
-RE	REU	262	Saint-Denis	EUR	🇷🇪	fr-RE	Réunion	638	.re
-RO	ROU	40	Bucharest	RON	🇷🇴	ro,hu,rom	Romania	642	.ro
-RS	SRB	381	Belgrade	RSD	🇷🇸	sr,hu,bs,rom	Serbia	688	.rs
-RU	RUS	7	Moscow	RUB	🇷🇺	ru,tt,xal,cau,ady,kv,ce,tyv,cv,udm,tut,mns,bua,myv,mdf,chm,ba,inh,kbd,krc,av,sah,nog	Russia	643	.ru
-RW	RWA	250	Kigali	RWF	🇷🇼	rw,en-RW,fr-RW,sw	Rwanda	646	.rw
-SA	SAU	966	Riyadh	SAR	🇸🇦	ar-SA	Saudi Arabia	682	.sa
-SB	SLB	677	Honiara	SBD	🇸🇧	en-SB,tpi	Solomon Islands	090	.sb
-SC	SYC	248	Victoria	SCR	🇸🇨	en-SC,fr-SC	Seychelles	690	.sc
-SD	SDN	249	Khartoum	SDG	🇸🇩	ar-SD,en,fia	Sudan	729	.sd
-SE	SWE	46	Stockholm	SEK	🇸🇪	sv-SE,se,sma,fi-SE	Sweden	752	.se
-SG	SGP	65	Singapore	SGD	🇸🇬	cmn,en-SG,ms-SG,ta-SG,zh-SG	Singapore	702	.sg
-SH	SHN	290	Jamestown	SHP	🇸🇭	en-SH	St. Helena	654	.sh
-SI	SVN	386	Ljubljana	EUR	🇸🇮	sl,sh	Slovenia	705	.si
-SJ	SJM	47	Longyearbyen	NOK	🇸🇯	no,ru	Svalbard & Jan Mayen	744	.sj
-SK	SVK	421	Bratislava	EUR	🇸🇰	sk,hu	Slovakia	703	.sk
-SL	SLE	232	Freetown	SLE	🇸🇱	en-SL,men,tem	Sierra Leone	694	.sl
-SM	SMR	378	San Marino	EUR	🇸🇲	it-SM	San Marino	674	.sm
-SN	SEN	221	Dakar	XOF	🇸🇳	fr-SN,wo,fuc,mnk	Senegal	686	.sn
-SO	SOM	252	Mogadishu	SOS	🇸🇴	so-SO,ar-SO,it,en-SO	Somalia	706	.so
-SR	SUR	597	Paramaribo	SRD	🇸🇷	nl-SR,en,srn,hns,jv	Suriname	740	.sr
-SS	SSD	211	Juba	SSP	🇸🇸	en	South Sudan	728	.ss
-ST	STP	239	Sao Tome	STN	🇸🇹	pt-ST	São Tomé & Príncipe	678	.st
-SV	SLV	503	San Salvador	SVC	🇸🇻	es-SV	El Salvador	222	.sv
-SX	SXM	1-721	Philipsburg	XCG	🇸🇽	nl,en	Sint Maarten	534	.sx
-SY	SYR	963	Damascus	SYP	🇸🇾	ar-SY,ku,hy,arc,fr,en	Syria	760	.sy
-SZ	SWZ	268	Mbabane	SZL	🇸🇿	en-SZ,ss-SZ	Eswatini	748	.sz
-TC	TCA	1-649	Cockburn Town	USD	🇹🇨	en-TC	Turks & Caicos Islands	796	.tc
-TD	TCD	235	N'Djamena	XAF	🇹🇩	fr-TD,ar-TD,sre	Chad	148	.td
-TF	ATF	262	Port-aux-Francais	EUR	🇹🇫	fr	French Southern Territories	260	.tf
-TG	TGO	228	Lome	XOF	🇹🇬	fr-TG,ee,hna,kbp,dag,ha	Togo	768	.tg
-TH	THA	66	Bangkok	THB	🇹🇭	th,en	Thailand	764	.th
-TJ	TJK	992	Dushanbe	TJS	🇹🇯	tg,ru	Tajikistan	762	.tj
-TL	TLS	670	Dili	USD	🇹🇱	tet,pt-TL,id,en	Timor-Leste	626	.tl
-TM	TKM	993	Ashgabat	TMT	🇹🇲	tk,ru,uz	Turkmenistan	795	.tm
-TN	TUN	216	Tunis	TND	🇹🇳	ar-TN,fr	Tunisia	788	.tn
-TO	TON	676	Nuku'alofa	TOP	🇹🇴	to,en-TO	Tonga	776	.to
-TR	TUR	90	Ankara	TRY	🇹🇷	tr-TR,ku,diq,az,av	Türkiye	792	.tr
-TT	TTO	1-868	Port of Spain	TTD	🇹🇹	en-TT,hns,fr,es,zh	Trinidad & Tobago	780	.tt
-TV	TUV	688	Funafuti	AUD	🇹🇻	tvl,en,sm,gil	Tuvalu	798	.tv
-TW	TWN	886	Taipei	TWD	🇹🇼	zh-TW,zh,nan,hak	Taiwan	158	.tw
-TZ	TZA	255	Dodoma	TZS	🇹🇿	sw-TZ,en,ar	Tanzania	834	.tz
-UA	UKR	380	Kyiv	UAH	🇺🇦	uk,ru-UA,rom,pl,hu	Ukraine	804	.ua
-UG	UGA	256	Kampala	UGX	🇺🇬	en-UG,lg,sw,ar	Uganda	800	.ug
-US	USA	1	Washington	USD	🇺🇸	en-US,es-US,haw,fr	US	840	.us
-UY	URY	598	Montevideo	UYU	🇺🇾	es-UY	Uruguay	858	.uy
-UZ	UZB	998	Tashkent	UZS	🇺🇿	uz,ru,tg	Uzbekistan	860	.uz
-VA	VAT	39	Vatican City	EUR	🇻🇦	la,it,fr	Vatican City	336	.va
-VC	VCT	1-784	Kingstown	XCD	🇻🇨	en-VC,fr	St. Vincent & Grenadines	670	.vc
-VE	VEN	58	Caracas	VES	🇻🇪	es-VE	Venezuela	862	.ve
-VG	VGB	1-284	Road Town	USD	🇻🇬	en-VG	British Virgin Islands	092	.vg
-VI	VIR	1-340	Charlotte Amalie	USD	🇻🇮	en-VI	U.S. Virgin Islands	850	.vi
-VN	VNM	84	Hanoi	VND	🇻🇳	vi,en,fr,zh,km	Vietnam	704	.vn
-VU	VUT	678	Port Vila	VUV	🇻🇺	bi,en-VU,fr-VU	Vanuatu	548	.vu
-WF	WLF	681	Mata Utu	XPF	🇼🇫	wls,fud,fr-WF	Wallis & Futuna	876	.wf
-WS	WSM	685	Apia	WST	🇼🇸	sm,en-WS	Samoa	882	.ws
-YE	YEM	967	Sanaa	YER	🇾🇪	ar-YE	Yemen	887	.ye
-YT	MYT	262	Mamoudzou	EUR	🇾🇹	fr-YT	Mayotte	175	.yt
-ZA	ZAF	27	Pretoria	ZAR	🇿🇦	zu,xh,af,nso,en-ZA,tn,st,ts,ss,ve,nr	South Africa	710	.za
-ZM	ZMB	260	Lusaka	ZMW	🇿🇲	en-ZM,bem,loz,lun,lue,ny,toi	Zambia	894	.zm
-ZW	ZWE	263	Harare	ZWG	🇿🇼	en-ZW,sn,nr,nd	Zimbabwe	716	.zw

data/misc/datetime.json

@@ -0,0 +1 @@
+"{date(2000-01-01,2029-12-31,'2006-01-02T15:04:05Z')}"

data/misc/httpmethod.json

@@ -0,0 +1 @@
+{ "format": "{method}", "rows": "httpmethod.tsv", "key": "method" }

data/misc/httpmethod.tsv

@@ -0,0 +1,10 @@
+idempotent	method	safe
+false	CONNECT	false
+true	DELETE	false
+true	GET	true
+true	HEAD	true
+true	OPTIONS	true
+false	PATCH	false
+false	POST	false
+true	PUT	false
+true	TRACE	true

data/misc/httpstatus.tsv

@@ -1,17 +1,61 @@
 code	reason
+100	Continue
+101	Switching Protocols
+102	Processing
+103	Early Hints
 200	OK
 201	Created
+202	Accepted
+203	Non-Authoritative Information
 204	No Content
+205	Reset Content
+206	Partial Content
+207	Multi-Status
+208	Already Reported
+226	IM Used
+300	Multiple Choices
 301	Moved Permanently
 302	Found
+303	See Other
 304	Not Modified
+305	Use Proxy
+307	Temporary Redirect
+308	Permanent Redirect
 400	Bad Request
 401	Unauthorized
+402	Payment Required
 403	Forbidden
 404	Not Found
+405	Method Not Allowed
+406	Not Acceptable
+407	Proxy Authentication Required
+408	Request Timeout
 409	Conflict
-422	Unprocessable Entity
+410	Gone
+411	Length Required
+412	Precondition Failed
+413	Content Too Large
+414	URI Too Long
+415	Unsupported Media Type
+416	Range Not Satisfiable
+417	Expectation Failed
+421	Misdirected Request
+422	Unprocessable Content
+423	Locked
+424	Failed Dependency
+425	Too Early
+426	Upgrade Required
+428	Precondition Required
 429	Too Many Requests
+431	Request Header Fields Too Large
+451	Unavailable For Legal Reasons
 500	Internal Server Error
+501	Not Implemented
 502	Bad Gateway
 503	Service Unavailable
+504	Gateway Timeout
+505	HTTP Version Not Supported
+506	Variant Also Negotiates
+507	Insufficient Storage
+508	Loop Detected
+511	Network Authentication Required

data/misc/language.tsv

@@ -1,17 +1,184 @@
-code	name
+code	code3	name
-ar	Arabic
+aa	aar	Afar
-zh	Chinese
+ab	abk	Abkhazian
-da	Danish
+ae	ave	Avestan
-nl	Dutch
+af	afr	Afrikaans
-en	English
+ak	aka	Akan
-fi	Finnish
+am	amh	Amharic
-fr	French
+an	arg	Aragonese
-de	German
+ar	ara	Arabic
-it	Italian
+as	asm	Assamese
-ja	Japanese
+av	ava	Avaric
-no	Norwegian
+ay	aym	Aymara
-pl	Polish
+az	aze	Azerbaijani
-pt	Portuguese
+ba	bak	Bashkir
-ru	Russian
+be	bel	Belarusian
-es	Spanish
+bg	bul	Bulgarian
-sv	Swedish
+bi	bis	Bislama
+bm	bam	Bambara
+bn	ben	Bengali
+bo	bod	Tibetan
+br	bre	Breton
+bs	bos	Bosnian
+ca	cat	Catalan
+ce	che	Chechen
+ch	cha	Chamorro
+co	cos	Corsican
+cr	cre	Cree
+cs	ces	Czech
+cu	chu	Church Slavic
+cv	chv	Chuvash
+cy	cym	Welsh
+da	dan	Danish
+de	deu	German
+dv	div	Divehi
+dz	dzo	Dzongkha
+ee	ewe	Ewe
+el	ell	Modern Greek
+en	eng	English
+eo	epo	Esperanto
+es	spa	Spanish
+et	est	Estonian
+eu	eus	Basque
+fa	fas	Persian
+ff	ful	Fulah
+fi	fin	Finnish
+fj	fij	Fijian
+fo	fao	Faroese
+fr	fra	French
+fy	fry	Western Frisian
+ga	gle	Irish
+gd	gla	Gaelic
+gl	glg	Galician
+gn	grn	Guarani
+gu	guj	Gujarati
+gv	glv	Manx
+ha	hau	Hausa
+he	heb	Hebrew
+hi	hin	Hindi
+ho	hmo	Hiri Motu
+hr	hrv	Croatian
+ht	hat	Haitian
+hu	hun	Hungarian
+hy	hye	Armenian
+hz	her	Herero
+ia	ina	Interlingua
+id	ind	Indonesian
+ie	ile	Interlingue
+ig	ibo	Igbo
+ii	iii	Sichuan Yi
+ik	ipk	Inupiaq
+io	ido	Ido
+is	isl	Icelandic
+it	ita	Italian
+iu	iku	Inuktitut
+ja	jpn	Japanese
+jv	jav	Javanese
+ka	kat	Georgian
+kg	kon	Kongo
+ki	kik	Kikuyu
+kj	kua	Kuanyama
+kk	kaz	Kazakh
+kl	kal	Kalaallisut
+km	khm	Central Khmer
+kn	kan	Kannada
+ko	kor	Korean
+kr	kau	Kanuri
+ks	kas	Kashmiri
+ku	kur	Kurdish
+kv	kom	Komi
+kw	cor	Cornish
+ky	kir	Kirghiz
+la	lat	Latin
+lb	ltz	Luxembourgish
+lg	lug	Ganda
+li	lim	Limburgan
+ln	lin	Lingala
+lo	lao	Lao
+lt	lit	Lithuanian
+lu	lub	Luba-Katanga
+lv	lav	Latvian
+mg	mlg	Malagasy
+mh	mah	Marshallese
+mi	mri	Maori
+mk	mkd	Macedonian
+ml	mal	Malayalam
+mn	mon	Mongolian
+mr	mar	Marathi
+ms	msa	Malay
+mt	mlt	Maltese
+my	mya	Burmese
+na	nau	Nauru
+nb	nob	Norwegian Bokmål
+nd	nde	North Ndebele
+ne	nep	Nepali
+ng	ndo	Ndonga
+nl	nld	Dutch
+nn	nno	Norwegian Nynorsk
+no	nor	Norwegian
+nr	nbl	South Ndebele
+nv	nav	Navajo
+ny	nya	Chichewa
+oc	oci	Occitan
+oj	oji	Ojibwa
+om	orm	Oromo
+or	ori	Oriya
+os	oss	Ossetian
+pa	pan	Panjabi
+pi	pli	Pali
+pl	pol	Polish
+ps	pus	Pushto
+pt	por	Portuguese
+qu	que	Quechua
+rm	roh	Romansh
+rn	run	Rundi
+ro	ron	Romanian
+ru	rus	Russian
+rw	kin	Kinyarwanda
+sa	san	Sanskrit
+sc	srd	Sardinian
+sd	snd	Sindhi
+se	sme	Northern Sami
+sg	sag	Sango
+si	sin	Sinhala
+sk	slk	Slovak
+sl	slv	Slovenian
+sm	smo	Samoan
+sn	sna	Shona
+so	som	Somali
+sq	sqi	Albanian
+sr	srp	Serbian
+ss	ssw	Swati
+st	sot	Sotho, Southern
+su	sun	Sundanese
+sv	swe	Swedish
+sw	swa	Swahili
+ta	tam	Tamil
+te	tel	Telugu
+tg	tgk	Tajik
+th	tha	Thai
+ti	tir	Tigrinya
+tk	tuk	Turkmen
+tl	tgl	Tagalog
+tn	tsn	Tswana
+to	ton	Tonga
+tr	tur	Turkish
+ts	tso	Tsonga
+tt	tat	Tatar
+tw	twi	Twi
+ty	tah	Tahitian
+ug	uig	Uighur
+uk	ukr	Ukrainian
+ur	urd	Urdu
+uz	uzb	Uzbek
+ve	ven	Venda
+vi	vie	Vietnamese
+vo	vol	Volapük
+wa	wln	Walloon
+wo	wol	Wolof
+xh	xho	Xhosa
+yi	yid	Yiddish
+yo	yor	Yoruba
+za	zha	Zhuang
+zh	zho	Chinese
+zu	zul	Zulu

data/misc/loglevel.json

@@ -0,0 +1 @@
+{ "format": "{keyword}", "rows": "loglevel.tsv", "key": "code", "name": "keyword" }

data/misc/loglevel.tsv

@@ -0,0 +1,9 @@
+code	keyword	severity
+0	emerg	Emergency
+1	alert	Alert
+2	crit	Critical
+3	err	Error
+4	warning	Warning
+5	notice	Notice
+6	info	Informational
+7	debug	Debug

data/misc/mimetype.tsv

@@ -1,16 +1,702 @@
 ext	type
+.ez	application/andrew-inset
+.atom	application/atom+xml
+.atomcat	application/atomcat+xml
+.atomdeleted	application/atomdeleted+xml
+.atomsvc	application/atomsvc+xml
+.dwd	application/atsc-dwd+xml
+.held	application/atsc-held+xml
+.rsat	application/atsc-rsat+xml
+.aml	application/automationml-aml+xml
+.amlx	application/automationml-amlx+zip
+.xcs	application/calendar+xml
+.ccxml	application/ccxml+xml
+.cdfx	application/cdfx+xml
+.cdmia	application/cdmi-capability
+.cdmic	application/cdmi-container
+.cdmid	application/cdmi-domain
+.cdmio	application/cdmi-object
+.cdmiq	application/cdmi-queue
+.cpl	application/cpl+xml
+.cwl	application/cwl
+.mpd	application/dash+xml
+.mpp	application/dash-patch+xml
+.davmount	application/davmount+xml
+.dcm	application/dicom
+.dssc	application/dssc+der
+.xdssc	application/dssc+xml
+.emma	application/emma+xml
+.emotionml	application/emotionml+xml
+.epub	application/epub+zip
+.exi	application/exi
+.exp	application/express
+.fdf	application/fdf
+.fdt	application/fdt+xml
+.pfr	application/font-tdpfr
+.geojson	application/geo+json
+.gml	application/gml+xml
 .gz	application/gzip
+.stk	application/hyperstudio
+.ink	application/inkml+xml
+.ipfix	application/ipfix
+.its	application/its+xml
+.jar	application/java-archive
 .json	application/json
+.kbl	application/kbl+xml
+.jsonld	application/ld+json
+.lgr	application/lgr+xml
+.lostxml	application/lost+xml
+.hqx	application/mac-binhex40
+.mads	application/mads+xml
+.webmanifest	application/manifest+json
+.mrc	application/marc
+.mrcx	application/marcxml+xml
+.ma	application/mathematica
+.mathml	application/mathml+xml
+.mbox	application/mbox
+.mpf	application/media-policy-dataset+xml
+.mscml	application/mediaservercontrol+xml
+.meta4	application/metalink4+xml
+.mets	application/mets+xml
+.maei	application/mmt-aei+xml
+.musd	application/mmt-usd+xml
+.mods	application/mods+xml
+.m21	application/mp21
+.mp4s	application/mp4
+.doc	application/msword
+.mxf	application/mxf
+.nq	application/n-quads
+.nt	application/n-triples
+.cjs	application/node
+.bin	application/octet-stream
+.oda	application/oda
+.opf	application/oebps-package+xml
+.ogx	application/ogg
+.oxps	application/oxps
+.relo	application/p2p-overlay+xml
+.xer	application/patch-ops-error+xml
 .pdf	application/pdf
+.pgp	application/pgp-encrypted
+.asc	application/pgp-keys
+.sig	application/pgp-signature
+.p10	application/pkcs10
+.p7m	application/pkcs7-mime
+.p7s	application/pkcs7-signature
+.p8	application/pkcs8
+.ac	application/pkix-attr-cert
+.cer	application/pkix-cert
+.crl	application/pkix-crl
+.pkipath	application/pkix-pkipath
+.pki	application/pkixcmp
+.pls	application/pls+xml
+.ai	application/postscript
+.provx	application/provenance+xml
+.cww	application/prs.cww
+.xsf	application/prs.xsf+xml
+.pskcxml	application/pskc+xml
+.rdf	application/rdf+xml
+.rif	application/reginfo+xml
+.rnc	application/relax-ng-compact-syntax
+.rl	application/resource-lists+xml
+.rld	application/resource-lists-diff+xml
+.rs	application/rls-services+xml
+.rapd	application/route-apd+xml
+.sls	application/route-s-tsid+xml
+.rusd	application/route-usd+xml
+.gbr	application/rpki-ghostbusters
+.mft	application/rpki-manifest
+.roa	application/rpki-roa
+.rtf	application/rtf
+.sbml	application/sbml+xml
+.scq	application/scvp-cv-request
+.scs	application/scvp-cv-response
+.spq	application/scvp-vp-request
+.spp	application/scvp-vp-response
+.sdp	application/sdp
+.senmlx	application/senml+xml
+.sensmlx	application/sensml+xml
+.setpay	application/set-payment-initiation
+.setreg	application/set-registration-initiation
+.shf	application/shf+xml
+.siv	application/sieve
+.smi	application/smil+xml
+.rq	application/sparql-query
+.srx	application/sparql-results+xml
+.sql	application/sql
+.gram	application/srgs
+.grxml	application/srgs+xml
+.sru	application/sru+xml
+.ssml	application/ssml+xml
+.swidtag	application/swid+xml
+.tei	application/tei+xml
+.tfi	application/thraud+xml
+.tsd	application/timestamped-data
+.toml	application/toml
+.trig	application/trig
+.ttml	application/ttml+xml
+.rsheet	application/urc-ressheet+xml
+.td	application/urc-targetdesc+xml
+.vec	application/vec+xml
+.1km	application/vnd.1000minds.decision-model+xml
+.plb	application/vnd.3gpp.pic-bw-large
+.psb	application/vnd.3gpp.pic-bw-small
+.pvb	application/vnd.3gpp.pic-bw-var
+.tcap	application/vnd.3gpp2.tcap
+.pwn	application/vnd.3m.post-it-notes
+.aso	application/vnd.accpac.simply.aso
+.imp	application/vnd.accpac.simply.imp
+.acu	application/vnd.acucobol
+.atc	application/vnd.acucorp
+.fcdt	application/vnd.adobe.formscentral.fcdt
+.fxp	application/vnd.adobe.fxp
+.xdp	application/vnd.adobe.xdp+xml
+.age	application/vnd.age
+.ahead	application/vnd.ahead.space
+.azf	application/vnd.airzip.filesecure.azf
+.azs	application/vnd.airzip.filesecure.azs
+.acc	application/vnd.americandynamics.acc
+.ami	application/vnd.amiga.ami
+.cii	application/vnd.anser-web-certificate-issue-initiation
+.atx	application/vnd.antix.game-component
+.parquet	application/vnd.apache.parquet
+.mpkg	application/vnd.apple.installer+xml
+.key	application/vnd.apple.keynote
+.m3u8	application/vnd.apple.mpegurl
+.numbers	application/vnd.apple.numbers
+.pages	application/vnd.apple.pages
+.swi	application/vnd.aristanetworks.swi
+.iota	application/vnd.astraea-software.iota
+.aep	application/vnd.audiograph
+.bmml	application/vnd.balsamiq.bmml+xml
+.mpm	application/vnd.blueice.multipass
+.bmi	application/vnd.bmi
+.rep	application/vnd.businessobjects
+.cdxml	application/vnd.chemdraw+xml
+.mmd	application/vnd.chipnuts.karaoke-mmd
+.cdy	application/vnd.cinderella
+.csl	application/vnd.citationstyles.style+xml
+.cla	application/vnd.claymore
+.rp9	application/vnd.cloanto.rp9
+.c4g	application/vnd.clonk.c4group
+.c11amc	application/vnd.cluetrust.cartomobile-config
+.c11amz	application/vnd.cluetrust.cartomobile-config-pkg
+.csp	application/vnd.commonspace
+.cdbcmsg	application/vnd.contact.cmsg
+.cmc	application/vnd.cosmocaller
+.clkx	application/vnd.crick.clicker
+.clkk	application/vnd.crick.clicker.keyboard
+.clkp	application/vnd.crick.clicker.palette
+.clkt	application/vnd.crick.clicker.template
+.clkw	application/vnd.crick.clicker.wordbank
+.wbs	application/vnd.criticaltools.wbs+xml
+.pml	application/vnd.ctc-posml
+.ppd	application/vnd.cups-ppd
+.dart	application/vnd.dart
+.rdz	application/vnd.data-vision.rdz
+.dbf	application/vnd.dbf
+.dcmp	application/vnd.dcmp+xml
+.uvf	application/vnd.dece.data
+.uvt	application/vnd.dece.ttml+xml
+.uvx	application/vnd.dece.unspecified
+.uvz	application/vnd.dece.zip
+.fe_launch	application/vnd.denovo.fcselayout-link
+.dna	application/vnd.dna
+.dpg	application/vnd.dpgraph
+.dfac	application/vnd.dreamfactory
+.ait	application/vnd.dvb.ait
+.svc	application/vnd.dvb.service
+.geo	application/vnd.dynageo
+.mag	application/vnd.ecowin.chart
+.nml	application/vnd.enliven
+.esf	application/vnd.epson.esf
+.msf	application/vnd.epson.msf
+.qam	application/vnd.epson.quickanime
+.slt	application/vnd.epson.salt
+.ssf	application/vnd.epson.ssf
+.es3	application/vnd.eszigno3+xml
+.ez2	application/vnd.ezpix-album
+.ez3	application/vnd.ezpix-package
+.mseed	application/vnd.fdsn.mseed
+.seed	application/vnd.fdsn.seed
+.gph	application/vnd.flographit
+.ftc	application/vnd.fluxtime.clip
+.fm	application/vnd.framemaker
+.fsc	application/vnd.fsc.weblaunch
+.oas	application/vnd.fujitsu.oasys
+.oa2	application/vnd.fujitsu.oasys2
+.oa3	application/vnd.fujitsu.oasys3
+.fg5	application/vnd.fujitsu.oasysgp
+.bh2	application/vnd.fujitsu.oasysprs
+.ddd	application/vnd.fujixerox.ddd
+.xdw	application/vnd.fujixerox.docuworks
+.xbd	application/vnd.fujixerox.docuworks.binder
+.fzs	application/vnd.fuzzysheet
+.txd	application/vnd.genomatix.tuxedo
+.ggb	application/vnd.geogebra.file
+.ggs	application/vnd.geogebra.slides
+.ggt	application/vnd.geogebra.tool
+.gex	application/vnd.geometry-explorer
+.gxt	application/vnd.geonext
+.g2w	application/vnd.geoplan
+.g3w	application/vnd.geospace
+.gmx	application/vnd.gmx
+.kml	application/vnd.google-earth.kml+xml
+.kmz	application/vnd.google-earth.kmz
+.xdcf	application/vnd.gov.sk.xmldatacontainer+xml
+.gqf	application/vnd.grafeq
+.gac	application/vnd.groove-account
+.ghf	application/vnd.groove-help
+.gim	application/vnd.groove-identity-message
+.grv	application/vnd.groove-injector
+.gtm	application/vnd.groove-tool-message
+.tpl	application/vnd.groove-tool-template
+.vcg	application/vnd.groove-vcard
+.hal	application/vnd.hal+xml
+.zmm	application/vnd.handheld-entertainment+xml
+.hbci	application/vnd.hbci
+.les	application/vnd.hhe.lesson-player
+.hpgl	application/vnd.hp-hpgl
+.hpid	application/vnd.hp-hpid
+.hps	application/vnd.hp-hps
+.jlt	application/vnd.hp-jlyt
+.pcl	application/vnd.hp-pcl
+.pclxl	application/vnd.hp-pclxl
+.sfd-hdstx	application/vnd.hydrostatix.sof-data
+.mpy	application/vnd.ibm.minipay
+.irm	application/vnd.ibm.rights-management
+.sc	application/vnd.ibm.secure-container
+.icc	application/vnd.iccprofile
+.igl	application/vnd.igloader
+.ivp	application/vnd.immervision-ivp
+.ivu	application/vnd.immervision-ivu
+.igm	application/vnd.insors.igm
+.xpw	application/vnd.intercon.formnet
+.i2g	application/vnd.intergeo
+.qbo	application/vnd.intu.qbo
+.qfx	application/vnd.intu.qfx
+.rcprofile	application/vnd.ipunplugged.rcprofile
+.irp	application/vnd.irepository.package+xml
+.xpr	application/vnd.is-xpr
+.fcs	application/vnd.isac.fcs
+.jam	application/vnd.jam
+.rms	application/vnd.jcp.javame.midlet-rms
+.jisp	application/vnd.jisp
+.joda	application/vnd.joost.joda-archive
+.ktz	application/vnd.kahootz
+.karbon	application/vnd.kde.karbon
+.chrt	application/vnd.kde.kchart
+.kfo	application/vnd.kde.kformula
+.flw	application/vnd.kde.kivio
+.kon	application/vnd.kde.kontour
+.kpr	application/vnd.kde.kpresenter
+.ksp	application/vnd.kde.kspread
+.kwd	application/vnd.kde.kword
+.htke	application/vnd.kenameaapp
+.kia	application/vnd.kidspiration
+.kne	application/vnd.kinar
+.skp	application/vnd.koan
+.sse	application/vnd.kodak-descriptor
+.lasxml	application/vnd.las.las+xml
+.lbd	application/vnd.llamagraphics.life-balance.desktop
+.lbe	application/vnd.llamagraphics.life-balance.exchange+xml
+.123	application/vnd.lotus-1-2-3
+.apr	application/vnd.lotus-approach
+.pre	application/vnd.lotus-freelance
+.nsf	application/vnd.lotus-notes
+.org	application/vnd.lotus-organizer
+.scm	application/vnd.lotus-screencam
+.lwp	application/vnd.lotus-wordpro
+.portpkg	application/vnd.macports.portpkg
+.mvt	application/vnd.mapbox-vector-tile
+.mcd	application/vnd.mcd
+.mc1	application/vnd.medcalcdata
+.cdkey	application/vnd.mediastation.cdkey
+.mwf	application/vnd.mfer
+.mfm	application/vnd.mfmp
+.flo	application/vnd.micrografx.flo
+.igx	application/vnd.micrografx.igx
+.mif	application/vnd.mif
+.daf	application/vnd.mobius.daf
+.dis	application/vnd.mobius.dis
+.mbk	application/vnd.mobius.mbk
+.mqy	application/vnd.mobius.mqy
+.msl	application/vnd.mobius.msl
+.plc	application/vnd.mobius.plc
+.txf	application/vnd.mobius.txf
+.mpn	application/vnd.mophun.application
+.mpc	application/vnd.mophun.certificate
+.xul	application/vnd.mozilla.xul+xml
+.cil	application/vnd.ms-artgalry
+.cab	application/vnd.ms-cab-compressed
+.xls	application/vnd.ms-excel
+.xlam	application/vnd.ms-excel.addin.macroenabled.12
+.xlsb	application/vnd.ms-excel.sheet.binary.macroenabled.12
+.xlsm	application/vnd.ms-excel.sheet.macroenabled.12
+.xltm	application/vnd.ms-excel.template.macroenabled.12
+.eot	application/vnd.ms-fontobject
+.chm	application/vnd.ms-htmlhelp
+.ims	application/vnd.ms-ims
+.lrm	application/vnd.ms-lrm
+.thmx	application/vnd.ms-officetheme
+.ppt	application/vnd.ms-powerpoint
+.ppam	application/vnd.ms-powerpoint.addin.macroenabled.12
+.pptm	application/vnd.ms-powerpoint.presentation.macroenabled.12
+.sldm	application/vnd.ms-powerpoint.slide.macroenabled.12
+.ppsm	application/vnd.ms-powerpoint.slideshow.macroenabled.12
+.potm	application/vnd.ms-powerpoint.template.macroenabled.12
+.mpp	application/vnd.ms-project
+.docm	application/vnd.ms-word.document.macroenabled.12
+.dotm	application/vnd.ms-word.template.macroenabled.12
+.wps	application/vnd.ms-works
+.wpl	application/vnd.ms-wpl
+.xps	application/vnd.ms-xpsdocument
+.mseq	application/vnd.mseq
+.mus	application/vnd.musician
+.msty	application/vnd.muvee.style
+.taglet	application/vnd.mynfc
+.bdo	application/vnd.nato.bindingdataobject+xml
+.nlu	application/vnd.neurolanguage.nlu
+.ntf	application/vnd.nitf
+.nnd	application/vnd.noblenet-directory
+.nns	application/vnd.noblenet-sealer
+.nnw	application/vnd.noblenet-web
+.ac	application/vnd.nokia.n-gage.ac+xml
+.ngdat	application/vnd.nokia.n-gage.data
+.rpst	application/vnd.nokia.radio-preset
+.rpss	application/vnd.nokia.radio-presets
+.edm	application/vnd.novadigm.edm
+.edx	application/vnd.novadigm.edx
+.ext	application/vnd.novadigm.ext
+.odc	application/vnd.oasis.opendocument.chart
+.otc	application/vnd.oasis.opendocument.chart-template
+.odf	application/vnd.oasis.opendocument.formula
+.odft	application/vnd.oasis.opendocument.formula-template
+.odg	application/vnd.oasis.opendocument.graphics
+.otg	application/vnd.oasis.opendocument.graphics-template
+.odi	application/vnd.oasis.opendocument.image
+.oti	application/vnd.oasis.opendocument.image-template
+.odp	application/vnd.oasis.opendocument.presentation
+.otp	application/vnd.oasis.opendocument.presentation-template
+.ods	application/vnd.oasis.opendocument.spreadsheet
+.ots	application/vnd.oasis.opendocument.spreadsheet-template
+.odt	application/vnd.oasis.opendocument.text
+.odm	application/vnd.oasis.opendocument.text-master
+.ott	application/vnd.oasis.opendocument.text-template
+.oth	application/vnd.oasis.opendocument.text-web
+.xo	application/vnd.olpc-sugar
+.dd2	application/vnd.oma.dd2+xml
+.obgx	application/vnd.openblox.game+xml
+.osm	application/vnd.openstreetmap.data+xml
+.pptx	application/vnd.openxmlformats-officedocument.presentationml.presentation
+.sldx	application/vnd.openxmlformats-officedocument.presentationml.slide
+.ppsx	application/vnd.openxmlformats-officedocument.presentationml.slideshow
+.potx	application/vnd.openxmlformats-officedocument.presentationml.template
+.xlsx	application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
+.xltx	application/vnd.openxmlformats-officedocument.spreadsheetml.template
+.docx	application/vnd.openxmlformats-officedocument.wordprocessingml.document
+.dotx	application/vnd.openxmlformats-officedocument.wordprocessingml.template
+.mgp	application/vnd.osgeo.mapguide.package
+.dp	application/vnd.osgi.dp
+.esa	application/vnd.osgi.subsystem
+.pdb	application/vnd.palm
+.paw	application/vnd.pawaafile
+.str	application/vnd.pg.format
+.ei6	application/vnd.pg.osasli
+.efif	application/vnd.picsel
+.wg	application/vnd.pmi.widget
+.plf	application/vnd.pocketlearn
+.pbd	application/vnd.powerbuilder6
+.systemverify	application/vnd.pp.systemverify+xml
+.box	application/vnd.previewsystems.box
+.mgz	application/vnd.proteus.magazine
+.qps	application/vnd.publishare-delta-tree
+.ptid	application/vnd.pvi.ptid1
+.xhtm	application/vnd.pwg-xhtml-print+xml
+.qxd	application/vnd.quark.quarkxpress
+.rar	application/vnd.rar
+.bed	application/vnd.realvnc.bed
+.mxl	application/vnd.recordare.musicxml
+.musicxml	application/vnd.recordare.musicxml+xml
+.cryptonote	application/vnd.rig.cryptonote
+.link66	application/vnd.route66.link66+xml
+.st	application/vnd.sailingtracker.track
+.see	application/vnd.seemail
+.sema	application/vnd.sema
+.semd	application/vnd.semd
+.semf	application/vnd.semf
+.ifm	application/vnd.shana.informed.formdata
+.itp	application/vnd.shana.informed.formtemplate
+.iif	application/vnd.shana.informed.interchange
+.ipk	application/vnd.shana.informed.package
+.twd	application/vnd.simtech-mindmapper
+.mmf	application/vnd.smaf
+.teacher	application/vnd.smart.teacher
+.fo	application/vnd.software602.filler.form+xml
+.sdkm	application/vnd.solent.sdkm+xml
+.dxp	application/vnd.spotfire.dxp
+.sfs	application/vnd.spotfire.sfs
+.sqlite	application/vnd.sqlite3
+.smzip	application/vnd.stepmania.package
+.sm	application/vnd.stepmania.stepchart
+.wadl	application/vnd.sun.wadl+xml
+.sus	application/vnd.sus-calendar
+.svd	application/vnd.svd
+.xsm	application/vnd.syncml+xml
+.bdm	application/vnd.syncml.dm+wbxml
+.xdm	application/vnd.syncml.dm+xml
+.ddf	application/vnd.syncml.dmddf+xml
+.tao	application/vnd.tao.intent-module-archive
+.pcap	application/vnd.tcpdump.pcap
+.tmo	application/vnd.tmobile-livetv
+.tpt	application/vnd.trid.tpt
+.mxs	application/vnd.triscape.mxs
+.tra	application/vnd.trueapp
+.ufd	application/vnd.ufdl
+.utz	application/vnd.uiq.theme
+.umj	application/vnd.umajin
+.unityweb	application/vnd.unity
+.uoml	application/vnd.uoml+xml
+.vcx	application/vnd.vcx
+.vsd	application/vnd.visio
+.vis	application/vnd.visionary
+.vsf	application/vnd.vsf
+.wbxml	application/vnd.wap.wbxml
+.wmlc	application/vnd.wap.wmlc
+.wmlsc	application/vnd.wap.wmlscriptc
+.wtb	application/vnd.webturbo
+.nbp	application/vnd.wolfram.player
+.wpd	application/vnd.wordperfect
+.wqd	application/vnd.wqd
+.stf	application/vnd.wt.stf
+.xar	application/vnd.xara
+.xfdl	application/vnd.xfdl
+.hvd	application/vnd.yamaha.hv-dic
+.hvs	application/vnd.yamaha.hv-script
+.hvp	application/vnd.yamaha.hv-voice
+.osf	application/vnd.yamaha.openscoreformat
+.osfpvg	application/vnd.yamaha.openscoreformat.osfpvg+xml
+.saf	application/vnd.yamaha.smaf-audio
+.spf	application/vnd.yamaha.smaf-phrase
+.cmp	application/vnd.yellowriver-custom-menu
+.zir	application/vnd.zul
+.zaz	application/vnd.zzazz.deck+xml
+.vxml	application/voicexml+xml
+.wasm	application/wasm
+.wif	application/watcherinfo+xml
+.wgt	application/widget
+.wsdl	application/wsdl+xml
+.wspolicy	application/wspolicy+xml
+.der	application/x-x509-ca-cert
+.xav	application/xcap-att+xml
+.xca	application/xcap-caps+xml
+.xdf	application/xcap-diff+xml
+.xel	application/xcap-el+xml
+.xns	application/xcap-ns+xml
+.xenc	application/xenc+xml
+.xfdf	application/xfdf
+.xhtml	application/xhtml+xml
+.xlf	application/xliff+xml
 .xml	application/xml
+.dtd	application/xml-dtd
+.xop	application/xop+xml
+.xsl	application/xslt+xml
+.mxml	application/xv+xml
+.yang	application/yang
+.yin	application/yin+xml
 .zip	application/zip
+.3gpp	audio/3gpp
+.adts	audio/aac
+.amr	audio/amr
+.au	audio/basic
+.mka	audio/matroska
+.mxmf	audio/mobile-xmf
+.m4a	audio/mp4
 .mp3	audio/mpeg
+.ogg	audio/ogg
+.uva	audio/vnd.dece.audio
+.eol	audio/vnd.digital-winds
+.dra	audio/vnd.dra
+.dts	audio/vnd.dts
+.dtshd	audio/vnd.dts.hd
+.lvp	audio/vnd.lucent.voice
+.pya	audio/vnd.ms-playready.media.pya
+.ecelp4800	audio/vnd.nuera.ecelp4800
+.ecelp7470	audio/vnd.nuera.ecelp7470
+.ecelp9600	audio/vnd.nuera.ecelp9600
+.rip	audio/vnd.rip
+.ttc	font/collection
+.otf	font/otf
+.ttf	font/ttf
+.woff	font/woff
+.woff2	font/woff2
+.exr	image/aces
+.apng	image/apng
+.avci	image/avci
+.avcs	image/avcs
+.avif	image/avif
+.bmp	image/bmp
+.cgm	image/cgm
+.drle	image/dicom-rle
+.dpx	image/dpx
+.emf	image/emf
+.fits	image/fits
+.g3	image/g3fax
 .gif	image/gif
+.heic	image/heic
+.heics	image/heic-sequence
+.heif	image/heif
+.heifs	image/heif-sequence
+.hej2	image/hej2k
+.ief	image/ief
+.jaii	image/jaii
+.jais	image/jais
+.jls	image/jls
+.jp2	image/jp2
 .jpg	image/jpeg
+.jph	image/jph
+.jhc	image/jphc
+.jpm	image/jpm
+.jpx	image/jpx
+.jxl	image/jxl
+.jxr	image/jxr
+.jxra	image/jxra
+.jxrs	image/jxrs
+.jxs	image/jxs
+.jxsc	image/jxsc
+.jxsi	image/jxsi
+.jxss	image/jxss
+.ktx	image/ktx
+.ktx2	image/ktx2
 .png	image/png
+.btif	image/prs.btif
+.pti	image/prs.pti
 .svg	image/svg+xml
+.t38	image/t38
+.tif	image/tiff
+.tfx	image/tiff-fx
+.psd	image/vnd.adobe.photoshop
+.azv	image/vnd.airzip.accelerator.azv
+.facti	image/vnd.blockfact.facti
+.uvi	image/vnd.dece.graphic
+.djvu	image/vnd.djvu
+.sub	image/vnd.dvb.subtitle
+.dwg	image/vnd.dwg
+.dxf	image/vnd.dxf
+.fbs	image/vnd.fastbidsheet
+.fpx	image/vnd.fpx
+.fst	image/vnd.fst
+.mmr	image/vnd.fujixerox.edmics-mmr
+.rlc	image/vnd.fujixerox.edmics-rlc
+.ico	image/vnd.microsoft.icon
+.mdi	image/vnd.ms-modi
+.npx	image/vnd.net-fpx
+.b16	image/vnd.pco.b16
+.tap	image/vnd.tencent.tap
+.vtf	image/vnd.valve.source.texture
+.wbmp	image/vnd.wap.wbmp
+.xif	image/vnd.xiff
+.pcx	image/vnd.zbrush.pcx
 .webp	image/webp
+.wmf	image/wmf
+.disposition-notification	message/disposition-notification
+.u8msg	message/global
+.u8dsn	message/global-delivery-status
+.u8mdn	message/global-disposition-notification
+.u8hdr	message/global-headers
+.eml	message/rfc822
+.wsc	message/vnd.wfa.wsc
+.3mf	model/3mf
+.gltf	model/gltf+json
+.glb	model/gltf-binary
+.igs	model/iges
+.jt	model/jt
+.msh	model/mesh
+.mtl	model/mtl
+.obj	model/obj
+.prc	model/prc
+.step	model/step
+.stpx	model/step+xml
+.stpz	model/step+zip
+.stpxz	model/step-xml+zip
+.stl	model/stl
+.u3d	model/u3d
+.bary	model/vnd.bary
+.cld	model/vnd.cld
+.dae	model/vnd.collada+xml
+.dwf	model/vnd.dwf
+.gdl	model/vnd.gdl
+.gtw	model/vnd.gtw
+.mts	model/vnd.mts
+.ogex	model/vnd.opengex
+.x_b	model/vnd.parasolid.transmit.binary
+.x_t	model/vnd.parasolid.transmit.text
+.pyo	model/vnd.pytha.pyox
+.vds	model/vnd.sap.vds
+.usda	model/vnd.usda
+.usdz	model/vnd.usdz+zip
+.bsp	model/vnd.valve.source.compiled-map
+.vtu	model/vnd.vtu
+.wrl	model/vrml
+.x3db	model/x3d+fastinfoset
+.x3d	model/x3d+xml
+.x3dv	model/x3d-vrml
+.appcache	text/cache-manifest
+.ics	text/calendar
+.css	text/css
 .csv	text/csv
 .html	text/html
+.js	text/javascript
+.md	text/markdown
+.n3	text/n3
 .txt	text/plain
+.dsc	text/prs.lines.tag
+.rtx	text/richtext
+.rtf	text/rtf
+.sgml	text/sgml
+.shex	text/shex
+.spdx	text/spdx
+.tsv	text/tab-separated-values
+.t	text/troff
+.ttl	text/turtle
+.uri	text/uri-list
+.vcard	text/vcard
+.curl	text/vnd.curl
+.sub	text/vnd.dvb.subtitle
+.ged	text/vnd.familysearch.gedcom
+.fly	text/vnd.fly
+.flx	text/vnd.fmi.flexstor
+.gv	text/vnd.graphviz
+.3dml	text/vnd.in3d.3dml
+.spot	text/vnd.in3d.spot
+.jad	text/vnd.sun.j2me.app-descriptor
+.wml	text/vnd.wap.wml
+.wmls	text/vnd.wap.wmlscript
+.vtt	text/vtt
+.wgsl	text/wgsl
+.xml	text/xml
+.3gp	video/3gpp
+.3g2	video/3gpp2
+.h261	video/h261
+.h263	video/h263
+.h264	video/h264
+.m4s	video/iso.segment
+.jpgv	video/jpeg
+.mkv	video/matroska
+.mk3d	video/matroska-3d
+.mj2	video/mj2
+.ts	video/mp2t
 .mp4	video/mp4
+.mpeg	video/mpeg
+.ogv	video/ogg
+.mov	video/quicktime
+.uvh	video/vnd.dece.hd
+.uvm	video/vnd.dece.mobile
+.uvp	video/vnd.dece.pd
+.uvs	video/vnd.dece.sd
+.uvv	video/vnd.dece.video
+.dvb	video/vnd.dvb.file
+.fvt	video/vnd.fvt
+.mxu	video/vnd.mpegurl
+.pyv	video/vnd.ms-playready.media.pyv
+.uvu	video/vnd.uvvu.mp4
+.viv	video/vnd.vivo

data/misc/port.json

@@ -0,0 +1 @@
+{ "format": "{number}", "rows": "port.tsv", "key": "number" }

data/misc/protocol.json

@@ -0,0 +1 @@
+{ "format": "{keyword}", "rows": "protocol.tsv", "key": "keyword", "name": "name" }

data/misc/protocol.tsv

@@ -0,0 +1,140 @@
+keyword	name	number
+HOPOPT	IPv6 Hop-by-Hop Option	0
+ICMP	Internet Control Message	1
+IGMP	Internet Group Management	2
+GGP	Gateway-to-Gateway	3
+IPv4	IPv4 encapsulation	4
+ST	Stream	5
+TCP	Transmission Control	6
+CBT	CBT	7
+EGP	Exterior Gateway Protocol	8
+IGP	any private interior gateway (used by Cisco for their IGRP)	9
+BBN-RCC-MON	BBN RCC Monitoring	10
+NVP-II	Network Voice Protocol	11
+PUP	PUP	12
+EMCON	EMCON	14
+XNET	Cross Net Debugger	15
+CHAOS	Chaos	16
+UDP	User Datagram	17
+MUX	Multiplexing	18
+DCN-MEAS	DCN Measurement Subsystems	19
+HMP	Host Monitoring	20
+PRM	Packet Radio Measurement	21
+XNS-IDP	XEROX NS IDP	22
+TRUNK-1	Trunk-1	23
+TRUNK-2	Trunk-2	24
+LEAF-1	Leaf-1	25
+LEAF-2	Leaf-2	26
+RDP	Reliable Data Protocol	27
+IRTP	Internet Reliable Transaction	28
+ISO-TP4	ISO Transport Protocol Class 4	29
+NETBLT	Bulk Data Transfer Protocol	30
+MFE-NSP	MFE Network Services Protocol	31
+MERIT-INP	MERIT Internodal Protocol	32
+DCCP	Datagram Congestion Control Protocol	33
+3PC	Third Party Connect Protocol	34
+IDPR	Inter-Domain Policy Routing Protocol	35
+XTP	XTP	36
+DDP	Datagram Delivery Protocol	37
+IDPR-CMTP	IDPR Control Message Transport Proto	38
+TP++	TP++ Transport Protocol	39
+IL	IL Transport Protocol	40
+IPv6	IPv6 encapsulation	41
+SDRP	Source Demand Routing Protocol	42
+IPv6-Route	Routing Header for IPv6	43
+IPv6-Frag	Fragment Header for IPv6	44
+IDRP	Inter-Domain Routing Protocol	45
+RSVP	Reservation Protocol	46
+GRE	Generic Routing Encapsulation	47
+DSR	Dynamic Source Routing Protocol	48
+BNA	BNA	49
+ESP	Encap Security Payload	50
+AH	Authentication Header	51
+I-NLSP	Integrated Net Layer Security TUBA	52
+NARP	NBMA Address Resolution Protocol	54
+Min-IPv4	Minimal IPv4 Encapsulation	55
+TLSP	Transport Layer Security Protocol using Kryptonet key management	56
+SKIP	SKIP	57
+IPv6-ICMP	ICMP for IPv6	58
+IPv6-NoNxt	No Next Header for IPv6	59
+IPv6-Opts	Destination Options for IPv6	60
+CFTP	CFTP	62
+SAT-EXPAK	SATNET and Backroom EXPAK	64
+KRYPTOLAN	Kryptolan	65
+RVD	MIT Remote Virtual Disk Protocol	66
+IPPC	Internet Pluribus Packet Core	67
+SAT-MON	SATNET Monitoring	69
+VISA	VISA Protocol	70
+IPCV	Internet Packet Core Utility	71
+CPNX	Computer Protocol Network Executive	72
+CPHB	Computer Protocol Heart Beat	73
+WSN	Wang Span Network	74
+PVP	Packet Video Protocol	75
+BR-SAT-MON	Backroom SATNET Monitoring	76
+SUN-ND	SUN ND PROTOCOL-Temporary	77
+WB-MON	WIDEBAND Monitoring	78
+WB-EXPAK	WIDEBAND EXPAK	79
+ISO-IP	ISO Internet Protocol	80
+VMTP	VMTP	81
+SECURE-VMTP	SECURE-VMTP	82
+VINES	VINES	83
+IPTM	Internet Protocol Traffic Manager	84
+NSFNET-IGP	NSFNET-IGP	85
+DGP	Dissimilar Gateway Protocol	86
+TCF	TCF	87
+EIGRP	EIGRP	88
+OSPFIGP	OSPFIGP	89
+Sprite-RPC	Sprite RPC Protocol	90
+LARP	Locus Address Resolution Protocol	91
+MTP	Multicast Transport Protocol	92
+AX.25	AX.25 Frames	93
+IPIP	IP-within-IP Encapsulation Protocol	94
+SCC-SP	Semaphore Communications Sec. Pro.	96
+ETHERIP	Ethernet-within-IP Encapsulation	97
+ENCAP	Encapsulation Header	98
+GMTP	GMTP	100
+IFMP	Ipsilon Flow Management Protocol	101
+PNNI	PNNI over IP	102
+PIM	Protocol Independent Multicast	103
+ARIS	ARIS	104
+SCPS	SCPS	105
+QNX	QNX	106
+A/N	Active Networks	107
+IPComp	IP Payload Compression Protocol	108
+SNP	Sitara Networks Protocol	109
+Compaq-Peer	Compaq Peer Protocol	110
+IPX-in-IP	IPX in IP	111
+VRRP	Virtual Router Redundancy Protocol	112
+PGM	PGM Reliable Transport Protocol	113
+L2TP	Layer Two Tunneling Protocol	115
+DDX	D-II Data Exchange (DDX)	116
+IATP	Interactive Agent Transfer Protocol	117
+STP	Schedule Transfer Protocol	118
+SRP	SpectraLink Radio Protocol	119
+UTI	UTI	120
+SMP	Simple Message Protocol	121
+PTP	Performance Transparency Protocol	123
+ISIS over IPv4	ISIS over IPv4	124
+FIRE	FIRE	125
+CRTP	Combat Radio Transport Protocol	126
+CRUDP	Combat Radio User Datagram	127
+SSCOPMCE	SSCOPMCE	128
+IPLT	IPLT	129
+SPS	Secure Packet Shield	130
+PIPE	Private IP Encapsulation within IP	131
+SCTP	Stream Control Transmission Protocol	132
+FC	Fibre Channel	133
+RSVP-E2E-IGNORE	RSVP-E2E-IGNORE	134
+Mobility Header	Mobility Header	135
+UDPLite	UDPLite	136
+MPLS-in-IP	MPLS-in-IP	137
+manet	MANET Protocols	138
+HIP	Host Identity Protocol	139
+Shim6	Shim6 Protocol	140
+WESP	Wrapped Encapsulating Security Payload	141
+ROHC	Robust Header Compression	142
+Ethernet	Ethernet	143
+AGGFRAG	AGGFRAG encapsulation payload for ESP	144
+NSH	Network Service Header	145
+Homa	Homa	146
+BIT-EMU	Bit-stream Emulation	147

data/misc/territory.json

@@ -0,0 +1 @@
+{ "format": "{name}", "rows": "territory.tsv", "key": "alpha2", "name": "name" }

data/misc/territory.tsv

@@ -0,0 +1,242 @@
+alpha2	alpha3	calling-code	capital	country	currency	flag	languages	name	numeric	tld
+AD	AND	376	Andorra la Vella	AD	EUR	🇦🇩	ca	Andorra	020	.ad
+AE	ARE	971	Abu Dhabi	AE	AED	🇦🇪	ar-AE,fa,en,hi,ur	United Arab Emirates	784	.ae
+AF	AFG	93	Kabul	AF	AFN	🇦🇫	fa-AF,ps,uz-AF,tk	Afghanistan	004	.af
+AG	ATG	1-268	St. John's	AG	XCD	🇦🇬	en-AG	Antigua & Barbuda	028	.ag
+AI	AIA	1-264	The Valley	GB	XCD	🇦🇮	en-AI	Anguilla	660	.ai
+AL	ALB	355	Tirana	AL	ALL	🇦🇱	sq,el	Albania	008	.al
+AM	ARM	374	Yerevan	AM	AMD	🇦🇲	hy	Armenia	051	.am
+AO	AGO	244	Luanda	AO	AOA	🇦🇴	pt-AO	Angola	024	.ao
+AR	ARG	54	Buenos Aires	AR	ARS	🇦🇷	es-AR,en,it,de,fr,gn	Argentina	032	.ar
+AS	ASM	1-684	Pago Pago	US	USD	🇦🇸	en-AS,sm,to	American Samoa	016	.as
+AT	AUT	43	Vienna	AT	EUR	🇦🇹	de-AT,hr,hu,sl	Austria	040	.at
+AU	AUS	61	Canberra	AU	AUD	🇦🇺	en-AU	Australia	036	.au
+AW	ABW	297	Oranjestad	NL	AWG	🇦🇼	nl-AW,pap,es,en	Aruba	533	.aw
+AX	ALA	358	Mariehamn	FI	EUR	🇦🇽	sv-AX	Åland Islands	248	.ax
+AZ	AZE	994	Baku	AZ	AZN	🇦🇿	az,ru,hy	Azerbaijan	031	.az
+BA	BIH	387	Sarajevo	BA	BAM	🇧🇦	bs,hr-BA,sr-BA	Bosnia	070	.ba
+BB	BRB	1-246	Bridgetown	BB	BBD	🇧🇧	en-BB	Barbados	052	.bb
+BD	BGD	880	Dhaka	BD	BDT	🇧🇩	bn-BD,en	Bangladesh	050	.bd
+BE	BEL	32	Brussels	BE	EUR	🇧🇪	nl-BE,fr-BE,de-BE	Belgium	056	.be
+BF	BFA	226	Ouagadougou	BF	XOF	🇧🇫	fr-BF,mos	Burkina Faso	854	.bf
+BG	BGR	359	Sofia	BG	EUR	🇧🇬	bg,tr-BG,rom	Bulgaria	100	.bg
+BH	BHR	973	Manama	BH	BHD	🇧🇭	ar-BH,en,fa,ur	Bahrain	048	.bh
+BI	BDI	257	Gitega	BI	BIF	🇧🇮	fr-BI,rn	Burundi	108	.bi
+BJ	BEN	229	Porto-Novo	BJ	XOF	🇧🇯	fr-BJ	Benin	204	.bj
+BL	BLM	590	Gustavia	FR	EUR	🇧🇱	fr	St. Barthélemy	652	.gp
+BM	BMU	1-441	Hamilton	GB	BMD	🇧🇲	en-BM,pt	Bermuda	060	.bm
+BN	BRN	673	Bandar Seri Begawan	BN	BND	🇧🇳	ms-BN,en-BN	Brunei	096	.bn
+BO	BOL	591	Sucre	BO	BOB	🇧🇴	es-BO,qu,ay	Bolivia	068	.bo
+BR	BRA	55	Brasilia	BR	BRL	🇧🇷	pt-BR,es,en,fr	Brazil	076	.br
+BS	BHS	1-242	Nassau	BS	BSD	🇧🇸	en-BS	Bahamas	044	.bs
+BT	BTN	975	Thimphu	BT	INR	🇧🇹	dz	Bhutan	064	.bt
+BW	BWA	267	Gaborone	BW	BWP	🇧🇼	en-BW,tn-BW	Botswana	072	.bw
+BY	BLR	375	Minsk	BY	BYN	🇧🇾	be,ru	Belarus	112	.by
+BZ	BLZ	501	Belmopan	BZ	BZD	🇧🇿	en-BZ,es	Belize	084	.bz
+CA	CAN	1	Ottawa	CA	CAD	🇨🇦	en-CA,fr-CA,iu	Canada	124	.ca
+CC	CCK	61	West Island	AU	AUD	🇨🇨	ms-CC,en	Cocos Islands	166	.cc
+CD	COD	243	Kinshasa	CD	CDF	🇨🇩	fr-CD,ln,ktu,kg,sw,lua	Congo - Kinshasa	180	.cd
+CF	CAF	236	Bangui	CF	XAF	🇨🇫	fr-CF,sg,ln,kg	Central African Republic	140	.cf
+CG	COG	242	Brazzaville	CG	XAF	🇨🇬	fr-CG,kg,ln-CG	Congo - Brazzaville	178	.cg
+CH	CHE	41	Bern	CH	CHF	🇨🇭	de-CH,fr-CH,it-CH,rm	Switzerland	756	.ch
+CI	CIV	225	Yamoussoukro	CI	XOF	🇨🇮	fr-CI	Côte d’Ivoire	384	.ci
+CK	COK	682	Avarua	NZ	NZD	🇨🇰	en-CK,mi	Cook Islands	184	.ck
+CL	CHL	56	Santiago	CL	CLP	🇨🇱	es-CL	Chile	152	.cl
+CM	CMR	237	Yaounde	CM	XAF	🇨🇲	en-CM,fr-CM	Cameroon	120	.cm
+CN	CHN	86	Beijing	CN	CNY	🇨🇳	zh-CN,yue,wuu,dta,ug,za	China	156	.cn
+CO	COL	57	Bogota	CO	COP	🇨🇴	es-CO	Colombia	170	.co
+CR	CRI	506	San Jose	CR	CRC	🇨🇷	es-CR,en	Costa Rica	188	.cr
+CU	CUB	53	Havana	CU	CUP	🇨🇺	es-CU,pap	Cuba	192	.cu
+CV	CPV	238	Praia	CV	CVE	🇨🇻	pt-CV	Cape Verde	132	.cv
+CW	CUW	599	 Willemstad	NL	XCG	🇨🇼	nl,pap	Curaçao	531	.cw
+CX	CXR	61	Flying Fish Cove	AU	AUD	🇨🇽	en,zh,ms-CX	Christmas Island	162	.cx
+CY	CYP	357	Nicosia	CY	EUR	🇨🇾	el-CY,tr-CY,en	Cyprus	196	.cy
+CZ	CZE	420	Prague	CZ	CZK	🇨🇿	cs,sk	Czechia	203	.cz
+DE	DEU	49	Berlin	DE	EUR	🇩🇪	de	Germany	276	.de
+DJ	DJI	253	Djibouti	DJ	DJF	🇩🇯	fr-DJ,ar,so-DJ,aa	Djibouti	262	.dj
+DK	DNK	45	Copenhagen	DK	DKK	🇩🇰	da-DK,en,fo,de-DK	Denmark	208	.dk
+DM	DMA	1-767	Roseau	DM	XCD	🇩🇲	en-DM	Dominica	212	.dm
+DO	DOM	1-809	Santo Domingo	DO	DOP	🇩🇴	es-DO	Dominican Republic	214	.do
+DZ	DZA	213	Algiers	DZ	DZD	🇩🇿	ar-DZ	Algeria	012	.dz
+EC	ECU	593	Quito	EC	USD	🇪🇨	es-EC	Ecuador	218	.ec
+EE	EST	372	Tallinn	EE	EUR	🇪🇪	et,ru	Estonia	233	.ee
+EG	EGY	20	Cairo	EG	EGP	🇪🇬	ar-EG,en,fr	Egypt	818	.eg
+EH	ESH	212	El-Aaiun	EH	MAD	🇪🇭	ar,mey	Western Sahara	732	.eh
+ER	ERI	291	Asmara	ER	ERN	🇪🇷	aa-ER,ar,tig,kun,ti-ER	Eritrea	232	.er
+ES	ESP	34	Madrid	ES	EUR	🇪🇸	es-ES,ca,gl,eu,oc	Spain	724	.es
+ET	ETH	251	Addis Ababa	ET	ETB	🇪🇹	am,en-ET,om-ET,ti-ET,so-ET,sid	Ethiopia	231	.et
+FI	FIN	358	Helsinki	FI	EUR	🇫🇮	fi-FI,sv-FI,smn	Finland	246	.fi
+FJ	FJI	679	Suva	FJ	FJD	🇫🇯	en-FJ,fj	Fiji	242	.fj
+FK	FLK	500	Stanley	GB	FKP	🇫🇰	en-FK	Falkland Islands	238	.fk
+FM	FSM	691	Palikir	FM	USD	🇫🇲	en-FM,chk,pon,yap,kos,uli,woe,nkr,kpg	Micronesia	583	.fm
+FO	FRO	298	Torshavn	DK	DKK	🇫🇴	fo,da-FO	Faroe Islands	234	.fo
+FR	FRA	33	Paris	FR	EUR	🇫🇷	fr-FR,frp,br,co,ca,eu,oc	France	250	.fr
+GA	GAB	241	Libreville	GA	XAF	🇬🇦	fr-GA	Gabon	266	.ga
+GB	GBR	44	London	GB	GBP	🇬🇧	en-GB,cy-GB,gd	UK	826	.uk
+GD	GRD	1-473	St. George's	GD	XCD	🇬🇩	en-GD	Grenada	308	.gd
+GE	GEO	995	Tbilisi	GE	GEL	🇬🇪	ka,ru,hy,az	Georgia	268	.ge
+GF	GUF	594	Cayenne	FR	EUR	🇬🇫	fr-GF	French Guiana	254	.gf
+GG	GGY	44	St Peter Port	GB	GBP	🇬🇬	en,nrf	Guernsey	831	.gg
+GH	GHA	233	Accra	GH	GHS	🇬🇭	en-GH,ak,ee,tw	Ghana	288	.gh
+GI	GIB	350	Gibraltar	GB	GIP	🇬🇮	en-GI,es,it,pt	Gibraltar	292	.gi
+GL	GRL	299	Nuuk	DK	DKK	🇬🇱	kl,da-GL,en	Greenland	304	.gl
+GM	GMB	220	Banjul	GM	GMD	🇬🇲	en-GM,mnk,wof,wo,ff	Gambia	270	.gm
+GN	GIN	224	Conakry	GN	GNF	🇬🇳	fr-GN	Guinea	324	.gn
+GP	GLP	590	Basse-Terre	FR	EUR	🇬🇵	fr-GP	Guadeloupe	312	.gp
+GQ	GNQ	240	Ciudad de la Paz	GQ	XAF	🇬🇶	es-GQ,fr,pt	Equatorial Guinea	226	.gq
+GR	GRC	30	Athens	GR	EUR	🇬🇷	el-GR,en,fr	Greece	300	.gr
+GT	GTM	502	Guatemala City	GT	GTQ	🇬🇹	es-GT	Guatemala	320	.gt
+GU	GUM	1-671	Hagatna	US	USD	🇬🇺	en-GU,ch-GU	Guam	316	.gu
+GW	GNB	245	Bissau	GW	XOF	🇬🇼	pt-GW,pov	Guinea-Bissau	624	.gw
+GY	GUY	592	Georgetown	GY	GYD	🇬🇾	en-GY	Guyana	328	.gy
+HK	HKG	852	Hong Kong	CN	HKD	🇭🇰	zh-HK,yue,zh,en	Hong Kong	344	.hk
+HN	HND	504	Tegucigalpa	HN	HNL	🇭🇳	es-HN,cab,miq	Honduras	340	.hn
+HR	HRV	385	Zagreb	HR	EUR	🇭🇷	hr-HR,sr	Croatia	191	.hr
+HT	HTI	509	Port-au-Prince	HT	HTG	🇭🇹	ht,fr-HT	Haiti	332	.ht
+HU	HUN	36	Budapest	HU	HUF	🇭🇺	hu-HU	Hungary	348	.hu
+ID	IDN	62	Jakarta	ID	IDR	🇮🇩	id,en,nl,jv	Indonesia	360	.id
+IE	IRL	353	Dublin	IE	EUR	🇮🇪	en-IE,ga-IE	Ireland	372	.ie
+IL	ISR	972	Jerusalem	IL	ILS	🇮🇱	he,ar-IL,en-IL	Israel	376	.il
+IM	IMN	44	Douglas	GB	GBP	🇮🇲	en,gv	Isle of Man	833	.im
+IN	IND	91	New Delhi	IN	INR	🇮🇳	en-IN,hi,bn,te,mr,ta,ur,gu,kn,ml,or,pa,as,bh,sat,ks,ne,sd,kok,doi,mni,sit,sa,fr,lus,inc	India	356	.in
+IO	IOT	246	Diego Garcia	GB	USD	🇮🇴	en-IO	British Indian Ocean Territory	086	.io
+IQ	IRQ	964	Baghdad	IQ	IQD	🇮🇶	ar-IQ,ku,hy	Iraq	368	.iq
+IR	IRN	98	Tehran	IR	IRR	🇮🇷	fa-IR,ku	Iran	364	.ir
+IS	ISL	354	Reykjavik	IS	ISK	🇮🇸	is,en,de,da,sv,no	Iceland	352	.is
+IT	ITA	39	Rome	IT	EUR	🇮🇹	it-IT,de-IT,fr-IT,sc,ca,co,sl	Italy	380	.it
+JE	JEY	44	Saint Helier	GB	GBP	🇯🇪	en,fr,nrf	Jersey	832	.je
+JM	JAM	1-876	Kingston	JM	JMD	🇯🇲	en-JM	Jamaica	388	.jm
+JO	JOR	962	Amman	JO	JOD	🇯🇴	ar-JO,en	Jordan	400	.jo
+JP	JPN	81	Tokyo	JP	JPY	🇯🇵	ja	Japan	392	.jp
+KE	KEN	254	Nairobi	KE	KES	🇰🇪	en-KE,sw-KE	Kenya	404	.ke
+KG	KGZ	996	Bishkek	KG	KGS	🇰🇬	ky,uz,ru	Kyrgyzstan	417	.kg
+KH	KHM	855	Phnom Penh	KH	KHR	🇰🇭	km,fr,en	Cambodia	116	.kh
+KI	KIR	686	Tarawa	KI	AUD	🇰🇮	en-KI,gil	Kiribati	296	.ki
+KM	COM	269	Moroni	KM	KMF	🇰🇲	ar,fr-KM	Comoros	174	.km
+KN	KNA	1-869	Basseterre	KN	XCD	🇰🇳	en-KN	St. Kitts & Nevis	659	.kn
+KP	PRK	850	Pyongyang	KP	KPW	🇰🇵	ko-KP	North Korea	408	.kp
+KR	KOR	82	Seoul	KR	KRW	🇰🇷	ko-KR,en	South Korea	410	.kr
+KW	KWT	965	Kuwait City	KW	KWD	🇰🇼	ar-KW,en	Kuwait	414	.kw
+KY	CYM	1-345	George Town	GB	KYD	🇰🇾	en-KY	Cayman Islands	136	.ky
+KZ	KAZ	7	Nur-Sultan	KZ	KZT	🇰🇿	kk,ru	Kazakhstan	398	.kz
+LA	LAO	856	Vientiane	LA	LAK	🇱🇦	lo,fr,en	Laos	418	.la
+LB	LBN	961	Beirut	LB	LBP	🇱🇧	ar-LB,fr-LB,en,hy	Lebanon	422	.lb
+LC	LCA	1-758	Castries	LC	XCD	🇱🇨	en-LC	St. Lucia	662	.lc
+LI	LIE	423	Vaduz	LI	CHF	🇱🇮	de-LI	Liechtenstein	438	.li
+LK	LKA	94	Colombo	LK	LKR	🇱🇰	si,ta,en	Sri Lanka	144	.lk
+LR	LBR	231	Monrovia	LR	LRD	🇱🇷	en-LR	Liberia	430	.lr
+LS	LSO	266	Maseru	LS	LSL	🇱🇸	en-LS,st,zu,xh	Lesotho	426	.ls
+LT	LTU	370	Vilnius	LT	EUR	🇱🇹	lt,ru,pl	Lithuania	440	.lt
+LU	LUX	352	Luxembourg	LU	EUR	🇱🇺	lb,de-LU,fr-LU	Luxembourg	442	.lu
+LV	LVA	371	Riga	LV	EUR	🇱🇻	lv,ru,lt	Latvia	428	.lv
+LY	LBY	218	Tripoli	LY	LYD	🇱🇾	ar-LY,it,en	Libya	434	.ly
+MA	MAR	212	Rabat	MA	MAD	🇲🇦	ar-MA,ber,fr	Morocco	504	.ma
+MC	MCO	377	Monaco	MC	EUR	🇲🇨	fr-MC,en,it	Monaco	492	.mc
+MD	MDA	373	Chisinau	MD	MDL	🇲🇩	ro,ru,gag,tr	Moldova	498	.md
+ME	MNE	382	Podgorica	ME	EUR	🇲🇪	sr,hu,bs,sq,hr,rom	Montenegro	499	.me
+MF	MAF	590	Marigot	FR	EUR	🇲🇫	fr	St. Martin	663	.gp
+MG	MDG	261	Antananarivo	MG	MGA	🇲🇬	fr-MG,mg	Madagascar	450	.mg
+MH	MHL	692	Majuro	MH	USD	🇲🇭	mh,en-MH	Marshall Islands	584	.mh
+MK	MKD	389	Skopje	MK	MKD	🇲🇰	mk,sq,tr,rmm,sr	North Macedonia	807	.mk
+ML	MLI	223	Bamako	ML	XOF	🇲🇱	fr-ML,bm	Mali	466	.ml
+MM	MMR	95	Nay Pyi Taw	MM	MMK	🇲🇲	my	Myanmar	104	.mm
+MN	MNG	976	Ulaanbaatar	MN	MNT	🇲🇳	mn,ru	Mongolia	496	.mn
+MO	MAC	853	Macao	CN	MOP	🇲🇴	zh,zh-MO,pt	Macao	446	.mo
+MP	MNP	1-670	Saipan	US	USD	🇲🇵	fil,tl,zh,ch-MP,en-MP	Northern Mariana Islands	580	.mp
+MQ	MTQ	596	Fort-de-France	FR	EUR	🇲🇶	fr-MQ	Martinique	474	.mq
+MR	MRT	222	Nouakchott	MR	MRU	🇲🇷	ar-MR,fuc,snk,fr,mey,wo	Mauritania	478	.mr
+MS	MSR	1-664	Plymouth	GB	XCD	🇲🇸	en-MS	Montserrat	500	.ms
+MT	MLT	356	Valletta	MT	EUR	🇲🇹	mt,en-MT	Malta	470	.mt
+MU	MUS	230	Port Louis	MU	MUR	🇲🇺	en-MU,bho,fr	Mauritius	480	.mu
+MV	MDV	960	Male	MV	MVR	🇲🇻	dv,en	Maldives	462	.mv
+MW	MWI	265	Lilongwe	MW	MWK	🇲🇼	ny,yao,tum,swk	Malawi	454	.mw
+MX	MEX	52	Mexico City	MX	MXN	🇲🇽	es-MX	Mexico	484	.mx
+MY	MYS	60	Kuala Lumpur	MY	MYR	🇲🇾	ms-MY,en,zh,ta,te,ml,pa,th	Malaysia	458	.my
+MZ	MOZ	258	Maputo	MZ	MZN	🇲🇿	pt-MZ,vmw	Mozambique	508	.mz
+NA	NAM	264	Windhoek	NA	NAD	🇳🇦	en-NA,af,de,hz,naq	Namibia	516	.na
+NC	NCL	687	Noumea	FR	XPF	🇳🇨	fr-NC	New Caledonia	540	.nc
+NE	NER	227	Niamey	NE	XOF	🇳🇪	fr-NE,ha,kr,dje	Niger	562	.ne
+NF	NFK	672	Kingston	AU	AUD	🇳🇫	en-NF	Norfolk Island	574	.nf
+NG	NGA	234	Abuja	NG	NGN	🇳🇬	en-NG,ha,yo,ig,ff	Nigeria	566	.ng
+NI	NIC	505	Managua	NI	NIO	🇳🇮	es-NI,en	Nicaragua	558	.ni
+NL	NLD	31	Amsterdam	NL	EUR	🇳🇱	nl-NL,fy-NL	Netherlands	528	.nl
+NO	NOR	47	Oslo	NO	NOK	🇳🇴	no,nb,nn,se,fi	Norway	578	.no
+NP	NPL	977	Kathmandu	NP	NPR	🇳🇵	ne,en	Nepal	524	.np
+NR	NRU	674	Yaren	NR	AUD	🇳🇷	na,en-NR	Nauru	520	.nr
+NU	NIU	683	Alofi	NZ	NZD	🇳🇺	niu,en-NU	Niue	570	.nu
+NZ	NZL	64	Wellington	NZ	NZD	🇳🇿	en-NZ,mi	New Zealand	554	.nz
+OM	OMN	968	Muscat	OM	OMR	🇴🇲	ar-OM,en,bal,ur	Oman	512	.om
+PA	PAN	507	Panama City	PA	PAB	🇵🇦	es-PA,en	Panama	591	.pa
+PE	PER	51	Lima	PE	PEN	🇵🇪	es-PE,qu,ay	Peru	604	.pe
+PF	PYF	689	Papeete	FR	XPF	🇵🇫	fr-PF,ty	French Polynesia	258	.pf
+PG	PNG	675	Port Moresby	PG	PGK	🇵🇬	en-PG,ho,meu,tpi	Papua New Guinea	598	.pg
+PH	PHL	63	Manila	PH	PHP	🇵🇭	tl,en-PH,fil,ceb,ilo,hil,war,pam,bik,bcl,pag,mrw,tsg,mdh,cbk,krj,sgd,msb,akl,ibg,yka,mta,abx	Philippines	608	.ph
+PK	PAK	92	Islamabad	PK	PKR	🇵🇰	ur-PK,en-PK,pa,sd,ps,brh	Pakistan	586	.pk
+PL	POL	48	Warsaw	PL	PLN	🇵🇱	pl	Poland	616	.pl
+PM	SPM	508	Saint-Pierre	FR	EUR	🇵🇲	fr-PM	St. Pierre & Miquelon	666	.pm
+PN	PCN	870	Adamstown	GB	NZD	🇵🇳	en-PN	Pitcairn	612	.pn
+PR	PRI	1	San Juan	US	USD	🇵🇷	en-PR,es-PR	Puerto Rico	630	.pr
+PT	PRT	351	Lisbon	PT	EUR	🇵🇹	pt-PT,mwl	Portugal	620	.pt
+PW	PLW	680	Melekeok	PW	USD	🇵🇼	pau,sov,en-PW,tox,ja,fil,zh	Palau	585	.pw
+PY	PRY	595	Asuncion	PY	PYG	🇵🇾	es-PY,gn	Paraguay	600	.py
+QA	QAT	974	Doha	QA	QAR	🇶🇦	ar-QA,es	Qatar	634	.qa
+RE	REU	262	Saint-Denis	FR	EUR	🇷🇪	fr-RE	Réunion	638	.re
+RO	ROU	40	Bucharest	RO	RON	🇷🇴	ro,hu,rom	Romania	642	.ro
+RS	SRB	381	Belgrade	RS	RSD	🇷🇸	sr,hu,bs,rom	Serbia	688	.rs
+RU	RUS	7	Moscow	RU	RUB	🇷🇺	ru,tt,xal,cau,ady,kv,ce,tyv,cv,udm,tut,mns,bua,myv,mdf,chm,ba,inh,kbd,krc,av,sah,nog	Russia	643	.ru
+RW	RWA	250	Kigali	RW	RWF	🇷🇼	rw,en-RW,fr-RW,sw	Rwanda	646	.rw
+SA	SAU	966	Riyadh	SA	SAR	🇸🇦	ar-SA	Saudi Arabia	682	.sa
+SB	SLB	677	Honiara	SB	SBD	🇸🇧	en-SB,tpi	Solomon Islands	090	.sb
+SC	SYC	248	Victoria	SC	SCR	🇸🇨	en-SC,fr-SC	Seychelles	690	.sc
+SD	SDN	249	Khartoum	SD	SDG	🇸🇩	ar-SD,en,fia	Sudan	729	.sd
+SE	SWE	46	Stockholm	SE	SEK	🇸🇪	sv-SE,se,sma,fi-SE	Sweden	752	.se
+SG	SGP	65	Singapore	SG	SGD	🇸🇬	cmn,en-SG,ms-SG,ta-SG,zh-SG	Singapore	702	.sg
+SH	SHN	290	Jamestown	GB	SHP	🇸🇭	en-SH	St. Helena	654	.sh
+SI	SVN	386	Ljubljana	SI	EUR	🇸🇮	sl,sh	Slovenia	705	.si
+SJ	SJM	47	Longyearbyen	NO	NOK	🇸🇯	no,ru	Svalbard & Jan Mayen	744	.sj
+SK	SVK	421	Bratislava	SK	EUR	🇸🇰	sk,hu	Slovakia	703	.sk
+SL	SLE	232	Freetown	SL	SLE	🇸🇱	en-SL,men,tem	Sierra Leone	694	.sl
+SM	SMR	378	San Marino	SM	EUR	🇸🇲	it-SM	San Marino	674	.sm
+SN	SEN	221	Dakar	SN	XOF	🇸🇳	fr-SN,wo,fuc,mnk	Senegal	686	.sn
+SO	SOM	252	Mogadishu	SO	SOS	🇸🇴	so-SO,ar-SO,it,en-SO	Somalia	706	.so
+SR	SUR	597	Paramaribo	SR	SRD	🇸🇷	nl-SR,en,srn,hns,jv	Suriname	740	.sr
+SS	SSD	211	Juba	SS	SSP	🇸🇸	en	South Sudan	728	.ss
+ST	STP	239	Sao Tome	ST	STN	🇸🇹	pt-ST	São Tomé & Príncipe	678	.st
+SV	SLV	503	San Salvador	SV	SVC	🇸🇻	es-SV	El Salvador	222	.sv
+SX	SXM	1-721	Philipsburg	NL	XCG	🇸🇽	nl,en	Sint Maarten	534	.sx
+SY	SYR	963	Damascus	SY	SYP	🇸🇾	ar-SY,ku,hy,arc,fr,en	Syria	760	.sy
+SZ	SWZ	268	Mbabane	SZ	SZL	🇸🇿	en-SZ,ss-SZ	Eswatini	748	.sz
+TC	TCA	1-649	Cockburn Town	GB	USD	🇹🇨	en-TC	Turks & Caicos Islands	796	.tc
+TD	TCD	235	N'Djamena	TD	XAF	🇹🇩	fr-TD,ar-TD,sre	Chad	148	.td
+TF	ATF	262	Port-aux-Francais	FR	EUR	🇹🇫	fr	French Southern Territories	260	.tf
+TG	TGO	228	Lome	TG	XOF	🇹🇬	fr-TG,ee,hna,kbp,dag,ha	Togo	768	.tg
+TH	THA	66	Bangkok	TH	THB	🇹🇭	th,en	Thailand	764	.th
+TJ	TJK	992	Dushanbe	TJ	TJS	🇹🇯	tg,ru	Tajikistan	762	.tj
+TL	TLS	670	Dili	TL	USD	🇹🇱	tet,pt-TL,id,en	Timor-Leste	626	.tl
+TM	TKM	993	Ashgabat	TM	TMT	🇹🇲	tk,ru,uz	Turkmenistan	795	.tm
+TN	TUN	216	Tunis	TN	TND	🇹🇳	ar-TN,fr	Tunisia	788	.tn
+TO	TON	676	Nuku'alofa	TO	TOP	🇹🇴	to,en-TO	Tonga	776	.to
+TR	TUR	90	Ankara	TR	TRY	🇹🇷	tr-TR,ku,diq,az,av	Türkiye	792	.tr
+TT	TTO	1-868	Port of Spain	TT	TTD	🇹🇹	en-TT,hns,fr,es,zh	Trinidad & Tobago	780	.tt
+TV	TUV	688	Funafuti	TV	AUD	🇹🇻	tvl,en,sm,gil	Tuvalu	798	.tv
+TW	TWN	886	Taipei	TW	TWD	🇹🇼	zh-TW,zh,nan,hak	Taiwan	158	.tw
+TZ	TZA	255	Dodoma	TZ	TZS	🇹🇿	sw-TZ,en,ar	Tanzania	834	.tz
+UA	UKR	380	Kyiv	UA	UAH	🇺🇦	uk,ru-UA,rom,pl,hu	Ukraine	804	.ua
+UG	UGA	256	Kampala	UG	UGX	🇺🇬	en-UG,lg,sw,ar	Uganda	800	.ug
+US	USA	1	Washington	US	USD	🇺🇸	en-US,es-US,haw,fr	US	840	.us
+UY	URY	598	Montevideo	UY	UYU	🇺🇾	es-UY	Uruguay	858	.uy
+UZ	UZB	998	Tashkent	UZ	UZS	🇺🇿	uz,ru,tg	Uzbekistan	860	.uz
+VA	VAT	39	Vatican City	VA	EUR	🇻🇦	la,it,fr	Vatican City	336	.va
+VC	VCT	1-784	Kingstown	VC	XCD	🇻🇨	en-VC,fr	St. Vincent & Grenadines	670	.vc
+VE	VEN	58	Caracas	VE	VES	🇻🇪	es-VE	Venezuela	862	.ve
+VG	VGB	1-284	Road Town	GB	USD	🇻🇬	en-VG	British Virgin Islands	092	.vg
+VI	VIR	1-340	Charlotte Amalie	US	USD	🇻🇮	en-VI	U.S. Virgin Islands	850	.vi
+VN	VNM	84	Hanoi	VN	VND	🇻🇳	vi,en,fr,zh,km	Vietnam	704	.vn
+VU	VUT	678	Port Vila	VU	VUV	🇻🇺	bi,en-VU,fr-VU	Vanuatu	548	.vu
+WF	WLF	681	Mata Utu	FR	XPF	🇼🇫	wls,fud,fr-WF	Wallis & Futuna	876	.wf
+WS	WSM	685	Apia	WS	WST	🇼🇸	sm,en-WS	Samoa	882	.ws
+YE	YEM	967	Sanaa	YE	YER	🇾🇪	ar-YE	Yemen	887	.ye
+YT	MYT	262	Mamoudzou	FR	EUR	🇾🇹	fr-YT	Mayotte	175	.yt
+ZA	ZAF	27	Pretoria	ZA	ZAR	🇿🇦	zu,xh,af,nso,en-ZA,tn,st,ts,ss,ve,nr	South Africa	710	.za
+ZM	ZMB	260	Lusaka	ZM	ZMW	🇿🇲	en-ZM,bem,loz,lun,lue,ny,toi	Zambia	894	.zm
+ZW	ZWE	263	Harare	ZW	ZWG	🇿🇼	en-ZW,sn,nr,nd	Zimbabwe	716	.zw

data/misc/timezone.json

@@ -1 +1 @@
-["UTC", "Africa/Cairo", "Africa/Johannesburg", "America/Chicago", "America/Denver", "America/Los_Angeles", "America/Mexico_City", "America/New_York", "America/Sao_Paulo", "America/Toronto", "Asia/Dubai", "Asia/Hong_Kong", "Asia/Kolkata", "Asia/Shanghai", "Asia/Singapore", "Asia/Tokyo", "Australia/Sydney", "Europe/Amsterdam", "Europe/Berlin", "Europe/London", "Europe/Madrid", "Europe/Moscow", "Europe/Oslo", "Europe/Paris", "Europe/Stockholm", "Europe/Warsaw", "Pacific/Auckland"]
+{ "format": "{zone}", "rows": "timezone.tsv", "key": "zone", "parent": "territory", "weight": "weight" }

data/misc/timezone.tsv

@@ -0,0 +1,402 @@
+offset	territory	weight	zone
++00:00	CI	17496811	Africa/Abidjan
++00:00	GH	11277813	Africa/Accra
++03:00	ET	18690487	Africa/Addis_Ababa
++01:00	DZ	20111539	Africa/Algiers
++03:00	ER	798575	Africa/Asmara
++00:00	ML	7673493	Africa/Bamako
++01:00	CF	2582076	Africa/Bangui
++00:00	GM	1305981	Africa/Banjul
++00:00	GW	904776	Africa/Bissau
++02:00	MW	3030905	Africa/Blantyre
++01:00	CG	3748808	Africa/Brazzaville
++02:00	BI	1251675	Africa/Bujumbura
++02:00	EG	47242427	Africa/Cairo
++00:00	MA	22808531	Africa/Casablanca
++01:00	ES	85985	Africa/Ceuta
++00:00	GN	6679241	Africa/Conakry
++00:00	SN	10055356	Africa/Dakar
++03:00	TZ	21350525	Africa/Dar_es_Salaam
++03:00	DJ	1316108	Africa/Djibouti
++01:00	CM	8330801	Africa/Douala
++00:00	EH	374107	Africa/El_Aaiun
++00:00	SL	1752243	Africa/Freetown
++02:00	BW	1262880	Africa/Gaborone
++02:00	ZW	4508424	Africa/Harare
++02:00	ZA	47939803	Africa/Johannesburg
++02:00	SS	1679674	Africa/Juba
++03:00	UG	9664318	Africa/Kampala
++02:00	SD	11781926	Africa/Khartoum
++02:00	RW	2910190	Africa/Kigali
++01:00	CD	20651243	Africa/Kinshasa
++01:00	NG	75725576	Africa/Lagos
++01:00	GA	1731692	Africa/Libreville
++00:00	TG	3089219	Africa/Lome
++01:00	AO	23567341	Africa/Luanda
++02:00	CD	16660579	Africa/Lubumbashi
++02:00	ZM	8422414	Africa/Lusaka
++01:00	GQ	355840	Africa/Malabo
++02:00	MZ	8598836	Africa/Maputo
++02:00	LS	739682	Africa/Maseru
++02:00	SZ	191312	Africa/Mbabane
++03:00	SO	6165052	Africa/Mogadishu
++00:00	LR	2127307	Africa/Monrovia
++03:00	KE	16920989	Africa/Nairobi
++01:00	TD	3363418	Africa/Ndjamena
++01:00	NE	3689982	Africa/Niamey
++00:00	MR	2157288	Africa/Nouakchott
++00:00	BF	6125427	Africa/Ouagadougou
++01:00	BJ	4001576	Africa/Porto-Novo
++00:00	ST	53300	Africa/Sao_Tome
++02:00	LY	6105792	Africa/Tripoli
++01:00	TN	7314557	Africa/Tunis
++02:00	NA	983097	Africa/Windhoek
+-10:00	US	15000	America/Adak
+-09:00	US	366200	America/Anchorage
+-04:00	AI	15000	America/Anguilla
+-04:00	AG	51737	America/Antigua
+-03:00	BR	858988	America/Araguaina
+-03:00	AR	11817784	America/Argentina/Buenos_Aires
+-03:00	AR	580969	America/Argentina/Catamarca
+-03:00	AR	8998034	America/Argentina/Cordoba
+-03:00	AR	413866	America/Argentina/Jujuy
+-03:00	AR	237745	America/Argentina/La_Rioja
+-03:00	AR	695484	America/Argentina/Mendoza
+-03:00	AR	241619	America/Argentina/Rio_Gallegos
+-03:00	AR	2120781	America/Argentina/Salta
+-03:00	AR	518624	America/Argentina/San_Juan
+-03:00	AR	266728	America/Argentina/San_Luis
+-03:00	AR	788916	America/Argentina/Tucuman
+-03:00	AR	109506	America/Argentina/Ushuaia
+-04:00	AW	90927	America/Aruba
+-03:00	PY	3653883	America/Asuncion
+-05:00	CA	15000	America/Atikokan
+-03:00	BR	13131907	America/Bahia
+-06:00	MX	36253	America/Bahia_Banderas
+-04:00	BB	98511	America/Barbados
+-03:00	BR	7977349	America/Belem
+-06:00	BZ	132559	America/Belize
+-04:00	CA	15000	America/Blanc-Sablon
+-04:00	BR	550434	America/Boa_Vista
+-05:00	CO	37243567	America/Bogota
+-07:00	US	748802	America/Boise
+-07:00	CA	15000	America/Cambridge_Bay
+-04:00	BR	2429053	America/Campo_Grande
+-05:00	MX	1328961	America/Cancun
+-04:00	VE	28324666	America/Caracas
+-03:00	GF	155245	America/Cayenne
+-05:00	KY	44705	America/Cayman
+-06:00	US	55510713	America/Chicago
+-06:00	MX	1533843	America/Chihuahua
+-07:00	MX	1512450	America/Ciudad_Juarez
+-06:00	CR	1458629	America/Costa_Rica
+-03:00	CL	62723	America/Coyhaique
+-07:00	CA	15000	America/Creston
+-04:00	BR	2960376	America/Cuiaba
+-04:00	CW	145838	America/Curacao
++00:00	GL	15000	America/Danmarkshavn
+-07:00	CA	15000	America/Dawson
+-07:00	CA	21465	America/Dawson_Creek
+-07:00	US	9069711	America/Denver
+-05:00	US	4344725	America/Detroit
+-04:00	DM	16571	America/Dominica
+-06:00	CA	3518059	America/Edmonton
+-05:00	BR	142900	America/Eirunepe
+-06:00	SV	2571611	America/El_Salvador
+-07:00	CA	15000	America/Fort_Nelson
+-03:00	BR	21824546	America/Fortaleza
+-04:00	CA	122883	America/Glace_Bay
+-04:00	CA	15000	America/Goose_Bay
+-05:00	TC	27489	America/Grand_Turk
+-04:00	GD	15000	America/Grenada
+-04:00	GP	250260	America/Guadeloupe
+-06:00	GT	7302205	America/Guatemala
+-05:00	EC	11564603	America/Guayaquil
+-04:00	GY	314746	America/Guyana
+-04:00	CA	843888	America/Halifax
+-05:00	CU	11972822	America/Havana
+-07:00	MX	2215048	America/Hermosillo
+-05:00	US	2672398	America/Indiana/Indianapolis
+-06:00	US	15000	America/Indiana/Knox
+-05:00	US	15000	America/Indiana/Marengo
+-05:00	US	15000	America/Indiana/Petersburg
+-06:00	US	15000	America/Indiana/Tell_City
+-05:00	US	15000	America/Indiana/Vevay
+-05:00	US	15451	America/Indiana/Vincennes
+-05:00	US	15000	America/Indiana/Winamac
+-06:00	CA	15000	America/Inuvik
+-05:00	CA	15000	America/Iqaluit
+-05:00	JM	2069735	America/Jamaica
+-09:00	US	31555	America/Juneau
+-05:00	US	909379	America/Kentucky/Louisville
+-05:00	US	15000	America/Kentucky/Monticello
+-04:00	BO	6958595	America/La_Paz
+-05:00	PE	21406340	America/Lima
+-08:00	US	46301242	America/Los_Angeles
+-04:00	SX	15000	America/Lower_Princes
+-03:00	BR	4805317	America/Maceio
+-06:00	NI	2621115	America/Managua
+-04:00	BR	3918958	America/Manaus
+-04:00	MF	15000	America/Marigot
+-04:00	MQ	243860	America/Martinique
+-06:00	MX	1899165	America/Matamoros
+-07:00	MX	2951216	America/Mazatlan
+-06:00	US	15000	America/Menominee
+-06:00	MX	4154924	America/Merida
+-09:00	US	15000	America/Metlakatla
+-06:00	MX	61603325	America/Mexico_City
+-03:00	PM	15000	America/Miquelon
+-04:00	CA	381536	America/Moncton
+-06:00	MX	9714766	America/Monterrey
+-03:00	UY	2457462	America/Montevideo
+-04:00	MS	15000	America/Montserrat
+-05:00	BS	319054	America/Nassau
+-05:00	US	88988896	America/New_York
+-09:00	US	15000	America/Nome
+-02:00	BR	15000	America/Noronha
+-06:00	US	15000	America/North_Dakota/Beulah
+-06:00	US	15000	America/North_Dakota/Center
+-06:00	US	21382	America/North_Dakota/New_Salem
+-02:00	GL	15000	America/Nuuk
+-06:00	MX	50784	America/Ojinaga
+-05:00	PA	1692726	America/Panama
+-03:00	SR	454064	America/Paramaribo
+-07:00	US	6541994	America/Phoenix
+-05:00	HT	5328623	America/Port-au-Prince
+-04:00	TT	456832	America/Port_of_Spain
+-04:00	BR	1551818	America/Porto_Velho
+-04:00	PR	1605968	America/Puerto_Rico
+-03:00	CL	135935	America/Punta_Arenas
+-06:00	CA	15000	America/Rankin_Inlet
+-03:00	BR	8832691	America/Recife
+-06:00	CA	982810	America/Regina
+-06:00	CA	15000	America/Resolute
+-05:00	BR	794527	America/Rio_Branco
+-03:00	BR	1403419	America/Santarem
+-04:00	CL	15912038	America/Santiago
+-04:00	DO	7521391	America/Santo_Domingo
+-03:00	BR	129186469	America/Sao_Paulo
+-02:00	GL	15000	America/Scoresbysund
+-09:00	US	15000	America/Sitka
+-04:00	BL	15000	America/St_Barthelemy
+-03:30	CA	202443	America/St_Johns
+-04:00	KN	15000	America/St_Kitts
+-04:00	LC	45210	America/St_Lucia
+-04:00	VI	70601	America/St_Thomas
+-04:00	VC	48723	America/St_Vincent
+-06:00	CA	16604	America/Swift_Current
+-06:00	HN	4808928	America/Tegucigalpa
+-04:00	GL	15000	America/Thule
+-08:00	MX	3797218	America/Tijuana
+-05:00	CA	24763654	America/Toronto
+-04:00	VG	15000	America/Tortola
+-07:00	CA	6340886	America/Vancouver
+-07:00	CA	28201	America/Whitehorse
+-06:00	CA	922566	America/Winnipeg
+-09:00	US	15000	America/Yakutat
++10:00	AU	15000	Antarctica/Macquarie
++01:00	SJ	15000	Arctic/Longyearbyen
++03:00	YE	6904010	Asia/Aden
++05:00	KZ	9055499	Asia/Almaty
++03:00	JO	4950859	Asia/Amman
++12:00	RU	15604	Asia/Anadyr
++05:00	KZ	282409	Asia/Aqtau
++05:00	KZ	598827	Asia/Aqtobe
++05:00	TM	2706037	Asia/Ashgabat
++05:00	KZ	388248	Asia/Atyrau
++03:00	IQ	29681856	Asia/Baghdad
++03:00	BH	815447	Asia/Bahrain
++04:00	AZ	5491302	Asia/Baku
++07:00	TH	49896297	Asia/Bangkok
++07:00	RU	1413356	Asia/Barnaul
++02:00	LB	4467748	Asia/Beirut
++06:00	KG	2233412	Asia/Bishkek
++08:00	BN	230416	Asia/Brunei
++09:00	RU	486311	Asia/Chita
++05:30	LK	4297175	Asia/Colombo
++03:00	SY	10939175	Asia/Damascus
++06:00	BD	35955225	Asia/Dhaka
++09:00	TL	311685	Asia/Dili
++04:00	AE	12931222	Asia/Dubai
++05:00	TJ	2578933	Asia/Dushanbe
++02:00	CY	109959	Asia/Famagusta
++07:00	VN	42129036	Asia/Ho_Chi_Minh
++08:00	HK	24678261	Asia/Hong_Kong
++07:00	MN	122174	Asia/Hovd
++08:00	RU	2231790	Asia/Irkutsk
++07:00	ID	68794489	Asia/Jakarta
++09:00	ID	2512221	Asia/Jayapura
++02:00	IL	7901151	Asia/Jerusalem
++04:30	AF	8764027	Asia/Kabul
++12:00	RU	247112	Asia/Kamchatka
++05:00	PK	88744953	Asia/Karachi
++05:45	NP	5913313	Asia/Kathmandu
++09:00	RU	15000	Asia/Khandyga
++05:30	IN	365373326	Asia/Kolkata
++07:00	RU	2615869	Asia/Krasnoyarsk
++08:00	MY	33322164	Asia/Kuala_Lumpur
++08:00	MY	4013140	Asia/Kuching
++03:00	KW	1662180	Asia/Kuwait
++08:00	MO	1236695	Asia/Macau
++11:00	RU	92782	Asia/Magadan
++08:00	ID	12458359	Asia/Makassar
++08:00	PH	64227523	Asia/Manila
++04:00	OM	3290021	Asia/Muscat
++02:00	CY	679420	Asia/Nicosia
++07:00	RU	2355475	Asia/Novokuznetsk
++07:00	RU	2119272	Asia/Novosibirsk
++06:00	RU	1243165	Asia/Omsk
++05:00	KZ	364799	Asia/Oral
++07:00	KH	4111780	Asia/Phnom_Penh
++07:00	ID	1819584	Asia/Pontianak
++09:00	KP	12876883	Asia/Pyongyang
++03:00	QA	1566088	Asia/Qatar
++05:00	KZ	529104	Asia/Qostanay
++05:00	KZ	511391	Asia/Qyzylorda
++03:00	SA	23913997	Asia/Riyadh
++11:00	RU	326982	Asia/Sakhalin
++05:00	UZ	3864475	Asia/Samarkand
++09:00	KR	47400555	Asia/Seoul
++08:00	CN	861293141	Asia/Shanghai
++08:00	SG	11621053	Asia/Singapore
++11:00	RU	15000	Asia/Srednekolymsk
++08:00	TW	26394699	Asia/Taipei
++05:00	UZ	7452125	Asia/Tashkent
++04:00	GE	1827911	Asia/Tbilisi
++03:30	IR	50648371	Asia/Tehran
++06:00	BT	170210	Asia/Thimphu
++09:00	JP	141330654	Asia/Tokyo
++07:00	RU	783576	Asia/Tomsk
++08:00	MN	1438738	Asia/Ulaanbaatar
++06:00	CN	12667986	Asia/Urumqi
++10:00	RU	15000	Asia/Ust-Nera
++07:00	LA	1561716	Asia/Vientiane
++10:00	RU	2534711	Asia/Vladivostok
++09:00	RU	902983	Asia/Yakutsk
++06:30	MM	18836338	Asia/Yangon
++05:00	RU	15407217	Asia/Yekaterinburg
++04:00	AM	2880620	Asia/Yerevan
+-01:00	PT	20056	Atlantic/Azores
+-04:00	BM	15000	Atlantic/Bermuda
++00:00	ES	2054431	Atlantic/Canary
+-01:00	CV	252678	Atlantic/Cape_Verde
++00:00	FO	15000	Atlantic/Faroe
++00:00	PT	213794	Atlantic/Madeira
++00:00	IS	245356	Atlantic/Reykjavik
++00:00	SH	15000	Atlantic/St_Helena
+-03:00	FK	15000	Atlantic/Stanley
++09:30	AU	1706678	Australia/Adelaide
++10:00	AU	5807246	Australia/Brisbane
++09:30	AU	17456	Australia/Broken_Hill
++09:30	AU	199509	Australia/Darwin
++08:45	AU	15000	Australia/Eucla
++10:00	AU	392450	Australia/Hobart
++10:00	AU	15000	Australia/Lindeman
++10:30	AU	15000	Australia/Lord_Howe
++10:00	AU	9304158	Australia/Melbourne
++08:00	AU	3086441	Australia/Perth
++10:00	AU	9352313	Australia/Sydney
++01:00	NL	13892513	Europe/Amsterdam
++01:00	AD	36283	Europe/Andorra
++04:00	RU	644812	Europe/Astrakhan
++02:00	GR	6286690	Europe/Athens
++01:00	RS	5756549	Europe/Belgrade
++01:00	DE	62740682	Europe/Berlin
++01:00	SK	2548401	Europe/Bratislava
++01:00	BE	9038306	Europe/Brussels
++02:00	RO	11070040	Europe/Bucharest
++01:00	HU	9366333	Europe/Budapest
++01:00	DE	15000	Europe/Busingen
++02:00	MD	1460347	Europe/Chisinau
++01:00	DK	3885672	Europe/Copenhagen
++01:00	IE	2720758	Europe/Dublin
++01:00	GI	26544	Europe/Gibraltar
++00:00	GG	16488	Europe/Guernsey
++02:00	FI	5265184	Europe/Helsinki
++00:00	IM	26218	Europe/Isle_of_Man
++03:00	TR	77554557	Europe/Istanbul
++00:00	JE	28000	Europe/Jersey
++02:00	RU	664545	Europe/Kaliningrad
++03:00	RU	782604	Europe/Kirov
++02:00	UA	35633582	Europe/Kyiv
++00:00	PT	6372176	Europe/Lisbon
++01:00	SI	552127	Europe/Ljubljana
++00:00	GB	57817786	Europe/London
++01:00	LU	131322	Europe/Luxembourg
++01:00	ES	42918247	Europe/Madrid
++01:00	MT	141714	Europe/Malta
++02:00	AX	15000	Europe/Mariehamn
++03:00	BY	6105410	Europe/Minsk
++01:00	MC	48977	Europe/Monaco
++03:00	RU	69342062	Europe/Moscow
++01:00	NO	3296711	Europe/Oslo
++01:00	FR	33093827	Europe/Paris
++01:00	ME	398494	Europe/Podgorica
++01:00	CZ	5924565	Europe/Prague
++02:00	LV	1206422	Europe/Riga
++01:00	IT	34033666	Europe/Rome
++04:00	RU	3546813	Europe/Samara
++01:00	SM	15000	Europe/San_Marino
++01:00	BA	1905798	Europe/Sarajevo
++04:00	RU	1705529	Europe/Saratov
++03:00	UA	1620906	Europe/Simferopol
++01:00	MK	1754303	Europe/Skopje
++02:00	BG	4111057	Europe/Sofia
++01:00	SE	7149541	Europe/Stockholm
++02:00	EE	1106943	Europe/Tallinn
++01:00	AL	1629643	Europe/Tirane
++04:00	RU	814189	Europe/Ulyanovsk
++01:00	LI	15000	Europe/Vaduz
++01:00	VA	15000	Europe/Vatican
++01:00	AT	5274040	Europe/Vienna
++02:00	LT	2103133	Europe/Vilnius
++03:00	RU	1837053	Europe/Volgograd
++01:00	PL	22720110	Europe/Warsaw
++01:00	HR	1650334	Europe/Zagreb
++01:00	CH	3461664	Europe/Zurich
++03:00	MG	6115943	Indian/Antananarivo
++06:00	IO	15000	Indian/Chagos
++07:00	CX	15000	Indian/Christmas
++06:30	CC	15000	Indian/Cocos
++03:00	KM	118353	Indian/Comoro
++05:00	TF	15000	Indian/Kerguelen
++04:00	SC	22881	Indian/Mahe
++05:00	MV	103693	Indian/Maldives
++04:00	MU	701493	Indian/Mauritius
++03:00	YT	138967	Indian/Mayotte
++04:00	RE	804329	Indian/Reunion
++13:00	WS	40407	Pacific/Apia
++12:00	NZ	5203989	Pacific/Auckland
++11:00	PG	40266	Pacific/Bougainville
++12:45	NZ	15000	Pacific/Chatham
++10:00	FM	15000	Pacific/Chuuk
+-06:00	CL	15000	Pacific/Easter
++11:00	VU	35901	Pacific/Efate
++12:00	FJ	334403	Pacific/Fiji
++12:00	TV	15000	Pacific/Funafuti
+-06:00	EC	15000	Pacific/Galapagos
+-09:00	PF	15000	Pacific/Gambier
++11:00	SB	194707	Pacific/Guadalcanal
++10:00	GU	121094	Pacific/Guam
+-10:00	US	1569453	Pacific/Honolulu
++13:00	KI	15000	Pacific/Kanton
++14:00	KI	15000	Pacific/Kiritimati
++11:00	FM	15000	Pacific/Kosrae
++12:00	MH	15000	Pacific/Kwajalein
++12:00	MH	45701	Pacific/Majuro
+-09:30	PF	15000	Pacific/Marquesas
++12:00	NR	15000	Pacific/Nauru
+-11:00	NU	15000	Pacific/Niue
++11:00	NF	15000	Pacific/Norfolk
++11:00	NC	137086	Pacific/Noumea
+-11:00	AS	15000	Pacific/Pago_Pago
++09:00	PW	15000	Pacific/Palau
+-08:00	PN	15000	Pacific/Pitcairn
++11:00	FM	15000	Pacific/Pohnpei
++10:00	PG	588589	Pacific/Port_Moresby
+-10:00	CK	15000	Pacific/Rarotonga
++10:00	MP	48220	Pacific/Saipan
+-10:00	PF	81958	Pacific/Tahiti
++12:00	KI	40311	Pacific/Tarawa
++13:00	TO	22400	Pacific/Tongatapu
++12:00	WF	15000	Pacific/Wallis

data/misc/tld.json

@@ -0,0 +1 @@
+{ "format": "{tld}", "rows": "tld.tsv", "key": "tld", "name": "unicode" }

data/misc/useragent.json

@@ -1 +1 @@
-["Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36", "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/123.0.0.0 Safari/537.36", "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36", "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:124.0) Gecko/20100101 Firefox/124.0", "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:123.0) Gecko/20100101 Firefox/123.0", "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.4 Safari/605.1.15", "Mozilla/5.0 (iPhone; CPU iPhone OS 17_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.4 Mobile/15E148 Safari/604.1", "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36 Edg/124.0.0.0"]
+{ "format": "{ua}", "rows": "useragent.tsv" }

data/misc/useragent.tsv

@@ -0,0 +1,97 @@
+browser	device	os	ua
+Chrome	mobile	Android	Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/138.0.0.0 Mobile Safari/537.36
+Chrome	mobile	Android	Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/148.0.0.0 Mobile Safari/537.36
+Chrome	mobile	Android	Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/150.0.0.0 Mobile Safari/537.36
+Chrome	mobile	Android	Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/151.0.0.0 Mobile Safari/537.36
+Chrome	mobile	Android	Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/152.0.0.0 Mobile Safari/537.36
+Chrome	mobile	Android	Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/152.0.0.0 Safari/537.36
+Chrome	mobile	Android	Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/153.0.0.0 Mobile Safari/537.36
+Chrome	mobile	Android	Mozilla/5.0 (Linux; Android 12; Pixel 6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/114.0.0.0 Mobile Safari/537.36
+Chrome	desktop	ChromeOS	Mozilla/5.0 (X11; CrOS x86_64 14541.0.0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/151.0.0.0 Safari/537.36
+Chrome	desktop	Linux	Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36
+Chrome	desktop	Linux	Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/148.0.0.0 Safari/537.36
+Chrome	desktop	Linux	Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/150.0.0.0 Safari/537.36
+Chrome	desktop	Linux	Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/151.0.0.0 Safari/537.36
+Chrome	desktop	Linux	Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/152.0.0.0 Safari/537.36
+Chrome	desktop	Linux	Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/153.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/117.0.5938.132 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/130.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/137.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/138.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/139.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/140.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/141.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/142.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/143.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/144.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/145.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/146.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/147.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/148.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/149.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/150.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/150.0.0.0 YaBrowser/26.8.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/151.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/151.0.7922.34 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/152.0.0.0 Safari/537.36
+Chrome	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/153.0.0.0 Safari/537.36
+Chrome	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 26_6_1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) CriOS/152.0.7977.64 Mobile/15E148 Safari/604.1
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/116.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/137.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/138.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/139.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/144.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/145.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/146.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/147.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/148.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/149.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/150.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/151.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/152.0.0.0 Safari/537.36
+Chrome	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/153.0.0.0 Safari/537.36
+Edge	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/151.0.0.0 Safari/537.36 Edg/151.0.0.0
+Edge	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/152.0.0.0 Safari/537.36 Edg/152.0.0.0
+Edge	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/153.0.0.0 Safari/537.36 Edg/153.0.0.0
+Edge	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/152.0.0.0 Safari/537.36 Edg/152.0.0.0
+Firefox	desktop	Linux	Mozilla/5.0 (X11; Linux x86_64; rv:155.0) Gecko/20100101 Firefox/155.0
+Firefox	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:126.0) Gecko/20100101 Firefox/126.0
+Firefox	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:154.0) Gecko/20100101 Firefox/154.0
+Firefox	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:155.0) Gecko/20100101 Firefox/155.0
+Firefox	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:154.0) Gecko/20100101 Firefox/154.0
+Firefox	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:155.0) Gecko/20100101 Firefox/155.0
+Opera	desktop	Windows	Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/151.0.0.0 Safari/537.36 OPR/135.0.0.0
+Opera	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/150.0.0.0 Safari/537.36 OPR/134.0.0.0
+Safari	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 18_5 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/18.5 Mobile/15E148 Safari/604.1
+Safari	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 18_7 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/18.7.5 Mobile/15E148 Safari/604.1
+Safari	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 18_7 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.2 Mobile/15E148 Safari/604.1
+Safari	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 18_7 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.3 Mobile/15E148 Safari/604.1
+Safari	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 18_7 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.5 Mobile/15E148 Safari/604.1
+Safari	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 18_7 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.5.2 Mobile/15E148 Safari/604.1
+Safari	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 18_7 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.6 Mobile/15E148 Safari/604.1
+Safari	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 18_7 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.6.1 Mobile/15E148 Safari/604.1
+Safari	mobile	iOS	Mozilla/5.0 (iPhone; CPU iPhone OS 18_7 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/27.0 Mobile/15E148 Safari/604.1
+Safari	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.4 Safari/605.1.15
+Safari	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/18.6 Safari/605.1.15
+Safari	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.2 Safari/605.1.15
+Safari	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.5 Safari/605.1.15
+Safari	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.5.2 Safari/605.1.15
+Safari	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.6 Safari/605.1.15
+Safari	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.6.1 Safari/605.1.15
+Safari	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/26.6.2 Safari/605.1.15
+Safari	desktop	macOS	Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/27.0 Safari/605.1.15
+Samsung Internet	mobile	Android	Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) SamsungBrowser/30.0 Chrome/143.0.0.0 Mobile Safari/537.36

data/sv_SE/birth-number.json

@@ -0,0 +1 @@
+{ "format": "{number}", "rows": "birth-number.tsv", "parent": "sex" }

data/sv_SE/birth-number.tsv

@@ -0,0 +1,3 @@
+sex	number
+f	238
+m	239

data/sv_SE/date.json

@@ -1,13 +1 @@
-{
+"{date(1970-01-01,2029-12-31,'2006-01-02')}"
-  "format": "{year}-{month}-{day}",
-  "month": ["01", "02", "03", "04", "05", "06", "07", "08", "09", "10", "11", "12"],
-  "day": ["01", "02", "03", "04", "05", "06", "07", "08", "09", "10", "11", "12", "13", "14", "15", "16", "17", "18", "19", "20", "21", "22", "23", "24", "25", "26", "27", "28"],
-  "year": [
-    "197{digits(1)}",
-    "198{digits(1)}",
-    { "format": "199{digits(1)}", "weight": 2 },
-    { "format": "200{digits(1)}", "weight": 3 },
-    { "format": "201{digits(1)}", "weight": 3 },
-    { "format": "202{digits(1)}", "weight": 2 }
-  ]
-}

data/sv_SE/first-name.json

@@ -0,0 +1 @@
+{ "format": "{name}", "rows": "first-name.tsv", "name": "name", "parent": "sex", "weight": "count" }

data/sv_SE/last-name.json

@@ -0,0 +1 @@
+{ "format": "{name}", "rows": "last-name.tsv", "key": "name", "weight": "count" }

data/sv_SE/person.json

@@ -1,32 +1,7 @@
 {
-  "format": "{prefix}{femalefirst|malefirst} {last}",
+  "format": "{prefix}{first} {last}",
-  "femalefirst": ["Agnes", "Alice", "Alicia", "Alma", "Amanda", "Anna", "Astrid", "Cornelia", "Ebba", "Elin", "Ella", "Ellen", "Elsa", "Emilia", "Emma", "Ester", "Eva", "Frida", "Hanna", "Hedda", "Ida", "Ingrid", "Isabelle", "Julia", "Karin", "Klara", "Kristina", "Lena", "Linnéa", "Lova", "Maja", "Maria", "Moa", "Molly", "Märta", "Nellie", "Nora", "Olivia", "Saga", "Sara", "Selma", "Signe", "Siri", "Sofia", "Stina", "Tilde", "Tuva", "Wilma", "Ylva", "Åsa"],
+  "first": "{.first-name.name}",
-  "malefirst": ["Adam", "Albin", "Alexander", "Alfred", "Anders", "Anton", "Arvid", "Axel", "Bengt", "Bo", "Carl", "David", "Edvin", "Elias", "Emil", "Erik", "Filip", "Folke", "Fredrik", "Gustav", "Göran", "Hampus", "Hans", "Henrik", "Hugo", "Isak", "Johan", "Jonas", "Karl", "Kjell", "Lars", "Leo", "Linus", "Love", "Magnus", "Mats", "Mattias", "Mikael", "Nils", "Olle", "Oskar", "Otto", "Patrik", "Per", "Pontus", "Rasmus", "Samuel", "Sixten", "Stefan", "Sten", "Sven", "Theodor", "Tobias", "Viktor", "William", "Åke"],
+  "last": "{.last-name.name}",
-  "last": [
+  "prefix": [null, { "format": "{.title.name} ", "weight": 0.05 }],
-    {
+  "sex": "{.sex.name}"
-      "format": "{first}sson",
-      "weight": 5,
-      "first": ["Ander", "Arvid", "Bengt", "Daniel", "David", "Erik", "Fredrik", "Gustaf", "Henrik", "Håkan", "Isak", "Jakob", "Johan", "Jon", "Jön", "Karl", "Lar", "Magnu", "Matt", "Mikael", "Mårten", "Nil", "Ol", "Per", "Petter", "Samuel", "Sven"]
-    },
-    {
-      "format": "{first}{last}",
-      "weight": 3.6,
-      "first": ["Ahl", "Berg", "Blom", "Ceder", "Dahl", "Ek", "Eng", "Falk", "Fors", "Gran", "Hag", "Hed", "Holm", "Lind", "Lund", "Mal", "Ny", "Rosen", "Sand", "Sjö", "Skog", "Söder", "Sten", "Strand", "Sund", "Wik", "Öst"],
-      "last": ["berg", "blad", "crona", "dahl", "ed", "fors", "gren", "holm", "in", "kvist", "löf", "lund", "man", "mark", "qvist", "roth", "stedt", "sten", "strand", "ström", "vall"]
-    },
-    {
-      "format": "{first}{last}",
-      "weight": 0.267,
-      "first": ["Hell", "Wall"],
-      "last": ["berg", "blad", "crona", "dahl", "ed", "fors", "gren", "holm", "in", "kvist", "man", "mark", "qvist", "roth", "stedt", "sten", "strand", "ström", "vall"]
-    },
-    {
-      "format": "{first}{last}",
-      "weight": 0.133,
-      "first": "Norr",
-      "last": ["berg", "blad", "crona", "dahl", "ed", "fors", "gren", "holm", "in", "kvist", "löf", "lund", "man", "mark", "qvist", "stedt", "sten", "strand", "ström", "vall"]
-    },
-    ["Berg", "Blom", "Eismar", "Falk", "Holm", "Lind", "Norberg", "Strand", "Ström", "von Flemming", "Åberg", "Öberg"]
-  ],
-  "prefix": ["", { "format": "{string} ", "string": ["dr", "prof"], "weight": 0.05 }]
 }

data/sv_SE/personnummer.json

@@ -0,0 +1 @@
+"{date(1930-01-01,2025-12-31,'060102')}-{.birth-number.number}{luhn()}"

data/sv_SE/samordningsnummer.json

@@ -0,0 +1 @@
+"{date(1930-01-01,2025-12-31,'0601')}{int(61,88)}-{.birth-number.number}{luhn()}"

data/sv_SE/sex.json

@@ -0,0 +1 @@
+{ "format": "{name}", "rows": "sex.tsv", "key": "code", "name": "name" }

data/sv_SE/sex.tsv

@@ -0,0 +1,3 @@
+code	name
+f	kvinna
+m	man

data/sv_SE/ssn.json

@@ -1,22 +0,0 @@
-{
-  "format": "{digits(2)}{mmdd}-{digits(3)}{luhn()}",
-  "mmdd": [
-    {
-      "format": "{m}{d}",
-      "weight": 7,
-      "m": ["01", "03", "05", "07", "08", "10", "12"],
-      "d": ["01", "02", "03", "04", "05", "06", "07", "08", "09", "10", "11", "12", "13", "14", "15", "16", "17", "18", "19", "20", "21", "22", "23", "24", "25", "26", "27", "28", "29", "30", "31"]
-    },
-    {
-      "format": "{m}{d}",
-      "weight": 4,
-      "m": ["04", "06", "09", "11"],
-      "d": ["01", "02", "03", "04", "05", "06", "07", "08", "09", "10", "11", "12", "13", "14", "15", "16", "17", "18", "19", "20", "21", "22", "23", "24", "25", "26", "27", "28", "29", "30"]
-    },
-    {
-      "format": "{m}{d}",
-      "m": "02",
-      "d": ["01", "02", "03", "04", "05", "06", "07", "08", "09", "10", "11", "12", "13", "14", "15", "16", "17", "18", "19", "20", "21", "22", "23", "24", "25", "26", "27", "28"]
-    }
-  ]
-}

data/sv_SE/time.json

@@ -1,17 +1 @@
-{
+"{time('15:04')}"
-  "format": "{hour}:{minute}{sec}",
-  "hour": [
-    { "format": "0{digits(1)}", "weight": 4 },
-    { "format": "1{digits(1)}", "weight": 4 },
-    { "format": "2{t}", "t": ["0", "1", "2", "3"], "weight": 2 }
-  ],
-  "minute": { "format": "{t}{digits(1)}", "t": ["0", "1", "2", "3", "4", "5"] },
-  "sec": [
-    "",
-    {
-      "format": ":{s}",
-      "s": { "format": "{t}{digits(1)}", "t": ["0", "1", "2", "3", "4", "5"] },
-      "weight": 0.4
-    }
-  ]
-}

data/sv_SE/title.json

@@ -0,0 +1 @@
+{ "format": "{name}", "rows": "title.tsv", "key": "name", "weight": "share" }

data/sv_SE/title.tsv

@@ -0,0 +1,3 @@
+name	share
+dr	3
+prof	1

data_test.go

@@ -10,7 +10,7 @@ import (
 
 // TestShippedDataCategories asserts every shipped locale emits only well-formed
 // values for each data category. en and sv patterns differ where the format is
-// locale-specific (date, time, ssn, company, price, ...); the rest are shared.
+// locale-specific (date, time, company, price, ...); the rest are shared.
 func TestShippedDataCategories(t *testing.T) {
 	letters := regexp.MustCompile(`^[\pL'-]+$`)
 	semver := regexp.MustCompile(`^v?\d+\.\d+\.\d+(-(alpha|beta|rc)\.\d+)?$`)
@@ -33,10 +33,7 @@ func TestShippedDataCategories(t *testing.T) {
 			regexp.MustCompile(`^\d{4}-(0[1-9]|1[0-2])-(0[1-9]|[12]\d|3[01])$`)},
 		{"time",
 			regexp.MustCompile(`^([1-9]|1[0-2]):[0-5]\d (AM|PM)$`),
-			regexp.MustCompile(`^([01]\d|2[0-3]):[0-5]\d(:[0-5]\d)?$`)},
+			regexp.MustCompile(`^([01]\d|2[0-3]):[0-5]\d$`)},
-		{"ssn",
-			regexp.MustCompile(`^[1-9]\d{2}-\d{2}-\d{4}$`),
-			regexp.MustCompile(`^\d{2}(0[1-9]|1[0-2])(0[1-9]|[12]\d|3[01])-\d{4}$`)},
 		{"version", semver, semver},
 		{"email", email, email},
 		{"ip", ip, ip},
@@ -72,7 +69,11 @@ func TestShippedMiscCategories(t *testing.T) {
 	v4 := regexp.MustCompile(`^[0-9a-f]{8}-[0-9a-f]{4}-4[0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$`)
 	mac := regexp.MustCompile(`^([0-9a-f]{2}:){5}[0-9a-f]{2}$`)
 	objectid := regexp.MustCompile(`^[0-9a-f]{24}$`)
+	datetime := regexp.MustCompile(`^\d{4}-\d\d-\d\dT\d\d:\d\d:\d\dZ$`)
 	for i := 0; i < 200; i++ {
+		if v := fake(t, f, "datetime"); !datetime.MatchString(v) {
+			t.Fatalf("misc datetime %q is not an RFC 3339 UTC instant", v)
+		}
 		if v := fake(t, f, "objectid"); !objectid.MatchString(v) {
 			t.Fatalf("misc objectid %q is not 24 hex chars", v)
 		}
@@ -97,19 +98,41 @@ func TestShippedMiscReferenceData(t *testing.T) {
 		"currency":              regexp.MustCompile(`^[A-Z]{3}$`),
 		"currency.name":         regexp.MustCompile(`\p{L}`),
 		"currency.symbol":       regexp.MustCompile(`^\S+$`),
-		"country":         regexp.MustCompile(`\p{L}`),
+		"territory":             regexp.MustCompile(`\p{L}`),
-		"country.alpha2":  regexp.MustCompile(`^[A-Z]{2}$`),
+		"territory.alpha2":      regexp.MustCompile(`^[A-Z]{2}$`),
-		"country.alpha3":  regexp.MustCompile(`^[A-Z]{3}$`),
+		"territory.alpha3":      regexp.MustCompile(`^[A-Z]{3}$`),
+		"territory.country":     regexp.MustCompile(`^[A-Z]{2}$`),
 		"language":              regexp.MustCompile(`\p{L}`),
 		"language.code":         regexp.MustCompile(`^[a-z]{2}$`),
-		"timezone":        regexp.MustCompile(`^([A-Za-z_]+(/[A-Za-z_]+)+|UTC)$`),
+		"language.code3":        regexp.MustCompile(`^[a-z]{3}$`),
+		"timezone":              regexp.MustCompile(`^[A-Za-z]+(/[A-Za-z0-9_+-]+)+$`),
+		"timezone.offset":       regexp.MustCompile(`^[+-](0\d|1[0-4]):[0-5]\d$`),
+		"timezone.territory":    regexp.MustCompile(`^[A-Z]{2}$`),
+		"timezone.weight":       regexp.MustCompile(`^[1-9]\d*$`),
 		"mimetype":              regexp.MustCompile(`^[a-z]+/[a-z0-9.+-]+$`),
-		"mimetype.ext":    regexp.MustCompile(`^\.[a-z0-9]+$`),
+		"mimetype.ext":          regexp.MustCompile(`^\.[a-z0-9_-]+$`),
+		"httpmethod":            regexp.MustCompile(`^(CONNECT|DELETE|GET|HEAD|OPTIONS|PATCH|POST|PUT|TRACE)$`),
+		"httpmethod.safe":       regexp.MustCompile(`^(true|false)$`),
+		"httpmethod.idempotent": regexp.MustCompile(`^(true|false)$`),
+		"port":                  regexp.MustCompile(`^[1-9]\d{0,4}$`),
+		"port.service":          regexp.MustCompile(`^\S+$`),
+		"protocol":              regexp.MustCompile(`^\S+( \S+)*$`),
+		"protocol.name":         regexp.MustCompile(`\S`),
+		"protocol.number":       regexp.MustCompile(`^\d{1,3}$`),
+		"loglevel":              regexp.MustCompile(`^(alert|crit|debug|emerg|err|info|notice|warning)$`),
+		"loglevel.code":         regexp.MustCompile(`^[0-7]$`),
+		"loglevel.severity":     regexp.MustCompile(`^(Alert|Critical|Debug|Emergency|Error|Informational|Notice|Warning)$`),
+		"tld":                   regexp.MustCompile(`^\.[a-z0-9]([a-z0-9-]*[a-z0-9])?$`),
+		"tld.type":              regexp.MustCompile(`^(country-code|generic|generic-restricted|infrastructure|sponsored)$`),
+		"tld.unicode":           regexp.MustCompile(`^\.\S+$`),
 		"httpstatus":            regexp.MustCompile(`^[1-5]\d{2} \S.*$`),
 		"httpstatus.code":       regexp.MustCompile(`^[1-5]\d{2}$`),
 		"useragent":             regexp.MustCompile(`^Mozilla/5\.0 .+`),
+		"useragent.browser":     regexp.MustCompile(`^(Chrome|Edge|Firefox|Opera|Safari|Samsung Internet)$`),
+		"useragent.device":      regexp.MustCompile(`^(desktop|mobile)$`),
+		"useragent.os":          regexp.MustCompile(`^(Android|ChromeOS|Linux|Windows|iOS|macOS)$`),
 		"car":                   regexp.MustCompile(`^\S.* \S`),
-		"car.maker":       regexp.MustCompile(`^\S`),
+		"car.make":              regexp.MustCompile(`^\S`),
 	}
 	for i := 0; i < 100; i++ {
 		for p, rx := range re {
@@ -147,30 +170,122 @@ func TestSwedishPersonNamesHaveNoTripleLetter(t *testing.T) {
 	}
 }
 
-// TestSwedishPersonnummer checks the two rules the shape regex can't: the date
+// TestSwedishPersonnummer checks what the shape regex can't: the date is a real
-// is a real calendar date (so month-length variants never emit e.g. Apr 31 or
+// calendar date, the birth number is Skatteverket's test series (238 female, 239
-// Feb 30) and the trailing digit is a valid Luhn checksum over the other nine.
+// male), the trailing digit is a Luhn checksum over the other nine, and a
+// samordningsnummer is the same with 60 added to the day.
 func TestSwedishPersonnummer(t *testing.T) {
 	sv := newGenerator(t, "data", WithSeed(1))
-	sawLongMonthEnd := false
+	re := regexp.MustCompile(`^\d{6}-23[89]\d$`)
+	sawLongMonthEnd, birth := false, map[string]bool{}
 	for i := 0; i < 2000; i++ {
-		v := fake(t, sv, "sv_SE.ssn")
+		v := fake(t, sv, "sv_SE.personnummer")
 		d := digitsOnly(v)
-		if len(d) != 10 {
+		if !re.MatchString(v) {
-			t.Fatalf("ssn %q has %d digits, want 10", v, len(d))
+			t.Fatalf("personnummer %q, want YYMMDD-238C or YYMMDD-239C", v)
 		}
-		if _, err := time.Parse("060102", d[:6]); err != nil { // 2-digit year, real-date check
+		if _, err := time.Parse("060102", d[:6]); err != nil {
-			t.Fatalf("ssn %q is not a valid calendar date: %v", v, err)
+			t.Fatalf("personnummer %q is not a valid calendar date: %v", v, err)
 		}
 		if !luhnValid(d) {
-			t.Fatalf("ssn %q fails the Luhn check", v)
+			t.Fatalf("personnummer %q fails the Luhn check", v)
 		}
-		if d[4:6] == "31" {
+		sawLongMonthEnd = sawLongMonthEnd || d[4:6] == "31"
-			sawLongMonthEnd = true
+		birth[d[6:9]] = true
+		s := fake(t, sv, "sv_SE.samordningsnummer")
+		sd := digitsOnly(s)
+		day, err := strconv.Atoi(sd[4:6])
+		if !re.MatchString(s) || err != nil || day < 61 || day > 88 || !luhnValid(sd) {
+			t.Fatalf("samordningsnummer %q, want YYMM(61-88)-23[89]C, Luhn-valid", s)
+		}
+		if _, err := time.Parse("0601", sd[:4]); err != nil {
+			t.Fatalf("samordningsnummer %q is not a valid year and month: %v", s, err)
 		}
 	}
 	if !sawLongMonthEnd {
-		t.Fatal("never generated a 31st — 31-day months are not reaching their last day")
+		t.Fatal("never generated a 31st")
+	}
+	if !birth["238"] || !birth["239"] {
+		t.Fatalf("birth numbers drawn %v, want both 238 and 239", birth)
+	}
+	for i := 0; i < 200; i++ {
+		got := fakeTemplate(t, sv, `{/sv_SE.person.sex} {/sv_SE.personnummer}`)
+		sex, id, _ := strings.Cut(got, " ")
+		if want := map[string]string{"kvinna": "238", "man": "239"}[sex]; want == "" || !strings.Contains(id, "-"+want) {
+			t.Fatalf("%q: a person and a personnummer in one render disagree on sex", got)
+		}
+	}
+}
+
+// TestShippedUSTaxIds pins the SSA and IRS ranges: an SSN's area is 001-899 but
+// 666, its group 01-99 and its serial 0001-9999; an ITIN is 9XX-GG-XXXX with GG
+// in 50-65, 70-88, 90-92 or 94-99.
+func TestShippedUSTaxIds(t *testing.T) {
+	f := newGenerator(t, "data", WithSeed(5))
+	ssn := regexp.MustCompile(`^(\d{3})-(\d{2})-(\d{4})$`)
+	itin := regexp.MustCompile(`^9\d{2}-(5\d|6[0-5]|7\d|8[0-8]|9[0-24-9])-\d{4}$`)
+	for i := 0; i < 2000; i++ {
+		v := fake(t, f, "en_US.ssn")
+		m := ssn.FindStringSubmatch(v)
+		if m == nil {
+			t.Fatalf("ssn %q, want AAA-GG-SSSS", v)
+		}
+		area, _ := strconv.Atoi(m[1])
+		group, _ := strconv.Atoi(m[2])
+		serial, _ := strconv.Atoi(m[3])
+		if area < 1 || area > 899 || area == 666 || group < 1 || serial < 1 {
+			t.Fatalf("ssn %q is in a range the SSA never assigns", v)
+		}
+		if v := fake(t, f, "en_US.itin"); !itin.MatchString(v) {
+			t.Fatalf("itin %q, want %s", v, itin)
+		}
+	}
+}
+
+// TestShippedPersonNames pins the name tables: a sex pins its first names, a name
+// both sexes carry appears under both, and the record's sex agrees with its name.
+func TestShippedPersonNames(t *testing.T) {
+	f := newGenerator(t, "data", WithSeed(9))
+	for path, want := range map[string]string{
+		"sv_SE.sex[f]":                        "kvinna",
+		"sv_SE.sex[man].code":                 "m",
+		"en_US.sex[f]":                        "female",
+		"sv_SE.sex[f].first-name[Anna]":       "Anna",
+		"sv_SE.sex[m].first-name[Erik].sex":   "m",
+		"en_US.sex[m].first-name[James]":      "James",
+		"en_US.sex[f].first-name[Taylor].sex": "f",
+		"en_US.sex[m].first-name[Taylor].sex": "m",
+		"sv_SE.last-name[Andersson]":          "Andersson",
+		"sv_SE.last-name[Andersson].count":    "",
+		"en_US.last-name[Smith]":              "Smith",
+	} {
+		got := fake(t, f, path)
+		if want == "" && !regexp.MustCompile(`^[1-9]\d*$`).MatchString(got) {
+			t.Errorf("Fake(%q) = %q, want a count", path, got)
+		} else if want != "" && got != want {
+			t.Errorf("Fake(%q) = %q, want %q", path, got, want)
+		}
+	}
+	if _, err := f.Fake("en_US.first-name[Taylor]"); err == nil || !strings.Contains(err.Error(), "sex[f].first-name[Taylor]") {
+		t.Fatalf("Fake(en_US.first-name[Taylor]) = %v, want both sexes' rows listed", err)
+	}
+	for _, locale := range []string{"sv_SE", "en_US"} {
+		seen := map[string]bool{}
+		for i := 0; i < 2000; i++ {
+			seen[fake(t, f, locale+".sex[f].first-name")] = true
+			first, sex := fake(t, f, locale+".person.first"), fake(t, f, locale+".person.sex")
+			if first == "" || sex == "" {
+				t.Fatalf("%s.person lacks a first name or a sex", locale)
+			}
+		}
+		if len(seen) < 300 || !seen["Anna"] && locale == "sv_SE" || !seen["Mary"] && locale == "en_US" {
+			t.Fatalf("%s female first names: %d distinct in 2000, want a weighted register", locale, len(seen))
+		}
+		got := fakeTemplate(t, f, `{/`+locale+`.sex.code}|{/`+locale+`.person.first}`)
+		code, first, _ := strings.Cut(got, "|")
+		if v := fake(t, f, locale+".sex["+code+"].first-name["+first+"]"); v != first {
+			t.Fatalf("%s: %q drawn as a %s name is not one", locale, first, code)
+		}
 	}
 }
 
@@ -300,3 +415,31 @@ func TestShippedStreetNumberFormats(t *testing.T) {
 		}
 	}
 }
+
+// TestShippedUSTitleAgreesWithSex pins that one person's title and sex agree, and
+// that the everyday titles are reachable.
+func TestShippedUSTitleAgreesWithSex(t *testing.T) {
+	f := newGenerator(t, "data", WithSeed(4))
+	female := map[string]bool{"Miss": true, "Mrs": true, "Ms": true}
+	seen := map[string]bool{}
+	for i := 0; i < 3000; i++ {
+		got := fakeTemplate(t, f, `{/en_US.person.prefix}|{/en_US.person.sex}`)
+		title, sex, _ := strings.Cut(got, "|")
+		title = strings.TrimSpace(title)
+		seen[title] = true
+		if title == "Mr" && sex != "male" || female[title] && sex != "female" {
+			t.Fatalf("%q: the title contradicts the sex", got)
+		}
+	}
+	for _, want := range []string{"", "Mr", "Ms"} {
+		if !seen[want] {
+			t.Errorf("en_US.person.prefix never drew %q in 3000 draws", want)
+		}
+	}
+	swedish := map[string]bool{"dr": true, "prof": true}
+	for i := 0; i < 50; i++ {
+		if got := fake(t, f, "sv_SE.title"); !swedish[got] {
+			t.Fatalf("sv_SE.title = %q, want an unsexed Swedish honorific", got)
+		}
+	}
+}

docs/research/part1-names-finance-company.md

@@ -0,0 +1,148 @@
+# Part 1 — Names, finance, company: sources and rules
+
+Researched 2026-09-17. "Embed" = ship as a data file in an MIT repo. Everything not marked UNVERIFIED was read from the cited page or file this session.
+
+## 1. US person names and identifiers
+
+### SSA baby names (given names by year and sex)
+- Page: https://www.ssa.gov/oact/babynames/limits.html — national zip https://www.ssa.gov/oact/babynames/names.zip (7 MB), state zip https://www.ssa.gov/oact/babynames/state/namesbystate.zip (23 MB), territories 228 kB.
+- Format (per SSA page + readme as mirrored by https://github.com/hackerb9/ssa-baby-names): one `yobYYYY.txt` per year 1880–2025, CSV `name,sex,count`, name 2–15 chars, sex `M`/`F`, sorted by sex then count desc. Names with < 5 occurrences in a year are excluded.
+- Counts: ~100,364 distinct names, ~2.02 M rows across all years (hackerb9 snapshot; year of snapshot UNVERIFIED). Top 1000 names cover 71.5 % of 2025 births (SSA page).
+- Licence: data.gov catalog entry https://catalog.data.gov/dataset/baby-names-from-social-security-card-applications-national-data lists licence **CC0 1.0** (`https://creativecommons.org/publicdomain/zero/1.0/`), publisher SSA. US federal work, no attribution required; embeddable.
+- Note: ssa.gov returns 403 to non-browser user agents (curl/WebFetch); download with a browser UA or via the data.gov link.
+
+### Census 2010 surnames
+- Page: https://www.census.gov/topics/population/genealogy/data/2010_surnames.html
+- Files: full set https://www2.census.gov/topics/genealogy/2010surnames/names.zip (CSV + XLSX, `Names_2010Census.csv`), top 1000 `Names_2010Census_Top1000.xlsx`, docs https://www2.census.gov/topics/genealogy/2010surnames/surnames.pdf.
+- Fields (from surnames.pdf): `name, rank, count, prop100k, cum_prop100k, pctwhite, pctblack, pctapi, pctaian, pct2prace, pcthispanic`. `(S)` = suppressed percentage. Names are UPPERCASE.
+- Count: 162,253 surnames occurring ≥ 100 times (covers 90.1 % of people with a recorded surname); the CSV also has a trailing `ALL OTHER NAMES` row (UNVERIFIED — www2.census.gov returned 403 to curl this session; download with a browser).
+- Licence: US Census Bureau work → US federal government work, public domain in the US (17 U.S.C. §105). Census only asks for citation (https://www.census.gov/about/policies/citation.html: "U.S. Census Bureau, [Table], [Product], [Vintage], [URL], accessed on [date]"). Embeddable; cite source in the data file header.
+
+### Name prefixes / suffixes
+- No authoritative enumerated list exists. USPS Publication 28 §33 (https://pe.usps.com/text/pub28/28c3_011.htm) defines the fields: Name Prefix, First Name, Middle Name or Initial, Surname, **Suffix Title** = "maturity (e.g., JR, SR) and professional (e.g., PHD, DDS) suffixes"; example "MR. WALTER W. WITHERSPOON JR.".
+- Recommendation: hand-curate. Prefixes `Mr., Mrs., Ms., Mx., Dr., Rev.`; suffixes `Jr., Sr., II, III, IV, PhD, MD, DDS, Esq.`. Rule of thumb: generational suffix (Jr/Sr/II–IV) only with a male name in traditional US usage; never combine Jr. with Sr.; at most one generational suffix.
+
+### SSN
+- Structure: `AAA-GG-SSSS` (area 3, group 2, serial 4). No check digit (Wikipedia).
+- Randomization since **2011-06-25** (https://www.ssa.gov/employer/randomization.html): "Previously unassigned area numbers were introduced for assignment excluding area numbers 000, 666 and 900-999." Area no longer geographic.
+- Invalid per SSA POMS RM 10201.035 (https://secure.ssa.gov/apps10/poms.nsf/lnx/0110201035): area `000`, `666`, `900–999`; group `00`; serial `0000`.
+- Generator rule: area ∈ [001,899] \ {666}; group ∈ [01,99]; serial ∈ [0001,9999].
+- Advertising SSNs (https://www.ssa.gov/history/ssn/misused.html): `078-05-1120` (Woolworth wallet card, 1938; > 40,000 people claimed it) and `219-09-9999` (1940 Social Security Board pamphlet). Both are structurally valid — exclude from output.
+- `987-65-4320`–`987-65-4329` "reserved for advertising": repeated by searchbug.com, ssn-check.org, a Missouri state standard (https://oa.mo.gov/sites/default/files/CC-SocialSecurityNumberNamingStandardV050305.pdf, which even misprints it as 987-54-4329). **UNVERIFIED** — not found on any ssa.gov page or POMS this session; Wikipedia no longer carries it. Safe to use these ten as the library's "obviously fake" examples, but do not cite SSA for it.
+
+### ITIN (IRS)
+- IRS Publication 4757 (Rev. 6-2026), p. 5 (https://www.irs.gov/pub/irs-pdf/p4757.pdf): "All valid ITINs are nine-digit numbers in the same format as the SSN (9XX-8X-XXXX), beginning with a "9" and the 4th and 5th digits ranging from 50 to 65, 70 to 88, 90 to 92, and 94 to 99."
+- Group `93` = ATIN (adoption TIN); `89` reserved (Intuit/TaxSlayer support pages, not IRS — UNVERIFIED).
+- Generator rule: `9` + 2 digits + group ∈ {50–65, 70–88, 90–92, 94–99} + 4 digits (serial `0000` not documented as invalid for ITIN; avoid it anyway).
+
+### EIN (IRS)
+- Format `NN-NNNNNNN`. Page: https://www.irs.gov/businesses/small-businesses-self-employed/how-eins-are-assigned-and-valid-ein-prefixes ("The first two digits of your EIN show the IRS campus that assigned it or if you applied online").
+- Valid prefixes (union of the table, 2026-09-17):
+  `01 02 03 04 05 06 10 11 12 13 14 15 16 20 21 22 23 24 25 26 27 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 71 72 73 74 75 76 77 80 81 82 83 84 85 86 87 88 90 91 92 93 94 95 98 99`
+  By campus: Andover 10,12 · Atlanta 60,67 · Austin 50,53 · Brookhaven 01–06,11,13,14,16,21–23,25,34,51,52,54–59,65 · Cincinnati 30,32,35–38,61 · Fresno 15,24 · Kansas City 40,44 · Memphis 94,95 · Ogden 80,90 · Philadelphia 33,39,41–43,46,48,62–64,66,68,71–77,85–88,91–93,98,99 · Internet 20,26,27,33,39,41,42,45–47,81–88,92,93,99 · SBA 31.
+- Not valid: 00, 07, 08, 09, 17, 18, 19, 28, 29, 49, 69, 70, 78, 79, 89, 96, 97. No check digit.
+
+## 2. Finance
+
+### ABA routing number (RTN)
+- Wikipedia https://en.wikipedia.org/wiki/ABA_routing_transit_number: 9 digits; check: `(3(d1+d4+d7) + 7(d2+d5+d8) + (d3+d6+d9)) mod 10 = 0`.
+- First two digits: `00` US Government; `01–12` Federal Reserve districts (normal banks); `21–32` thrifts (historic, still valid); `61–72` non-bank payment processors/clearinghouses; `80` traveler's checks. Other prefixes unused (Wikipedia does not say so explicitly — inferred).
+- Generator rule: prefix ∈ {01–12, 21–32, 61–72}, 6 random digits, compute 9th as `(10 − (3d1+7d2+d3+3d4+7d5+d6+3d7+7d8) mod 10) mod 10`.
+- Open list of real RTNs: the Federal Reserve E-Payments Routing Directory (https://www.frbservices.org/resources/routing-number-directory/index.html) says "may not be sold, re-licensed, or otherwise used for commercial gain" → **not embeddable**. Generate synthetically.
+
+### Payment card numbers (Wikipedia "Payment card number", https://en.wikipedia.org/wiki/Payment_card_number; all Luhn-checked)
+| Network | IIN ranges | Length | CVV |
+|---|---|---|---|
+| Visa | 4 | 13, 16, 19 | 3 |
+| Mastercard | 51–55, 2221–2720 | 16 | 3 |
+| American Express | 34, 37 | 15 | 4 (front) |
+| Discover | 6011, 644–649, 65, 622126–622925 (UnionPay co-brand) | 16–19 | 3 |
+| JCB | 3528–3589 | 16–19 | 3 |
+| Diners Club International | 30 (covers 300–305, 3095), 36, 38, 39 | 14–19 | 3 |
+| Diners Club US & Canada | 55 (Mastercard co-brand) | 16 | 3 |
+| UnionPay | 62 | 16–19 | 3 |
+| Maestro | 5018, 5020, 5038, 5893, 6304, 6759, 6761, 6762, 6763 | 12–19 | 3 |
+| Maestro UK | 6759, 676770, 676774 | 12–19 | 3 |
+| Dankort 5019 · Mir 2200–2204 · RuPay 60,65,81,82,508 · Troy 65,9792 · UATP 1 (15) · Verve 506099–506198, 507865–507964, 650002–650027 (16/18/19) | | | |
+- CVV lengths: https://en.wikipedia.org/wiki/Card_security_code ("three-digit ... Visa, Mastercard, and Discover"; "American Express is a four-digit code on the front").
+- The requested "Maestro 50/56–58/6xxx" is a looser legacy rule; use Wikipedia's explicit IINs.
+- Wikipedia text is CC BY-SA 4.0 — copy the *facts* (not prose) into your table; facts are not copyrightable.
+
+### Published test cards
+- Stripe: https://docs.stripe.com/testing — Visa 4242424242424242, Mastercard 5555555555554444, Amex 378282246310005, Discover 6011111111111117, Diners 3056930009020004 (14), JCB 3566002020360505, UnionPay 6200000000000005; any future expiry, any CVC (4 for Amex).
+- PayPal: https://developer.paypal.com/tools/sandbox/card-testing/ — Visa 4005519200000004, 4012000033330026, 4012000077777777, 4012888888881881, 4217651111111119, 4500600000000061, 4772129056533503, 4915805038587737; Mastercard 2223000048400011; Amex 371449635398431, 376680816376961; Diners 36461510000039, 36461510000013; Maestro 6304000000000000, 5063516945005047; JCB 3636500000000260, 3636500000000989; CUP 6200680000000004, 6200680000000038.
+- Reuse: neither page carries an open licence (Stripe docs are under Stripe's site terms). The numbers themselves are Luhn-valid facts, widely reproduced; embedding the numbers with a "from Stripe/PayPal test docs" note is low-risk but **UNVERIFIED** as an explicit grant. Alternative: generate Luhn-valid numbers from the IIN table above — no dependency on either vendor.
+
+### BIC / SWIFT (ISO 9362)
+- https://en.wikipedia.org/wiki/ISO_9362: 4 letters bank code + 2 letters ISO 3166-1 country + 2 alphanumeric location + optional 3 alphanumeric branch (8 or 11 chars; `XXX` = primary office). Location second char `0` = test BIC, `1` = passive participant, `2` = reverse billing.
+- Generator rule: `[A-Z]{4}[A-Z]{2}[A-Z2-9][A-NP-Z0-9]([A-Z0-9]{3})?` — avoid `0`/`1` as the 8th char for a "live" BIC.
+- SWIFT's BIC Directory is proprietary (SwiftRef licence, redistribution needs a separate "SwiftRef Redistribution License": https://www.swift.com/myswift/ordering/order-products-services/swiftref-redistribution-license) → **not embeddable**.
+- Open alternative: GLEIF BIC-to-LEI mapping, https://www.gleif.org/en/lei-data/lei-mapping/download-bic-to-lei-relationship-files, files at https://mapping.gleif.org/api/v2/bic-lei/ (monthly zip; `LEI-BIC-20260828.zip` → `lei-bic-20260828T000000.csv`, columns `LEI,BIC`, 39,347 rows). Licence: "BIC/LEI Mapping Table License Agreement" (https://www.gleif.org/lei-data/lei-mapping/download-bic-to-lei-relationship-files/2017-12-21_annex-2_bic-to-lei-mapping-table-license-agreement_final.pdf) — a CC0-style grant "for any purpose whatsoever, including ... commercial" but **conditional on reproducing the notice** "SWIFT © and database rights [month year]. All rights reserved. This Mapping Table has been developed by SWIFT. Any use of the Mapping Table ... is subject to the BIC/LEI Mapping Table License Agreement ... For the latest BIC information and updates, always refer to www.swift.com/bic." Embeddable in an MIT repo with that notice in the data file; it is a list of real live BICs (bank names not included — join to GLEIF LEI data if names are wanted).
+
+### ISIN (https://en.wikipedia.org/wiki/International_Securities_Identification_Number)
+- 12 chars: 2-letter country + 9 alphanumeric NSIN + 1 check digit.
+- Check: convert letters A=10…Z=35 (ASCII − 55) to produce a digit string, then Luhn over that string (double every second digit from the right, sum digits, check makes total ≡ 0 mod 10). Pitfall: work on the *expanded* digit string; a transposed letter pair can pass.
+
+### CUSIP (https://en.wikipedia.org/wiki/CUSIP)
+- 9 chars: 6 issuer + 2 issue + 1 check. Values: digits as-is, A=10…Z=35, `*`=36, `@`=37, `#`=38.
+- Check: for i in 1..8, v = value; if i even, v ×= 2; sum += v div 10 + v mod 10; check = (10 − sum mod 10) mod 10.
+- Real CUSIPs are proprietary (CGS/FactSet); generate synthetic ones only.
+
+### IBAN (ISO 13616)
+- Validation (https://en.wikipedia.org/wiki/International_Bank_Account_Number): move first 4 chars to end, letters → 10–35, integer mod 97 must equal 1; max 34 chars; check digits 02–98.
+- SWIFT registry: https://www.swift.com/standards/data-standards/iban-international-bank-account-number, PDF release 101 https://www.swift.com/sites/default/files/files/iban-registry-v101.pdf (release 100 was Oct 2025); TXT also published (URL **UNVERIFIED** — swift.com returns 403/errors to automation this session). Terms: free of charge; no explicit licence text found (**UNVERIFIED**). Country count: 89 countries per Wikipedia (Dec 2024).
+- Embeddable per-country length tables already extracted from the registry:
+  - php-iban `registry.txt` (https://github.com/globalcitizen/php-iban, **LGPL-3.0**): pipe-separated, 121 rows (116 official + unofficial), columns `country_code|country_name|domestic_example|bban_example|bban_format_swift|bban_format_regex|bban_length|iban_example|iban_format_swift|iban_format_regex|iban_length|bban_bankid_start_offset|...|country_sepa|swift_official|...|currency_iso4217|central_bank_url|central_bank_name|membership`. LGPL data in an MIT repo is awkward; use it as a *reference* to build your own table (formats are facts) rather than copying the file.
+  - ibankit-js (https://github.com/koblas/ibankit-js, **Apache-2.0**, registry v95) — same caveat.
+  - Best route: own table `{country, length, bban_format}` derived from the SWIFT PDF (facts), cite SWIFT.
+
+### Currencies (ISO 4217)
+- ISO says use is free: https://www.iso.org/iso-4217-currency-codes.html — "ISO allows free-of-charge use of its country, currency and language codes from ISO 3166, ISO 4217 and ISO 639". Official list from SIX: https://www.six-group.com/dam/download/financial-information/data-center/iso-currrency/lists/list-one.xml (+ `.xls`, `list-three` historic). Fields `CtryNm, CcyNm, Ccy, CcyNbr, CcyMnrUnts`; published 2026-01-01; 280 entity rows, 178 distinct codes. No symbols. No licence text inside the XML.
+- Alternatives:
+  - datasets/currency-codes (https://github.com/datasets/currency-codes): **PDDL** (public domain); `data/codes-all.csv` fields `Entity, Currency, AlphabeticCode, NumericCode, MinorUnit, WithdrawalDate`; built from SIX list one + three; no symbols. Embeddable.
+  - npm `currency-codes` v2.2.0 (https://github.com/freeall/currency-codes): **MIT**; fields `code, number, digits, currency, countries[]`; generated from the SIX XML; no symbols. Embeddable.
+  - umpirsky/currency-list (https://github.com/umpirsky/currency-list): **MIT**; code → localised name only (311 entries in `data/en_US/currency.json`, includes historic); no symbols, no numeric codes, no decimals.
+  - Debian iso-codes (https://salsa.debian.org/iso-codes-team/iso-codes): **LGPL-2.1+**; `data/iso_4217.json` has `alpha_3, name, numeric` only (179 entries); no symbols/minor units. LGPL → avoid embedding.
+  - Symbols: none of the above carry them. Unicode CLDR (`common/main/en.xml` currency symbols, Unicode License, MIT-compatible with notice) is the usual source — **UNVERIFIED this session**.
+
+### Crypto addresses
+- Bitcoin (https://en.bitcoin.it/wiki/List_of_address_prefixes): P2PKH version 0x00 → leading `1`; P2SH 0x05 → leading `3`; Base58Check 25–34 chars (alphabet excludes `0OIl`; last 4 bytes = first 4 of double-SHA256 of version+payload). Testnet `m`/`n`, `2`, `tb1`.
+- Bech32 (BIP-173, https://github.com/bitcoin/bips/blob/master/bip-0173.mediawiki, BSD-2-Clause): HRP `bc`, separator `1`, charset `qpzry9x8gf2tvdw0s3jn54khce6mua7l`, all-lowercase (or all-uppercase), max 90 chars; P2WPKH `bc1q…` = 42 chars, P2WSH `bc1q…` = 62 chars.
+- Bech32m (BIP-350, https://github.com/bitcoin/bips/blob/master/bip-0350.mediawiki): witness v1+ (P2TR `bc1p…`, 62 chars), checksum constant `0x2bc830a3` instead of 1.
+- Ethereum EIP-55 (https://eips.ethereum.org/EIPS/eip-55, CC0): `0x` + 40 hex; keccak256 of the lowercase hex (no 0x, as ASCII); for each hex letter, uppercase iff the corresponding hash nibble ≥ 8 (bit 4·i set). Digits unchanged.
+
+## 3. Company
+
+### NAICS 2022 (US Census)
+- Page https://www.census.gov/naics/ ; files: 6-digit list https://www.census.gov/naics/2022NAICS/6-digit_2022_Codes.xlsx (82 kB; **1,012** six-digit codes, 111110…928120, verified by parsing), 2–6 digit https://www.census.gov/naics/2022NAICS/2-6%20digit_2022_Codes.xlsx, structure https://www.census.gov/naics/2022NAICS/2022_NAICS_Structure.xlsx, descriptions https://www.census.gov/naics/2022NAICS/2022_NAICS_Descriptions.xlsx, manual PDF https://www.census.gov/naics/reference_files_tools/2022_NAICS_Manual.pdf.
+- Format: xlsx, column A code (numeric), column B title (trailing spaces in some titles — trim).
+- Licence: no statement on the page or in the manual front matter; US federal government work → public domain in the US (17 U.S.C. §105). Embeddable; cite Census/OMB.
+
+### SEC EDGAR company names
+- `https://www.sec.gov/files/company_tickers.json`: object keyed `"0".."N"` of `{cik_str, ticker, title}`; **10,422** entries (2026-09-17). `company_tickers_exchange.json`: `{"fields":["cik","name","ticker","exchange"],"data":[...]}`.
+- `https://www.sec.gov/Archives/edgar/cik-lookup-data.txt`: 40 MB, **1,059,372** lines, `NAME:CIK:` (`COMPANY NAME:0001234567:`), all filers incl. individuals — filter to the ticker list for company names.
+- Access: declare `User-Agent: Company Name email@domain`, ≤ 10 req/s (https://www.sec.gov/os/webmaster-faq). Licence: "All Government-created content on sec.gov and EDGAR public filing content are free to access and reuse" (same FAQ). Embeddable; the ticker file (~800 kB) is the practical one.
+
+### DUNS
+- https://en.wikipedia.org/wiki/Data_Universal_Numbering_System: 9 digits, no significance; shown `NN-NNN-NNNN` or plain. Had a mod-10 check digit until ~Dec 2006; **dropped** since (expanded the pool by 800 M). So: any 9 digits are structurally valid today; no validation rule to enforce. Optional DUNS+4 suffix (4 alphanumerics) is user-assigned.
+
+### Legal entity suffixes per country
+- GLEIF ISO 20275 Entity Legal Forms code list v1.6 (2026-02-19): page https://www.gleif.org/en/lei-data/code-lists/iso-20275-entity-legal-forms-code-list ; CSV https://www.gleif.org/lei-data/code-lists/iso-20275-entity-legal-forms-code-list/2026-02-19-elf-code-list-v1.6.csv (758 kB; xlsx alongside).
+- Columns: `ELF Code, Country of formation, Country Code (ISO 3166-1), Jurisdiction of formation, Country sub-division code (ISO 3166-2), Entity Legal Form name Local name, Language, Language Code (ISO 639-1), Entity Legal Form name Transliterated name, Abbreviations Local language, Abbreviations transliterated, Date created, ELF Status ACTV/INAC, Modification, Modification date, Reason`.
+- Counts: 4,002 rows; 130 countries; 3,790 ACTV / 212 INAC; 1,518 rows have a local abbreviation (e.g. US `N.A.`, `FSA`; SE `AB`… multiple abbreviations separated by `;`). US has 737 rows (per-state forms). Abbreviations are the field you want for `Inc`, `LLC`, `GmbH`, `AB`, `SAS`, `Pty Ltd`; many rows have none, so filter on non-empty abbreviation and status ACTV.
+- Licence: GLEIF Open Data page (https://www.gleif.org/en/about/open-data): "The data on GLEIF's website is provided under a Creative Commons (CC0) license"; LEI Data Terms of Use: "provided under the CC0 licence, see CC0 1.0 Universal". The ELF page itself names no licence — that CC0 explicitly covers this code list is **UNVERIFIED**, but GLEIF treats all its published data this way. Embeddable.
+
+## Quick embed verdict
+| Item | Embed? | Licence / attribution |
+|---|---|---|
+| SSA baby names | yes | CC0 (data.gov) |
+| Census 2010 surnames | yes | US gov PD; cite Census |
+| EIN prefixes, ITIN/SSN rules | yes (rules) | facts from IRS/SSA |
+| NAICS 2022 | yes | US gov PD |
+| SEC company_tickers.json | yes | free to reuse (SEC) |
+| GLEIF ELF code list | yes | CC0 (GLEIF) |
+| GLEIF BIC-LEI (real BICs) | yes, with mandatory SWIFT notice | custom permissive licence |
+| ISO 4217 (SIX XML / datasets PDDL / npm MIT) | yes | free use (ISO), PDDL, MIT |
+| Fed routing directory, SwiftRef BIC directory, CUSIP lists | no | restricted/proprietary — generate synthetically |
+| php-iban / ibankit registries | copy facts only | LGPL-3.0 / Apache-2.0 |
+| Stripe / PayPal test cards | numbers only, note source | no explicit grant (UNVERIFIED) |

docs/research/part2-internet-vehicle-airline.md

@@ -0,0 +1,118 @@
+# Open data sources: Internet, Vehicle, Airline (researched 2026-09-17)
+
+Verdict key: **EMBED** = safe to vendor into an MIT repo; **EMBED+ATTR** = ok with a notice; **AVOID** = licence incompatible or unclear.
+
+## 4. Internet
+
+### IANA registries (all of them)
+- Licence: IANA/IETF statement at https://www.iana.org/help/licensing-terms — "IANA and IETF intend that the Protocol Registries may be freely used by any party for any purpose", data dedicated under **CC0 1.0**. No attribution required. **EMBED**.
+- All registries below share this licence; ship one `SOURCES`/NOTICE line pointing at that URL.
+
+| Registry | CSV URL | Rows (2026-09-16) | Notes |
+|---|---|---|---|
+| Media types | `https://www.iana.org/assignments/media-types/<type>.csv`, type ∈ application, audio, font, haptics, image, message, model, multipart, text, video (also `example`) | application 1800, audio 165, font 6, haptics 3, image 88, message 27, model 42, multipart 17, text 105, video 97 (header excluded) ≈ **2350** | Fields: `Name,Template,Reference`. Template is the full `type/subtype`; some rows are DEPRECATED/OBSOLETED in Name (filter). Format: RFC 6838 §4.2 — `type "/" subtype`, restricted chars, max 127 chars each, case-insensitive. |
+| HTTP status codes | https://www.iana.org/assignments/http-status-codes/http-status-codes-1.csv | 75 rows, **64 assigned** (306 and 418 are "(Unused)"; 104 is TEMPORARY) → 62 usable | Fields: `Value,Description,Reference`. Value is 3 digits 1xx–5xx. |
+| IPv4 special-purpose | https://www.iana.org/assignments/iana-ipv4-special-registry/iana-ipv4-special-registry-1.csv | 26 blocks | Fields: `Address Block,Name,RFC,Allocation Date,Termination Date,Source,Destination,Forwardable,Globally Reachable,Reserved-by-Protocol`. |
+| IPv6 special-purpose | https://www.iana.org/assignments/iana-ipv6-special-registry/iana-ipv6-special-registry-1.csv | 27 blocks | Same fields. Contains `2001:db8::/32` and `3fff::/20`. |
+| Service names / ports | https://www.iana.org/assignments/service-names-port-numbers/service-names-port-numbers.csv | 14,534 rows; 11,731 with both name and a single port; **6,102 distinct named ports**; 6,008 tcp-named; 707 tcp names < 1024 | Fields: `Service Name,Port Number,Transport Protocol,Description,Assignee,Contact,Registration Date,Modification Date,Reference,Service Code,Unauthorized Use Reported,Assignment Notes`. Port Number may be empty, single, or a range `a-b`. Well-known = 0–1023, registered 1024–49151, dynamic 49152–65535 (RFC 6335). ~1.1 MB; embed only the named-port subset. |
+| TLDs | https://data.iana.org/TLD/tlds-alpha-by-domain.txt | **1,401** TLDs (first line is a `# Version …` comment) | Upper-case ASCII; IDNs as `XN--…` punycode. Same IANA CC0 terms (UNVERIFIED that data.iana.org is explicitly covered by the licensing page; it is the same registry data). Lower-case on use. |
+
+### MIME extension mapping: jshttp/mime-db
+- https://github.com/jshttp/mime-db — **MIT** (GitHub API spdx MIT; npm `license: MIT`), version 1.54.0 on npm.
+- `db.json` (CDN: https://cdn.jsdelivr.net/npm/mime-db/db.json): object keyed by lower-case type → `{source, extensions[], compressible, charset}`.
+- Counts (1.54.0): **2,522 types**, **1,015 with extensions**; source = iana 2,136, apache 275, nginx 13, custom 98.
+- Sources: IANA registry (CC0), Apache httpd `mime.types` (Apache-2.0), nginx `mime.types` (BSD-2). **EMBED+ATTR** (keep the MIT copyright notice). Data updates are not semver-breaking per README.
+
+### IPv4 / IPv6 generation rules (RFCs, no data file needed)
+- RFC 5737 documentation: `192.0.2.0/24` (TEST-NET-1), `198.51.100.0/24` (TEST-NET-2), `203.0.113.0/24` (TEST-NET-3). https://www.rfc-editor.org/rfc/rfc5737.html
+- RFC 1918 private: `10.0.0.0/8`, `172.16.0.0/12`, `192.168.0.0/16`. Also useful: `100.64.0.0/10` CGNAT (RFC 6598), `198.18.0.0/15` benchmarking (RFC 2544), `169.254.0.0/16` link-local.
+- IPv6 documentation: `2001:db8::/32` (RFC 3849) and `3fff::/20` (RFC 9637, "expands on the existing 2001:db8::/32 … with the reservation of an additional, larger prefix"). ULA `fc00::/7` (use `fd00::/8`), link-local `fe80::/10`.
+- Validation: generate inside those blocks; avoid network/broadcast (`.0`/`.255`) if the consumer expects host addresses. Format IPv6 per RFC 5952 (lower-case hex, `::` once, longest zero run).
+
+### MAC OUI (IEEE MA-L)
+- CSV: https://standards-oui.ieee.org/oui/oui.csv (3.8 MB, **40,161 rows**, fields `Registry,Assignment,Organization Name,Organization Address`; Assignment is 6 hex chars). Server returns HTTP 418 to non-browser user agents; fetch with a browser UA. Also MA-M `oui28.csv`, MA-S `oui36.csv`, CID `cid.csv`.
+- Licence: no licence on the IEEE pages themselves (site footer "© IEEE – All rights reserved" is generic). Debian's `ieee-data` copyright file (https://metadata.ftp-master.debian.org/changelogs/main/i/ieee-data/unstable_copyright) records IEEE's 2014 statement: *"IEEE does not assert any copyright in the OUI Public Listing or attempt to restrict distribution of the listing in any way. The IEEE Registration Authority does, however, strongly encourage those who use the list to obtain it directly from IEEE…"* — same text in the FSF directory. Debian ships it in `main` on that basis. **EMBED+ATTR** (cite the statement); the org-address column is not needed — keep `Assignment,Organization Name` only (~1 MB → few hundred KB). UNVERIFIED: no current IEEE page restates it (regauth FAQ and IPR pages do not mention it).
+- Alternative needing no data: locally administered unicast MAC — first octet's bit 1 (U/L) = 1, bit 0 (I/G) = 0 → second hex digit ∈ {2, 6, A, E}, e.g. `x2:xx:xx:xx:xx:xx`. Guaranteed never to collide with a vendor OUI. Notations: `01:23:45:67:89:ab`, `01-23-45-67-89-AB`, Cisco `0123.4567.89ab`.
+
+### User-agent strings
+| Source | Licence | Verdict |
+|---|---|---|
+| https://www.useragents.me/ | No licence stated anywhere on the page ("© 2022–2025 Useragents.me"); `/api` returns 404 now; JSON blobs embedded in page, weekly updates | **AVOID** (unlicensed = all rights reserved) |
+| https://explore.whatismybrowser.com/useragents/explore/ + API legal https://developers.whatismybrowser.com/api/about/legal/ | Proprietary: "You may NOT share the unique database download URL, or the downloaded file itself … EXCEPT for making it publicly available on the internet" | **AVOID** |
+| npm `user-agents` (intoli) https://github.com/intoli/user-agents | **BSD-2-Clause**, v2.1.185, daily-rebuilt dataset from Intoli traffic, weighted by frequency; fields `userAgent, platform, vendor, appName, deviceCategory, screenWidth/Height, viewportWidth/Height, connection{…}, oscpu, cpuClass, pluginsLength` | **EMBED+ATTR** (keep BSD notice). Count not stated on the page (UNVERIFIED; dataset is a few thousand rows). |
+| npm `top-user-agents` (microlinkhq) https://github.com/microlinkhq/top-user-agents | **MIT** (GitHub API). Top 100 UA strings from microlink.io traffic, weekly; JSON: https://cdn.jsdelivr.net/gh/microlinkhq/top-user-agents@master/src/index.json (+ desktop.json, mobile.json) | **EMBED+ATTR** — smallest clean option |
+- None of the above is CC0. Alternative: generate UAs from templates (Chrome/Firefox/Safari/Edge strings are fixed-form; only version numbers vary), no data needed.
+
+### Programming languages: GitHub Linguist
+- https://raw.githubusercontent.com/github-linguist/linguist/main/lib/linguist/languages.yml — **MIT** (repo LICENSE, "Copyright (c) 2017 GitHub, Inc."). **EMBED+ATTR**.
+- **835** top-level entries; **562** with `type: programming` (others: data, markup, prose). Per-entry fields: `type, color, extensions, aliases, tm_scope, ace_mode, language_id, group, interpreters, …`.
+
+### Operating systems
+- No open dataset found beyond awesome-lists. OS names/versions are facts — hand-author a small table (Windows 10/11, Windows Server, macOS versions+names, Ubuntu/Debian/Fedora/Arch/Alpine, iOS, Android, FreeBSD…). `user-agents` (BSD-2) carries `platform`/`oscpu` values if a data-driven list is wanted.
+
+### Database engines / collations
+- Engine names: hand-author (facts).
+- MySQL manual (https://dev.mysql.com/doc/refman/8.4/en/preface.html): Oracle terms — "you may not … copy, reproduce … distribute … any part" except distributing the docs with the software. **AVOID copying tables from MySQL docs.** Instead dump `SHOW COLLATION` from a `mysql:8.4.x` container (server output is factual, not the manual) — ~286 collations / 41 charsets in 8.x (UNVERIFIED count). PostgreSQL docs are under the PostgreSQL licence (permissive, keep notice) — `pg_collation` dump or docs both fine. MariaDB KB is CC BY-SA — avoid.
+
+## 5. Vehicle
+
+### NHTSA vPIC
+- API base `https://vpic.nhtsa.dot.gov/api/vehicles/`, `?format=json|xml|csv`. Endpoints verified: `GetAllMakes` (**12,363 makes**, fields `Make_ID,Make_Name`; 612 KB JSON, mostly small trailer/custom builders — filter with `GetMakesForVehicleType/car` → 195 makes, `/truck` → 207), `GetModelsForMake/honda` (362 models, `Make_ID,Make_Name,Model_ID,Model_Name`), `DecodeWMI/1HG`, `GetWMIsForManufacturer/honda` (45 WMIs, fields `WMI,Name,Country,VehicleType,Id,…`), `GetVehicleVariableList`, `GetVehicleVariableValuesList/<display name>` (must be the display name, URL-encoded, e.g. `Fuel%20Type%20-%20Primary`; the camel-case name returns 0 rows).
+- Value lists (counts): **Fuel Type - Primary 14**, **Body Class 71**, **Transmission Style 12**, **Drive Type 23**, **Vehicle Type 9**. Fields `Id,Name`.
+- Bulk: https://vpic.nhtsa.dot.gov/downloads/ — `vPICList_lite_2026_08.bak.zip` (MS SQL Server 2019+ backup, 190 MB, updated 2026-08-14), also `.plain.zip` / `.custom.zip`. Page says standalone DB "limited to VIN decoding"; makes/models/variables still via API. Contains `Wmi`, `Make`, `Model`, `WMIYearValidChars` tables (UNVERIFIED table names).
+- Licence: FAQ https://vpic.nhtsa.dot.gov/api/home/index/faq — "No, NHTSA is a government agency and the services provided on the API are free for use by the public as an offering as a part of our Open Data initiatives"; no rate-limit number stated, batch jobs asked to run at night. Work of the US federal government → public domain under 17 USC §105; data.gov entry lists licence as "unknown". **EMBED** (note source; no attribution legally required).
+
+### VIN (ISO 3779 / 49 CFR 565, North America)
+- 17 chars from `0-9 A-H J-N P R-Z` (**I, O, Q never**). Positions: 1–3 WMI, 4–8 VDS, 9 check digit, 10 model year, 11 plant, 12–17 serial (last 4 must be digits for US, positions 12–14 alphanumeric — 49 CFR 565.15).
+- Check digit: transliterate `A=1 B=2 C=3 D=4 E=5 F=6 G=7 H=8 J=1 K=2 L=3 M=4 N=5 P=7 R=9 S=2 T=3 U=4 V=5 W=6 X=7 Y=8 Z=9`, digits as-is; weights `8 7 6 5 4 3 2 10 0 9 8 7 6 5 4 3 2`; sum mod 11 → digit, remainder 10 → `X`.
+- Model year (pos 10): excludes `I O Q U Z 0`; `A=2010 … H=2017 J=2018 K=2019 L=2020 M=2021 N=2022 P=2023 R=2024 S=2025 T=2026 V=2027 W=2028 X=2029 Y=2030`, `1=2031…9=2039` (30-year cycle; `1–9` also = 2001–2009). Source: https://en.wikipedia.org/wiki/Vehicle_identification_number (rules are facts; do not copy prose).
+
+### WMI list
+- SAE J1044 / SAE WMI database is paid (https://www.sae.org/standards/j1044_202501-world-manufacturer-identifier) — **AVOID**.
+- Wikipedia "List of WMIs" — CC BY-SA — **AVOID**.
+- Use vPIC: `GetWMIsForManufacturer/<name or id>` per manufacturer, or the `Wmi` table in the standalone `.bak`. Public domain. Country prefix rule (first char): `1,4,5` USA, `2` Canada, `3` Mexico, `J` Japan, `K` Korea, `S` UK, `W` Germany, `Y` Sweden/Finland, `Z` Italy, `L` China (ISO 3780 facts).
+
+### US licence plate formats
+| Source | Licence | Verdict |
+|---|---|---|
+| openalpr `runtime_data/postprocess/us.patterns` (https://github.com/openalpr/openalpr) | **AGPL-3.0** | **AVOID** (291 lines, `@`=letter `#`=digit — useful only as a reference to check hand-authored patterns) |
+| Wikipedia "United States license plate designs and serial formats" | CC BY-SA | **AVOID copying**; formats themselves are facts |
+| jonnii/platekit | MIT, but SVG rendering, no serial patterns | not useful |
+- No MIT/CC0 per-state pattern list found. Recommendation: hand-author a per-state table (facts); current standard passenger formats as listed by Wikipedia (verify against DMV pages before use): AL `0AXXXXX`/`00AXXXX`, AK `ABC 123`, AZ `XXX 1XX`, AR `ABC 12D`, CA `1ABC123` (digit-3 letters-3 digits), CO `ABC-D12`, CT `AB·12345`, DE `123456`, DC `AB-1234`, FL `ABC D12`, GA `ABC1234`, HI `ABC 123`, ID `A 1234U`, IL `AB 12345`, IN `123ABC`, IA `ABC 123`, KS `1234ABC`, KY `ABC123`, LA `123 ABC`, ME `123·ABC`, MD `1AB2345`, MA `1ABC 23`, MI `ABC 1234`, MN `ABC-123`, MS `ABC 123`, MO `AB1 C2D`, MT `0-AB1234`, NE `ABC 123`, NV `123·A45`, NH `123 4567`, NJ `D12-ABC`, NM `123-ABC`, NY `ABC-1234`, NC `ABC-1234`, ND `123 ABC`, OH `ABC 1234`, OK `ABC-123`, OR `123 ABC`, PA `ABC1234`, RI `1AB 234`, SC `123ABC`, SD `0A1 234`, TN `ABC 1234`, TX `ABC-1234`, UT `A12 3BC`, VT `ABC 123`, VA `ABC-1234`, WA `ABC1234`, WV `X1A 2345`, WI `ABC-1234`, WY `1A-123A`. Most states omit I/O/Q from serials (state-specific, UNVERIFIED per state).
+
+## 6. Airline
+
+### OurAirports
+- https://ourairports.com/data/ — "All data is released to the Public Domain"; mirror repo https://github.com/davidmegginson/ourairports-data is **Unlicense** (GitHub API). Nightly updates. **EMBED**.
+- Raw URLs: `https://davidmegginson.github.io/ourairports-data/{airports,airport-frequencies,runways,navaids,countries,regions}.csv`.
+
+| File | Rows | Fields |
+|---|---|---|
+| airports.csv (12.7 MB) | **86,089**; by type: small_airport 42,734, heliport 23,216, closed 13,524, medium 4,106, seaplane_base 1,273, large 1,174, balloonport 62 | `id,ident,type,name,latitude_deg,longitude_deg,elevation_ft,continent,iso_country,iso_region,municipality,scheduled_service,icao_code,iata_code,gps_code,local_code,home_link,wikipedia_link,keywords` |
+| — with `iata_code` | **9,055** (large 1,171, medium 3,398, small 4,233, seaplane 153, heliport 100); `scheduled_service=yes` 4,335; US with IATA 2,036 (873 large/medium) | Suggested subset: `iata_code != '' AND type IN (large_airport, medium_airport)` → 4,569 rows |
+| countries.csv | **249** | `id,code,name,continent,wikipedia_link,keywords` |
+| regions.csv | **3,987** | `id,code,local_code,name,continent,iso_country,wikipedia_link,keywords` (code = ISO 3166-2, e.g. `US-KS`) |
+- No airline file. Derived: datasets/airport-codes on datahub (PDDL) — no need, use upstream.
+
+### OpenFlights
+- https://openflights.org/data.php — airports/airlines/routes/planes under **ODbL 1.0** (+ DbCL); airline/plane data partly from Wikipedia (GFDL/CC BY-SA). 5,888 airlines (2012). **AVOID** (share-alike). npm `airline-codes` (npow, ISC) is a straight OpenFlights sync → inherits ODbL → **AVOID**.
+
+### Airline IATA/ICAO codes — open options
+| Source | Licence | Count | Verdict |
+|---|---|---|---|
+| Wikidata SPARQL (https://query.wikidata.org/sparql) | **CC0** (https://www.wikidata.org/wiki/Wikidata:Licensing) | **2,928** items with IATA code (P229), **3,644** with ICAO code (P230) | **EMBED**. Query: `SELECT ?item ?itemLabel ?iata ?icao ?callsign WHERE { ?item wdt:P229 ?iata . OPTIONAL { ?item wdt:P230 ?icao } OPTIONAL { ?item wdt:P432 ?callsign } SERVICE wikibase:label { bd:serviceParam wikibase:language "en" } }`. Includes defunct airlines — filter `MINUS { ?item wdt:P576 ?dissolved }` and/or P31 = Q46970 (airline). |
+| OpenTravelData `optd_airlines.csv` (https://github.com/opentraveldata/opentraveldata) | **CC BY 4.0** | 1,620 rows; `^`-separated; fields `pk^env_id^validity_from^validity_to^3char_code^2char_code^num_code^name^name2^alliance_code^alliance_status^type^wiki_link^flt_freq^alt_names^bases^key^version^parent_pk_list^successor_pk_list` | **EMBED+ATTR** (attribution notice required) |
+- Format rules: IATA designator = 2 alphanumeric chars (`[A-Z0-9]{2}`, at least one letter in practice; a third optional char exists in the standard but is unused); "controlled duplicates" share a code between non-overlapping regional carriers. ICAO designator = 3 letters `[A-Z]{3}`, unique. Accounting/prefix code = 3 digits (ticket number prefix, e.g. `016` United).
+
+### Aircraft types
+| Source | Licence | Verdict |
+|---|---|---|
+| ICAO Doc 8643 (https://www.icao.int/operational-safety/doc-8643-aircraft-type-designators) | ICAO copyright notice: "None of the materials … may be used, reproduced or transmitted … without permission in writing from ICAO"; API Data Service is paid (25 free trial calls) | **AVOID** — including GitHub mirrors such as ColtJD45/icao-aircraft-designator-list (MIT-labelled, 7,388 rows, fields `manufacturer,model,type_designator,description,engine_type,engine_count,wtc`) because the underlying data is scraped ICAO content (UNVERIFIED provenance) |
+| OpenSky aircraft database (https://opensky-network.org/data/aircraft; CSV https://s3.opensky-network.org/data-samples/metadata/aircraftDatabase.csv) | "The aircraft database is unlicensed and does not fall under our terms of use … offered as is"; built from registries, openflights.org and ICAO Doc 8643; citation requested; page says it is no longer up to date | **AVOID** ("unlicensed" ≠ open; upstream includes ODbL/ICAO). Fields for reference: `icao24,registration,manufacturericao,manufacturername,model,typecode,serialnumber,linenumber,icaoaircrafttype,operator,operatorcallsign,operatoricao,operatoriata,owner,…,engines,…,categoryDescription` |
+| Wikidata | **CC0** | **550** items with ICAO aircraft type designator (P8305); also IATA aircraft code P9040 (UNVERIFIED property id), manufacturer P176 | **EMBED**. Query: `SELECT ?item ?itemLabel ?icao ?mfrLabel WHERE { ?item wdt:P8305 ?icao . OPTIONAL { ?item wdt:P176 ?mfr } SERVICE wikibase:label { bd:serviceParam wikibase:language "en" } }` |
+- Format: ICAO type designator 2–4 alphanumerics (`[A-Z0-9]{2,4}`, e.g. `B738`, `A20N`, `E190`); IATA aircraft code 3 alphanumerics (`738`, `32N`).
+
+### Airport / flight / seat format rules
+- IATA airport code `[A-Z]{3}`; ICAO location indicator `[A-Z0-9]{4}` (US `K…`, Canada `C…`, UK `EG…`, Sweden `ES…`). Use OurAirports rows so codes are real.
+- Flight number: 2-char IATA designator + 1–4 digits, no leading zeros in display (`QF9`, `AA1234`); systems cap at 0001–9999. Conventions (not rules): 1–999 mainline, 3000–5999 regional affiliates, ≥6000 codeshares, 8xxx charters, ≥9000 ferry/positioning. ICAO form: 3-letter designator + 1–4 alphanumerics (`AFR1`). Source: https://en.wikipedia.org/wiki/Flight_number.
+- Seat: `row + letter`. Rows typically 1–~60 (A380 up to ~90); many airlines skip 13 (and 14/17 on some carriers). Letters `A–K` **skipping I** (looks like 1/l); narrow-body 3-3 = `ABC DEF` (A/F windows, C/D aisles); wide-body 3-4-3 = `ABC DEFG HJK`, 2-4-2 = `AB DEFG JK`. Regex: `^[1-9][0-9]?[A-HJK]$`.

docs/research/part3-science-text-dates.md

@@ -0,0 +1,149 @@
+# Part 3 — Science/nature, Text, Dates: open data sources for an MIT fake-data library
+
+Researched 2026-09-17. "OK to embed" = redistributable inside an MIT repo with the stated notice. Anything not fetched first-hand is marked **UNVERIFIED**.
+
+## 7. Science / nature
+
+### Periodic table
+
+| Source | Licence | Embed in MIT repo | Fields | Rows | Format |
+|---|---|---|---|---|---|
+| PubChem periodic table — CSV https://pubchem.ncbi.nlm.nih.gov/rest/pug/periodictable/CSV, JSON https://pubchem.ncbi.nlm.nih.gov/rest/pug/periodictable/JSON (page: https://pubchem.ncbi.nlm.nih.gov/periodic-table/) | US-government work, public domain. NLM policy (https://www.ncbi.nlm.nih.gov/home/about/policies/): "Information that is created by or for the US government on this site is within the public domain … may be freely distributed and copied. However, it is requested that in any subsequent use of this work, NLM be given appropriate acknowledgment." | Yes; add an acknowledgment line ("Element data: PubChem/NLM"). | `AtomicNumber, Symbol, Name, AtomicMass, CPKHexColor, ElectronConfiguration, Electronegativity, AtomicRadius, IonizationEnergy, ElectronAffinity, OxidationStates, StandardState, MeltingPoint, BoilingPoint, Density, GroupBlock, YearDiscovered` (17) | 118 (Z=1..118, verified) | CSV; JSON is `{Table:{Columns:{Column:[...]},Row:[{Cell:[...]}]}}` |
+| NIST periodic table https://www.nist.gov/pml/periodic-table-elements | Public domain (17 USC 105); NIST asks for citation (https://www.nist.gov/open/copyright-fair-use-and-licensing-statements-srd-data-software-and-technical-series-publications) | Yes, but | PDF only (no CSV/JSON) — not usable as a data source | 118 | PDF |
+| Bowserinator/Periodic-Table-JSON https://github.com/Bowserinator/Periodic-Table-JSON | **CC BY-SA 3.0** (LICENSE.md, verified; data derived from Wikipedia) | **No** — ShareAlike is incompatible with MIT redistribution | 33 keys incl. `name, symbol, number, atomic_mass, category, phase, period, group, block, shells, electron_configuration, summary, cpk-hex, …` | **119** (includes hypothetical Ununennium) | JSON, CSV |
+| faker-js `src/locales/en/science/chemical_element.ts` https://github.com/faker-js/faker (branch `next`) | MIT (LICENSE verified) | Yes, keep MIT notice | `{symbol, name, atomicNumber}` | 118 | TS |
+
+Validation rules: `AtomicNumber` 1..118 unique; `Symbol` 1–2 letters, first uppercase, unique; `Name` unique. PubChem leaves numeric fields empty for Z≥104 where unknown; `ElectronConfiguration` for Og is `"[Rn]7s2 7p6 5f14 6d10 (predicted)"` and `StandardState` is `"Expected to be a Gas"` — strip "(predicted)"/"Expected to be a" if you want an enum. `AtomicMass` for unstable elements is the mass number of the most stable isotope as a bare decimal (e.g. `295.216`), no brackets. `CPKHexColor` is 6 hex digits without `#`, empty for some.
+
+### SI units
+
+- Source: BIPM SI Brochure, 9th ed. (2019; current revision v4.01, 2026) https://www.bipm.org/en/publications/si-brochure ; PDF https://www.bipm.org/documents/20126/41483022/SI-Brochure-9-EN.pdf
+- Licence: **CC BY 4.0** — stated on the brochure page and on BIPM's copyright page (https://www.bipm.org/en/copyright): content may be adapted, copied, used in commercial products with "appropriate acknowledgement of the BIPM and its source"; BIPM name/logo must not imply endorsement. Verified via the HTML pages; the PDF's own front-matter licence line is **UNVERIFIED** (could not text-extract the PDF here).
+- Embed: Yes — the unit *names/symbols* are facts (not copyrightable); attribute "SI Brochure, BIPM, CC BY 4.0" anyway.
+- Data (cross-checked against https://en.wikipedia.org/wiki/International_System_of_Units):
+  - 7 base units: second s (time), metre m (length), kilogram kg (mass), ampere A (electric current), kelvin K (thermodynamic temperature), mole mol (amount of substance), candela cd (luminous intensity).
+  - 22 coherent derived units with special names: radian rad, steradian sr, hertz Hz, newton N, pascal Pa, joule J, watt W, coulomb C, volt V, farad F, ohm Ω, siemens S, weber Wb, tesla T, henry H, degree Celsius °C, lumen lm, lux lx, becquerel Bq, gray Gy, sievert Sv, katal kat.
+  - 24 prefixes: quetta Q 10^30, ronna R 10^27, yotta Y 10^24, zetta Z 10^21, exa E 10^18, peta P 10^15, tera T 10^12, giga G 10^9, mega M 10^6, kilo k 10^3, hecto h 10^2, deca da 10^1, deci d 10^-1, centi c 10^-2, milli m 10^-3, micro µ 10^-6, nano n 10^-9, pico p 10^-12, femto f 10^-15, atto a 10^-18, zepto z 10^-21, yocto y 10^-24, ronto r 10^-27, quecto q 10^-30.
+- Format rules for valid output: unit symbols are case-sensitive (`s` vs `S`, `m` vs `M`); prefix and unit symbol are joined without a space (`kJ`, `µs`); a space separates number and unit (`12 kg`), except the degree/minute/second of plane angle; `kg` already carries a prefix — never `µkg`; micro is Greek mu U+03BC (U+00B5 micro sign is the legacy compatibility character); ohm is U+03A9 (U+2126 OHM SIGN is deprecated); `°C` not `° C`.
+- MIT alternative: faker-js `en/science/unit.ts` — 29 `{name, symbol}` objects, MIT.
+
+### Animals by class
+
+- **Wikidata** (https://www.wikidata.org/wiki/Wikidata:Licensing): all structured data **CC0** — no attribution required. Embed: Yes.
+  - Feasibility, live SPARQL (https://query.wikidata.org/sparql), pattern `?t wdt:P31 wd:Q16521; wdt:P105 wd:Q7432; wdt:P171* wd:<class>; wdt:P1843 ?cn FILTER(LANG(?cn)='en')` (taxon, rank species, parent-taxon chain, English common name):
+    - Mammalia Q7377: **5,759** species with ≥1 English common name.
+    - Aves Q5113: **12,432**.
+    - Reptilia Q10811: **16,369** — suspiciously large; Wikidata's parent chain likely pulls birds/Sauropsida in. Use Squamata Q122422 + Testudines Q223044 + Crocodilia Q1387 (or check the chain) instead.
+    - Insecta Q1390 and Plantae Q756: **timed out (502/504) every attempt** — trees too big for the 60 s public endpoint. Options: paginate by order (`P171` one level at a time), use the weekly JSON dump, or take GBIF instead.
+  - Sample rows (mammals): `Notoryctes typhlops` → "Central Desert Marsupial Mole", "Itjaritjari", "Marsupial Mole", "Southern marsupial mole", "Southern Marsupial Mole" — i.e. **several P1843 values per taxon, inconsistent casing, and indigenous-language names tagged `en`**. Validation: pick one name per taxon (prefer the one matching the item's English `rdfs:label`), normalise case, drop names that equal the scientific name.
+- **GBIF Backbone Taxonomy** https://www.gbif.org/dataset/d7dddbf4-2cf0-4f39-9b2a-bb099caae36c — licence **CC BY 4.0** (verified via https://api.gbif.org/v1/dataset/d7dddbf4-2cf0-4f39-9b2a-bb099caae36c, `license: http://creativecommons.org/licenses/by/4.0/legalcode`); required citation text: "GBIF Secretariat (2023). GBIF Backbone Taxonomy. Checklist dataset https://doi.org/10.15468/39omei". Embed: Yes with that citation. Vernacular names come via `/v1/species/{key}/vernacularNames`, each row from a *different* source checklist with its own licence (some CC BY-NC) — record and check per row. GBIF's site terms page (https://www.gbif.org/terms) was unreachable (timeout) — **UNVERIFIED** beyond the API field.
+- **faker-js** `src/locales/en/animal/*` (MIT; counts = quoted string lines on branch `next`): bird 821, snake 533, dog 493, cow 467, horse 342, rodent 161, insect 129, fish 95, cat 55, cetacean 52, rabbit 49, type 44 (generic: "bat, bear, bee, …, zebra"), pet_name 42, crocodilia 24, bear 8, lion 7. These are breeds/species mixed with common names, no class field; fine as MIT drop-ins. (A WebFetch summariser claimed 1,036 dogs; the regex count is 493 — treat the exact count as UNVERIFIED.)
+
+### Plants
+
+- Same Wikidata approach (CC0) — but the Plantae query timed out; paginate by order or family (`?t wdt:P171 ?family . ?family wdt:P105 wd:Q35409`) or use the dump.
+- GBIF Plantae (kingdomKey 6) vernacular names, CC BY 4.0, same per-row-licence caveat.
+- faker-js has **no** plant list (en locale dirs: airline, animal, app, book, color, commerce, company, database, date, finance, food, hacker, internet, location, lorem, medical, music, person, phone_number, science, team, vehicle, word).
+- USDA PLANTS database (public domain, has common names) — **UNVERIFIED**, not fetched.
+
+### Planets / moons
+
+- NASA NSSDCA Planetary Fact Sheet https://nssdc.gsfc.nasa.gov/planetary/factsheet/ — 10 columns (Mercury, Venus, Earth, Moon, Mars, Jupiter, Saturn, Uranus, Neptune, Pluto) × 20 rows (mass 10^24 kg, diameter km, density, gravity, escape velocity, rotation period, day length, distance from Sun, perihelion, aphelion, orbital period, orbital velocity, inclination, eccentricity, obliquity, mean temp °C, surface pressure, number of moons, rings Y/N, magnetic field Y/N). HTML table only. NASA content "generally not subject to copyright in the United States" (https://www.nasa.gov/nasa-brand-center/images-and-media/); NASA asks to be acknowledged; the NASA insignia must not be used. Embed: Yes with "Source: NASA NSSDCA".
+- Wikidata (CC0): planets = `?p wdt:P31/wdt:P279* wd:Q634; wdt:P397 wd:Q525` → Mercury Q308, Venus Q313, Earth Q2, Mars Q111, Jupiter Q319, Saturn Q193, Uranus Q324, Neptune Q332 (plus hypothetical Theia Q1053432 — filter out). Moons = `?m wdt:P397 <planet>; wdt:P31/wdt:P279* wd:Q2537` (natural satellite; note **Q2199 is "dwarf planet"**, not moon): Saturn 165, Jupiter 85, Uranus 29, Neptune 16, Earth 5, Mars 2. Caveats: Wikidata lags IAU counts (Saturn has 274 recognised moons as of 2025 — UNVERIFIED figure); Earth's 5 includes quasi-satellites; provisional designations (`S/2004 S 3`) are labels too — filter `^S/\d{4}` if you want proper names only.
+
+### Blood type distribution
+
+- **US** (Stanford Blood Center https://stanfordbloodcenter.org/donate-blood/blood-donation-facts/blood-types/, verified; cites AABB Technical Manual 18th ed.): O+ 37.4 %, A+ 35.7 %, B+ 8.5 %, AB+ 3.4 %, O− 6.6 %, A− 6.3 %, B− 1.5 %, AB− 0.6 % (sums to 100.0). Facts — freely embeddable; cite Stanford/AABB.
+- American Red Cross https://www.redcrossblood.org/donate-blood/blood-types.html — **UNVERIFIED**: site returns HTTP 403 (Akamai) to both WebFetch and the headless browser. Search snippets say the Red Cross gives type O by ethnicity (Latino 57 %, African American 51 %, Caucasian 45 %) rather than an 8-way national table.
+- **Global**: no authoritative figure found. Wikipedia "Blood type distribution by country" gives a population-weighted row (O+ 38.4, A+ 27.3, B+ 8.1, AB+ 2.0, O− 13.1, A− 8.1, B− 2.0, AB− 0.01) that its own editors flag "unreliable source"; WorldAtlas gives O+ 42, A+ 31, B+ 15, AB+ 5, O− 3, A− 2.5, B− 1, AB− 0.5 with no source. Recommend shipping only the US table, or per-country tables from Wikipedia (facts; Wikipedia prose is CC BY-SA but the numbers are cited to national blood services — cite those).
+- Validation: ABO ∈ {A, B, AB, O}; Rh ∈ {+, −}; weights must sum to 100 ± rounding.
+
+## 8. Text
+
+### English word lists with part of speech
+
+| Source | Licence | Embed | Counts | Format |
+|---|---|---|---|---|
+| Princeton WordNet 3.0 / 3.1 https://wordnet.princeton.edu/license-and-commercial-use ; download https://wordnetcode.princeton.edu/wn3.1.dict.tar.gz | "WordNet License" (SPDX `WordNet`), BSD-style: permission to use/copy/modify/distribute "for any purpose and without fee or royalty … provided that … the following copyright notice and statements, including the disclaimer … appear on ALL copies … including modifications". Notice text: "WordNet 3.0 Copyright 2006 by Princeton University. All rights reserved." + AS-IS disclaimer; no use of Princeton's name in advertising. | Yes — ship the LICENSE text alongside the data | WordNet 3.0 (wnstats(7WN), verified): unique strings noun 117,798 / verb 11,529 / adj 21,479 / adv 4,481 (total 155,287); synsets 82,115 / 13,767 / 18,156 / 3,621 | WNDB: `index.noun|verb|adj|adv` (one lemma per line: `lemma pos synset_cnt p_cnt [ptr…] sense_cnt tagsense_cnt synset_offset…`), `data.<pos>` (synsets with glosses). Multi-word lemmas use `_` for spaces; lower-case; entries can contain digits/apostrophes/hyphens. |
+| Open English WordNet 2025 https://en-word.net/ ; https://github.com/globalwordnet/english-wordnet | **CC BY 4.0**; cite "Open English WordNet … derived from Princeton WordNet" | Yes, with attribution | 135,969 words, 107,519 synsets (README, verified) | `english-wordnet-2025.zip` (WNDB, 9.2 MB), `english-wordnet-2025.xml.gz` (WN-LMF, 10.8 MB), `english-wordnet-2025-json.zip` (9.5 MB), `.ttl.gz` (16.9 MB) |
+| SCOWL v2 / ESDB https://github.com/en-wl/wordlist (branch v2) ; classic http://wordlist.aspell.net/ | Custom MIT-like (Copyright 2000-2026 Kevin Atkinson: use/copy/modify/distribute/sell "provided that the above copyright notice appears in all copies and that both the above copyright notice and this notice appear in supporting documentation"). Sources 12dicts + ENABLE2K are public domain; lists above size 80 add the UKACD notice; **POS assignments partly from WordNet, so the WordNet notice "MIGHT apply"** (their words). Australian-English parts carry an extra notice. | Yes, ship the Copyright file | Sizes 35 (small), 50 (medium), 60 (spell-check default), 70 (large), 80 (incl. game words), 85 (archaic). No per-size counts published in the v2 README. Classic SCOWL sizes 10–95 — counts **UNVERIFIED** (README not reachable). | v2: `scowl.db` (SQLite), `scowl.txt`, Python `libscowl`. ESDB rows carry POS, spelling (A/B/Z/C/D) and region codes. |
+| 12dicts http://wordlist.aspell.net/12dicts-readme/ | Public domain ("I explicitly release them to the public domain, but request acknowledgment"), **except** `2of12inf` and the `2+2+3*` lists, which depend on AGID and inherit its terms | Yes (PD lists); acknowledge Alan Beale | 3esl ≈22k, 6of12 ≈32k, 2of12 ≈41k, 2of12inf ≈82k, 3of6game ≈65k, 5d+2a ≈68k, 3of6all ≈83k, 2+2+3lem ≈84k, 2of5core ≈4.7k, 6phrase ≈22k, neol2016 ≈600 | Plain text, one word per line; marker suffixes `+ ! ^ &` (not POS). **No POS tags** in any current 12dicts list. |
+| Moby Part-of-Speech II https://www.gutenberg.org/ebooks/3203 ; file https://www.gutenberg.org/files/3203/files/mobypos.txt | Public domain ("Public Domain material by grant from the author, January, 2001", in the package README) | Yes, no notice needed | **233,356** lines (verified) | One entry per line: `word\CODES`, CRLF line ends (e.g. `A-line\NA`, `a tempo\Avh`). Codes in priority order: `N` noun, `p` plural, `h` noun phrase, `V` verb (usu. participle), `t` transitive verb, `i` intransitive verb, `A` adjective, `v` adverb, `C` conjunction, `P` preposition, `!` interjection, `r` pronoun, `D` definite article, `I` indefinite article, `o` nominative. Many entries are phrases, proper nouns, or 1990s-era; non-ASCII entries use a legacy 8-bit encoding (e.g. `a bon march\v` lost its `é`) — restrict to ASCII `[a-z]+` for a clean list. |
+| faker-js `en/word/*` (MIT) | MIT | Yes | adjective 1000, noun 1000, verb 1000, adverb 325, preposition 109, conjunction 51, interjection 46 | TS string arrays |
+
+### Word frequency lists
+
+| Source | Licence | Verdict |
+|---|---|---|
+| Google Books Ngram data v2/v3 https://storage.googleapis.com/books/ngrams/books/datasetsv3.html | **CC BY 3.0** ("This compilation is licensed under a Creative Commons Attribution 3.0 Unported License") | Embeddable with attribution. Format `ngram TAB year TAB match_count TAB volume_count`; 1-gram files split by leading letter, GBs each — aggregate offline to a top-N list. |
+| wordfreq https://github.com/rspeer/wordfreq | Code **Apache 2.0** (not MIT). Data: "may be redistributed under a Creative Commons Attribution-ShareAlike 4.0 license" (mix of Google Books, Wikipedia, OpenSubtitles, SUBTLEX, Leeds, ParaCrawl, Twitter); maintainer notes the CSV export "does not follow the CC-By-SA license". Project in sunset mode (data snapshot ≈2021). | **Avoid** embedding the data (ShareAlike). |
+| Peter Norvig count_1w.txt https://norvig.com/ngrams/ | Page says "Code … under the MIT license"; the **data** comes from the Google Web 1T corpus distributed by LDC under LDC terms — no data licence stated. | **UNVERIFIED / avoid**. 333,333 words, `word TAB count`, lowercase. |
+| hermitdave/FrequencyWords https://github.com/hermitdave/FrequencyWords | "MIT License for code. CC-by-sa-4.0 for content." | **Avoid** (ShareAlike). |
+
+### Lorem ipsum
+
+- Canonical paragraph (Letraset, 1966; public domain — garbled Cicero): "Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum."
+- Source: Cicero, *De finibus bonorum et malorum* 1.10.32–33 (45 BC), public domain; the "dolorem ipsum" fragment begins "Neque porro quisquam est qui dolorem ipsum quia dolor sit amet, consectetur, adipisci velit…". Reference: https://en.wikipedia.org/wiki/Lorem_ipsum
+- MIT word pool alternative: faker-js `en/lorem/word.ts` — 999 words.
+
+### Quote collections
+
+- Project Gutenberg #27889 https://www.gutenberg.org/ebooks/27889 — Bartlett, *Familiar Quotations*, **9th edition** (title page "NINTH EDITION", copyright 1875/1882/1891/1903) — not the 1919 10th. "Public domain in the USA." Plain text 3.1 MB, HTML zip 50 MB, EPUB.
+- Project Gutenberg #16732 https://www.gutenberg.org/ebooks/16732 — an early (Hurst & Co. reprint) edition, no edition number given. PD in USA.
+- 10th edition (1914/1919, ed. Nathan Haskell Dole) exists on archive.org (https://archive.org/details/bartlettsfamilia0000john_o4b4) — not found on Gutenberg; **UNVERIFIED** availability as clean text.
+- Either Bartlett text needs parsing (quote / author / source headings); no structured file. Wikiquote is CC BY-SA — avoid.
+- Validation: strip Gutenberg header/footer (`*** START OF THE PROJECT GUTENBERG EBOOK` … `*** END …`) before extraction; Gutenberg's trademark terms only bind if you keep the "Project Gutenberg" name on the output — strip it.
+
+### Colour names
+
+| Source | Licence | Embed | Count / format |
+|---|---|---|---|
+| CSS Color Module Level 4 §named colors https://www.w3.org/TR/css-color-4/#named-colors | W3C Document License (2023) https://www.w3.org/copyright/document-license-2023/ — copy/distribute allowed for any purpose; derivatives allowed only "to facilitate implementation of the technical specifications"; notice required: "Copyright © 2023 W3C®. This software or document includes material copied from or derived from [title and URI]." | Yes — the name→hex table is factual data and is copied into every browser; include the W3C notice line with the spec title + URI. | **148** names (verified: aliceblue … yellowgreen, incl. `rebeccapurple #663399`); all names lowercase ASCII, hex 6 lowercase digits. |
+| XKCD colour survey https://xkcd.com/color/rgb.txt | **CC0** (file header: `# License: https://creativecommons.org/publicdomain/zero/1.0/`) | Yes, no notice needed | **949** lines (verified). Format `name TAB #rrggbb TAB` (note the trailing tab). Max name length 26; all hex valid lowercase; no duplicate names. Contains crude names ("poo", "baby poop green", "diarrhea", "vomit") — apply a blocklist. |
+| faker-js `en/color/human.ts` | MIT | Yes | 31 names |
+| Pantone | Proprietary | **Avoid** | — |
+
+## 9. Dates
+
+### IANA tz database
+
+- Source https://www.iana.org/time-zones ; files https://data.iana.org/time-zones/tzdb/ (current version **2026d**, verified). LICENSE: "Unless specified below, all files in the tz code and data (including this LICENSE file) are in the public domain" (only `date.c`, `newstrftime.3`, `strftime.c` are BSD-3). Embed: Yes, no notice needed.
+- `zone1970.tab` — **312** data rows; tab-separated; UTF-8; `#` comments. Columns: (1) comma-separated ISO 3166-1 alpha-2 codes of countries overlapping the zone, most-populous first; (2) coordinates in ISO 6709 `±DDMM±DDDMM` or `±DDMMSS±DDDMMSS` (lat then lon, no separator); (3) TZ name (`Europe/Zurich`); (4) comment, present iff the country has several zones. One row per timezone where civil time has agreed since 1970.
+- `zone.tab` — **418** rows, older one-country-per-row table (column 1 is a single country code); kept for backward compatibility. Prefer `zone1970.tab`.
+- `backward` — **252** `Link TARGET LINK-NAME` lines mapping old/merged names (`US/Eastern`, `Asia/Calcutta`, `Europe/Kiev`) to current ones; also a few `Zone` entries. Use it to (a) accept aliases as input and (b) never emit them as output.
+- UTC offsets: do **not** embed — offsets depend on DST rules that change several times a year. Derive at runtime: `new Intl.DateTimeFormat('en-US', { timeZone, timeZoneName: 'longOffset' }).formatToParts(date)` → part `timeZoneName` = `GMT-05:00` (or `GMT` for zero) — strip `GMT`, treat bare `GMT` as `+00:00`. Validation of a generated zone name: `Intl.supportedValuesOf('timeZone')` (canonical names only; whether links like `US/Eastern` appear depends on the ICU build) or `try { new Intl.DateTimeFormat(undefined, { timeZone }) } catch { invalid }`, which accepts links too.
+
+### ISO 8601 layouts (https://en.wikipedia.org/wiki/ISO_8601)
+
+| Layout | Extended | Basic | Example |
+|---|---|---|---|
+| Calendar date | `YYYY-MM-DD` | `YYYYMMDD` | 2009-01-06 |
+| Ordinal date | `YYYY-DDD` | `YYYYDDD` | 1981-095 |
+| Week date | `YYYY-Www-D` | `YYYYWwwD` | 2009-W01-1 (weeks start Monday, week 1 contains 4 Jan) |
+| Time | `hh:mm:ss[.fff]` | `hhmmss` | 13:47:30 (24 h; `T` prefix optional standalone) |
+| UTC / offset | `Z` or `±hh:mm` | `±hhmm`, `±hh` | 2007-04-05T14:30Z, 2024-06-01T09:00-05:00 |
+| Date-time | date `T` time offset | | 2007-04-05T14:30:00+02:00 |
+| Duration | `PnYnMnDTnHnMnS` / `PnW` | | P3Y6M4DT12H30M5S |
+| Interval | start`/`end, start`/`duration, duration`/`end | | 2007-03-01T13:00:00Z/2008-05-11T15:30:00Z |
+| Recurring | `Rn/`interval | | R5/2008-03-01T13:00:00Z/P1Y2M10DT2H30M |
+
+RFC 3339 profile (what JSON/APIs usually mean): extended forms only, `T`/`Z` may be lowercase, space allowed instead of `T`, `-00:00` = unknown offset, no durations/intervals/week/ordinal dates. Generate RFC 3339 by default.
+
+### Unicode CLDR (locale month/weekday names, date patterns)
+
+- Licence: **Unicode License v3** (SPDX `Unicode-3.0`), OSI-approved 2023-11-17 (https://opensource.org/license/unicode-3-0). Text https://www.unicode.org/license.txt — MIT-style: use/copy/modify/sell "provided that either (a) this copyright and permission notice appear with all copies of the Data Files or Software, or (b) this copyright and permission notice appear in associated Documentation"; no use of the Unicode name in advertising. Embed: Yes; put the notice in LICENSE/NOTICE.
+- npm: `cldr-core`, `cldr-dates-full`, `cldr-numbers-full` (peer of dates), etc.; current **48.2.0** (CLDR 48, verified via npm registry, `license: Unicode-3.0`); repo https://github.com/unicode-org/cldr-json .
+- Path is **`cldr-dates-full/main/<locale>/ca-gregorian.json`** (not `dates/gregorian.json`), rooted at `main.<locale>.dates.calendars.gregorian`. Keys (verified for `en`): `months`, `days`, `quarters`, `dayPeriods`, `eras`, `dateFormats`, `dateSkeletons`, `timeFormats`, `timeSkeletons`, `dateTimeFormats`, `dateTimeFormats-atTime`, `dateTimeFormats-relative`.
+  - `months.{format|stand-alone}.{abbreviated|narrow|wide}` keyed `"1".."12"`; `days.{format|stand-alone}.{abbreviated|narrow|short|wide}` keyed `sun..sat`.
+  - `dateFormats`: en = full `EEEE, MMMM d, y`, long `MMMM d, y`, medium `MMM d, y`, short `M/d/yy`.
+  - `timeFormats`: en = full `h:mm:ss a zzzz`, long `h:mm:ss a z`, medium `h:mm:ss a`, short `h:mm a` — **the `a` is preceded by U+202F NARROW NO-BREAK SPACE**; `*-alt-ascii` variants use a plain space. Pick one spelling (ASCII) and ignore the other.
+  - `dateTimeFormats.{full|long|medium|short}` = glue like `{1}, {0}` ({1} date, {0} time); `availableFormats` = skeleton→pattern map (`yMMMd` → `MMM d, y`), plus `intervalFormats`, `appendItems`.
+  - `dayPeriods.format.abbreviated`: `am: AM`, `pm: PM`, plus `midnight`, `noon`, `morning1`… variants.
+- Pattern letters follow LDML (https://unicode.org/reports/tr35/tr35-dates.html#Date_Field_Symbol_Table): `y` year, `M`/`L` month (format/stand-alone), `d` day, `E`/`c` weekday, `h` 1–12, `H` 0–23, `a` AM/PM, `z`/`zzzz` zone name, `x`/`X` ISO offset; literal text in single quotes. Same patterns drive `Intl.DateTimeFormat` internally, so runtime `Intl` can replace embedding for month/weekday names if bundle size matters.
+
+## Summary of licence verdicts
+
+- Embed freely (PD/CC0): PubChem periodic table, NIST, NASA fact sheet, Wikidata, Moby POS, 12dicts (PD lists), Bartlett via Gutenberg, XKCD colours, IANA tz, Lorem ipsum/Cicero.
+- Embed with notice: WordNet 3.x (WordNet License), Open English WordNet (CC BY 4.0), SCOWL/ESDB (custom MIT-like), GBIF backbone (CC BY 4.0 + citation), BIPM SI (CC BY 4.0), CSS named colours (W3C notice), CLDR (Unicode-3.0), Google Books ngrams (CC BY 3.0), faker-js lists (MIT).
+- Avoid: Bowserinator periodic table (CC BY-SA 3.0), wordfreq data (CC BY-SA 4.0), hermitdave (CC BY-SA 4.0), Norvig count_1w (LDC-derived, unstated), Wikiquote (CC BY-SA), Pantone.
+- UNVERIFIED: American Red Cross figures (site blocks fetches), any global blood-type table, BIPM PDF front-matter licence line, classic SCOWL per-size counts, USDA PLANTS, IAU moon counts, exact faker dog count, Bartlett 10th-edition text availability.

docs/research/part4-iso-commerce-books.md

@@ -0,0 +1,139 @@
+# Part 4 — ISO lists, commerce, books/music/food
+
+Researched 2026-09-17. "Embed?" = can the data ship inside an MIT-licensed repo. Anything marked UNVERIFIED was not confirmed against the primary source.
+
+## 10. ISO lists
+
+### ISO 3166-1 countries
+
+| Source | Licence | Embed in MIT repo? | Rows | Fields | Format |
+|---|---|---|---|---|---|
+| [datasets/country-codes](https://github.com/datasets/country-codes) | PDDL (README: "Public Domain Dedication and License"; GitHub API detects no licence file) | Yes, no attribution required. README caveat: ISO itself says its list is "for internal use and non-commercial purposes free of charge" — the repo argues no rights subsist in a list of facts | 249 | ISO3166-1-Alpha-2/-Alpha-3/-numeric, official_name_en, UNTERM names (ar/zh/en/fr/ru/es short+formal), Dial, TLD, Capital, Continent, Languages, ISO4217 currency code/name/numeric/minor_unit, Region/Sub-region/Intermediate names+codes (M49), FIPS, IOC, FIFA, MARC, ITU, WMO, GAUL, EDGAR, Geoname ID, wikidata_id, CLDR display name, is_independent, LDC/LLDC/SIDS flags | CSV `data/country-codes.csv` |
+| [annexare/Countries](https://github.com/annexare/Countries) (npm `countries-list` 3.4.1, 2026-07-14) | MIT | Yes; keep MIT notice | 252 countries, 185 languages (incl. non-ISO entries like AC, XK) | per country: name, native, phone[] (calling codes), continent, capital, currency[], languages[], optional alias[], partOf; separate ISO 4217 currency table (name, native, symbol, numeric, decimals); languages: name, native | TS source, exported JSON/CSV/SQL |
+| [Debian iso-codes](https://salsa.debian.org/iso-codes-team/iso-codes) | LGPL-2.1-or-later (Debian copyright file) | Not safely as embedded data in an MIT repo — LGPL applies to the data files; you would have to ship it as a separately licensed component with LGPL notice. Avoid unless accepting that | 3166-1: 249 | alpha_2, alpha_3, numeric, name, official_name, common_name, flag | JSON `data/iso_3166-1.json` (raw URL needs `?inline=false`) |
+| [mledoze/countries](https://github.com/mledoze/countries) | ODbL-1.0 | No — share-alike + attribution; derived databases must be ODbL. Avoid | ~250 | common/official names, cca2/cca3/ccn3/cioc, tld, currencies, idd (calling codes), capital, region/subregion, languages, translations, latlng, borders, area, demonyms, flag emoji, UN member | JSON/CSV/XML/YAML |
+| [restcountries](https://gitlab.com/restcountries/restcountries) | MPL-2.0 (source repo; hosted API v5 is proprietary/API-key) | File-level copyleft: a copied data file stays MPL and must carry the notice; OK to ship next to MIT code but not to relicense. Prefer PDDL/MIT sources | 250+ | name, tld, cca2, ccn3, cca3, cioc, fifa, independent, status, unMember, currencies, idd, capital, capitalInfo, altSpellings, region, subregion, continents, languages, translations, latlng, landlocked, borders, area, flag (emoji), demonyms, flags, coatOfArms, population, maps, gini, car, postalCode, startOfWeek, timezones | JSON `src/main/resources/countriesV3.1.json` |
+| [lukes/ISO-3166-Countries-with-Regional-Codes](https://github.com/lukes/ISO-3166-Countries-with-Regional-Codes) | CC BY-SA 4.0 (LICENSE.md) | No — share-alike. Avoid | 249 | name, alpha-2, alpha-3, country-code, iso_3166-2, region, sub-region, intermediate-region + codes | CSV/JSON/XML (all, slim-2, slim-3) |
+| [stefangabos/world_countries](https://github.com/stefangabos/world_countries) | Data: CC BY-SA 4.0 (README, GitHub licence detection); npm `world_countries_lists` package.json says LGPL-3.0-or-later — contradictory | No — share-alike either way. Avoid | 249 world / 193 UN; subdivisions 5046 | id (numeric), alpha2, alpha3, name (37 languages); subdivisions: country, code, name, name_en, type, parent | JSON/CSV/PHP/SQL/XML |
+| [i18n-iso-countries](https://www.npmjs.com/package/i18n-iso-countries) 7.14.0 | MIT | Yes | 78 language files | alpha2, alpha3, numeric, localized names | JSON per language |
+| Wikipedia ISO 3166-1 | CC BY-SA 4.0 | No (share-alike) | — | — | — |
+| [ISO OBP](https://www.iso.org/obp) | Proprietary; ISO grants free use only for "internal use and non-commercial purposes" (quoted in datasets/country-codes README) | No | — | — | — |
+
+- Flag emoji: derive, no data needed — for alpha-2 `XY`, emit U+1F1E6 + (X − 'A') and U+1F1E6 + (Y − 'A') (regional indicator symbols).
+- Recommendation: `datasets/country-codes` (PDDL) covers every requested field (name, alpha2, alpha3, numeric, calling code, TLD, capital, currency, languages) in one CSV; flag emoji derived. `annexare` (MIT) is the fallback for native names/calling codes if PDDL's ISO caveat worries you; it lacks TLD.
+- Validation rules: alpha-2 `^[A-Z]{2}$`, alpha-3 `^[A-Z]{3}$`, numeric `^\d{3}$` (zero-padded, keep as string); calling code 1–3 digits, NANP countries share `1`; TLD `^\.[a-z]{2}$` (ccTLD = lowercase alpha-2, exceptions: GB uses `.uk`).
+
+### ISO 639 languages
+
+| Source | Licence | Embed? | Rows | Fields | Format |
+|---|---|---|---|---|---|
+| [langs](https://www.npmjs.com/package/langs) 2.0.0 (2017, unmaintained) | MIT | Yes | 184 | name, local, 1 (639-1), 2, 2T, 2B, 3 | `data.js` array |
+| [iso-639-1](https://www.npmjs.com/package/iso-639-1) 3.1.6 (2026-07-02) | MIT | Yes | 183 | code, name, nativeName | `src/data.js` |
+| annexare/Countries languages | MIT | Yes | 185 | code (639-1), name, native | TS/JSON |
+| Debian iso-codes 639-2 / 639-3 / 639-5 | LGPL-2.1+ | See above (LGPL) | 639-2: 487; 639-3: 7923 | 639-2: alpha_2, alpha_3, bibliographic, name, common_name; 639-3: + inverted_name, scope, type | JSON |
+| [SIL ISO 639-3 tables](https://iso639-3.sil.org/code_tables/download_tables) | SIL terms: may incorporate into software (commercial or not) with attribution to iso639-3.sil.org, must not modify identifiers, and product "does not provide a means to redistribute the code set" | Risky — a git repo with the table IS a means to redistribute; avoid embedding the full table. Codes themselves are facts | 7927 | Id, Part2b, Part2t, Part1, Scope, Language_Type, Ref_Name, Comment | tab-delimited `iso-639-3.tab` |
+| [Library of Congress ISO 639-2](https://www.loc.gov/standards/iso639-2/ascii_8bits.html) | US federal government work → public domain in the US | Yes | ~487 (matches Debian 639-2 count) | pipe-delimited: alpha3-B, alpha3-T, alpha2, English name, French name — UNVERIFIED: loc.gov is behind a Cloudflare bot check (WebFetch, curl and Playwright all blocked); field layout from memory | `ISO-639-2_utf-8.txt` |
+
+- Recommendation: `langs` or `iso-639-1` (MIT) for 639-1 names; both are small and derived from public code tables. Anything 639-3-sized is LGPL or SIL-restricted.
+- Rules: 639-1 `^[a-z]{2}$`, 639-2/3 `^[a-z]{3}$`; 639-2 has B/T pairs (e.g. `fre`/`fra`, `ger`/`deu`) — pick one consistently (T codes = 639-3).
+
+### ISO 3166-2 subdivisions
+
+| Source | Licence | Embed? | Rows | Fields |
+|---|---|---|---|---|
+| Debian iso-codes 3166-2 | LGPL-2.1+ | LGPL caveat | 5046 | code, name, type, parent |
+| [olahol/iso-3166-2.json](https://github.com/olahol/iso-3166-2.json) | package.json says ISC; no LICENSE file; GitHub detects none; data scraped from eQuest xls (provenance unclear); last push 2021 | No — unclear licence and provenance | 237 countries / 3807 divisions | `{alpha2: {name, divisions: {code: name}}}` |
+| [esosedi/3166](https://github.com/esosedi/3166) (npm `iso3166-2-db` 2.3.11) | MIT | Yes; data mixes GeoNames (CC BY 4.0), Wikipedia (CC BY-SA), OSM (ODbL) — the MIT label may not survive scrutiny for the Wikipedia/OSM-derived parts. UNVERIFIED how much is derived from each | "all countries and regions" (count not stated) | ISO 3166-2 + FIPS codes, names in 12 languages, admin level, GeoNames/OSM/Wikipedia/WOF ids |
+| stefangabos subdivisions | CC BY-SA 4.0 | No | 5046 | country, code, name, name_en, type, parent |
+| [US Census state.txt](https://www2.census.gov/geo/docs/reference/state.txt) | US government work → public domain | Yes | 57 (50 states + DC + 6 territories) | STATE (FIPS), STUSAB, STATE_NAME, STATENS |
+
+- Recommendation: for an en_US library, ship only the US list (Census/USPS, public domain: 50 states + DC, optionally territories) — ISO 3166-2 code = `US-` + USPS abbreviation. Skip the ~5000-row world list; every open compilation is LGPL, CC BY-SA or unclear.
+- Rule: `^[A-Z]{2}-[A-Z0-9]{1,3}$`.
+
+## 11. Commerce
+
+### Word lists (product name / adjective / material)
+
+- [@faker-js/faker](https://github.com/faker-js/faker) 10.6.0 — MIT (LICENSE covers code and locale data; copyright Faker contributors 2022-2025 and Marak Squires 2011-2020; acknowledges Ruby Faker and Perl Data::Faker, both MIT). No separate data licence.
+- Provenance policy in [CONTRIBUTING.md](https://github.com/faker-js/faker/blob/next/CONTRIBUTING.md): "Faker must not contain copyrighted materials"; facts (finite known lists) are OK; compilations must not be copied from a single source (Wikipedia, "most popular" articles, government sites) because "a compilation of facts can be copyrighted" — contributors are told to draw on multiple sources and their own judgement.
+- Verdict: reusing faker's `en` commerce lists (`product_name.adjective/material/product`, `department`) is fine under MIT — keep the MIT notice with the copied data. The lists are short curated words (no external dataset). Nothing in the repo or docs restricts locale-data reuse beyond MIT.
+
+### Garment and shoe sizes
+
+- Letter sizes XS–XXL: EN 13402-3 letter codes with chest/bust ranges (Wikipedia [EN 13402](https://en.wikipedia.org/wiki/EN_13402), CC BY-SA — but the table is a handful of facts; the standard itself is paywalled). Men chest / women bust cm: XXS 70–78 / 66–74, XS 78–86 / 74–82, S 86–94 / 82–90, M 94–102 / 90–98, L 102–110 / 98–107, XL 110–118 / 107–119, XXL 118–129 / 119–131, 3XL 129–141 / 131–143.
+- US numeric women's sizes (0–20, even) come from ASTM D5585 (paywalled); no open dataset. Just generate the even-number series.
+- Shoe sizes: no open dataset found (search hits are ML demos and retailer charts). Use the formulas from Wikipedia [Shoe size](https://en.wikipedia.org/wiki/Shoe_size) (facts, not copyrightable): EU (Paris point) ≈ 1.5 × foot cm + 2 (i.e. 3/20 × mm + 2); UK adult ≈ 3 × foot in − 23; US men ≈ 3 × foot in − 22 = UK + 1; US women (common) = US men + 1.5 (FIA scale: 3 × foot in − 21); Mondopoint = foot length mm. ISO/TS 19407:2023 has the official conversion tables (paywalled). Generate from a foot length (men 24–31 cm, women 21–27 cm) in half-size steps so all systems agree.
+
+### Check-digit rules (all verified against Wikipedia articles; algorithms are facts)
+
+| Code | Rule |
+|---|---|
+| EAN-13 / UPC-A / GTIN-8/12/13/14 | Weights 3,1,3,1… starting from the rightmost data digit (the one next to the check digit); check = (10 − sum mod 10) mod 10. UPC-A = EAN-13 with a leading 0 (GS1 prefix of a 12-digit GTIN is `0` + first two digits). GTIN-14 first digit is an indicator 1–8 (packaging level) or 9 (variable measure); 0 is not valid there. |
+| UPC-A number system (first digit) | 0,1,6,7,8,9 regular; 2 variable-weight; 3 drugs (NDC); 4 in-store/local; 5 coupons. |
+| ISBN-10 | Σ(weights 10..2 × first 9 digits) + check ≡ 0 mod 11; check = (11 − sum mod 11) mod 11, 10 → `X`. |
+| ISBN-13 | EAN-13 with prefix 978 or 979. 979 groups in use: 979-8 USA, 979-10 France, 979-11 Korea, 979-12 Italy. 978 groups 0 and 1 = English-language. For an en_US generator: `978-0…`, `978-1…`, `979-8…`. The official range file (`https://www.isbn-international.org/export_rangemessage.xml`) is XML only and the site says "You must not republish material from our site … without our written permission" — do not embed it; group/registrant length only matters for hyphenation, not validity. |
+| ISSN | Format `NNNN-NNNC`; weights 8..2 over first 7 digits, sum mod 11; check = 0 if remainder 0 else 11 − remainder; 10 → `X`. EAN-13 form: `977` + 7 ISSN digits (no check) + 2 variant digits + EAN check. |
+| IMEI | 15 digits: TAC 8 (RBI 2 + 6) + serial 6 + Luhn check (ISO/IEC 7812; from the right, double every second digit, sum digits, total ≡ 0 mod 10). SVN never enters the check. IMEISV = 14 digits + 2-digit SVN, no check. RBI `00` = test IMEI (Wikipedia [Reporting Body Identifier](https://en.wikipedia.org/wiki/Reporting_Body_Identifier)); live RBIs: 01 CTIA/PTCRB (US), 35 TÜV SÜD/BABT (UK), 86 TAF (China), 99 GHA. For fake data use TAC `00xxxxxx` so it can never match a real handset — UNVERIFIED: third-party pages claim the GSMA TS.06 test-TAC shape is `00 44` + 4 digits; the TS.06 PDF could not be text-extracted here. |
+| ASIN | 10 chars `[A-Z0-9]{10}`; non-books start `B0` + 8 alphanumerics (`^B0[A-Z0-9]{8}$`); books reuse their ISBN-10 verbatim. No check digit (Wikipedia). |
+| SKU | Not standardised (Wikipedia). Convention only: 8–12 uppercase alphanumerics, e.g. `[A-Z]{3}-\d{4}-[A-Z]{2}`; avoid leading 0 and O/I if you want scanner-safe. |
+
+### GS1 prefixes
+
+- Official table: [gs1.org/standards/id-keys/company-prefix](https://www.gs1.org/standards/id-keys/company-prefix) (rendered via Playwright; WebFetch gets 403). Copyright notice ([gs1.org/terms-use](https://www.gs1.org/terms-use)): reproduction allowed only "in unaltered form … for your personal, non-commercial use or use within your organisation"; "You are not permitted to re-transmit, distribute or commercialise the information or material without seeking prior written approval from GS1." → Do not copy the GS1 page. The prefix-to-country mapping is factual and also on Wikipedia ([List of GS1 country codes](https://en.wikipedia.org/wiki/List_of_GS1_country_codes), CC BY-SA 4.0 — avoid verbatim copy). Safe path: embed only the US/special ranges you need, written from the facts below.
+- Ranges relevant to an en_US generator (from the GS1 page): 001–019, 030–039, 060–139 GS1 US (UPC-A compatible); 020–029 restricted circulation within a region; 040–049 restricted within a company; 050–059 GS1 US reserved; 200–299 restricted circulation (region); 952 "used for demonstrations and examples of the GS1 system"; 977 ISSN; 978–979 ISBN; 980 refund receipts; 981–983 coupons (common currency); 990–999 coupons. Prefixes do not identify country of origin. Selected others: 300–379 France, 400–440 Germany, 450–459 & 490–499 Japan, 500–509 UK, 690–699 China, 730–739 Sweden, 750 Mexico, 754–755 Canada, 760–769 Switzerland, 800–839 Italy, 840–849 Spain, 870–879 Netherlands, 880–881 South Korea, 890 India, 930–939 Australia, 940–949 New Zealand.
+- Generator rule: for fake UPC/EAN use prefix `952` (GS1 demo) or a `2xx` restricted-circulation prefix — never a real company prefix; for US-looking UPC-A use number system 0–1/6–8 with a random 5-digit manufacturer code and accept that it may collide with real products.
+
+## 12. Books, music, food
+
+### Project Gutenberg catalog
+
+- URL: `https://www.gutenberg.org/cache/epub/feeds/pg_catalog.csv` (20 MB) / `pg_catalog.csv.gz` (5.3 MB), regenerated weekly; also RDF (`rdf-files.tar.bz2`, 121 MB) and MARC.
+- Fields: `Text#, Type, Issued, Title, Language, Authors, Subjects, LoCC, Bookshelves` (comma CSV, quoted; Authors as `Last, First, birth-death`, `;`-separated).
+- Count (2026-09-17): 79 381 rows; Type: Text 78 130, Sound 1 114, Dataset 89, Image 33, other 12; Language: en 62 860, fr 4 190, fi 3 684, de 2 425, it 1 110.
+- Licence: each RDF record carries `<cc:Work><cc:license rdf:resource="https://creativecommons.org/publicdomain/zero/1.0/"/>` → catalog metadata is CC0 1.0 (verified on `cache/epub/1/pg1.rdf`). The CSV itself has no header notice; treat as CC0 by the same source. Terms of use: do not hammer their servers (bulk feeds are the sanctioned path), "Project Gutenberg" is a trademark — royalties for commercial use of the *name*; do not brand the data file with it beyond a source citation.
+- Embed: yes (CC0). For a fake-data library ship a filtered subset (Type=Text, Language=en, title + first author, maybe 5–10k rows) rather than 20 MB.
+
+### Music genres and instruments
+
+| Source | Licence | Embed? | Count | Notes |
+|---|---|---|---|---|
+| [MusicBrainz genre list](https://musicbrainz.org/genres) | Genre is a core entity (schema doc lists 13 core entities incl. Genre) → core data dump `mbdump.tar.bz2` is CC0. The genre→entity *associations* come via user tags (CC BY-NC-SA 3.0) — do not use those | Yes, names only | 2 202 (`/ws/2/genre/all?fmt=json`, `genre-count`) | WS API needs a descriptive User-Agent; list is a JSON array of `{id, name, disambiguation}` |
+| [Discogs data dumps](https://data.discogs.com/) | CC0 | Yes | genre/style vocab is embedded in release XML (monthly dumps, GBs) — no standalone style list; UNVERIFIED count (~15 genres, ~600 styles from memory) | Impractical to extract; prefer MusicBrainz |
+| ID3v1 genres | Names 0–79 from the 1999 ID3v1 spec, 80–125 Winamp, 126–191 Winamp 5.6 (2010) — a de-facto spec table, treated as public domain facts | Yes | 192 (0–191) | Wikipedia [List of ID3v1 genres](https://en.wikipedia.org/wiki/List_of_ID3v1_genres) (CC BY-SA page; the list itself is a spec enumeration). id3.org returned HTTP 500 during research — UNVERIFIED against the original spec page |
+| Wikidata instruments | CC0 | Yes | 9 194 items `P31/P279* Q34379` with English labels; 2 372 with a Hornbostel–Sachs number (P1762) | Filter by P1762 for a clean orchestral/folk instrument list |
+| Wikidata music genres | CC0 | Yes | 6 619 (`Q188451`) | Noisier than MusicBrainz |
+
+### USDA FoodData Central
+
+- Downloads: [fdc.nal.usda.gov/download-datasets](https://fdc.nal.usda.gov/download-datasets/) (April 2026 release):
+  - Foundation Foods: JSON 459 KB zip / 6.5 MB; CSV 3.7 MB zip / 32 MB — 394 foods (API `totalHits`, dataType=Foundation).
+  - SR Legacy (final, April 2018): JSON 12.3 MB zip / 205 MB; CSV 6.7 MB zip / 54 MB — 7 793 foods.
+  - FNDDS 2021-2023: CSV 200 MB zip / 1.6 GB — 5 432 foods.
+  - Branded: CSV 428 MB zip / 2.9 GB — 433 403 foods.
+  - Full: CSV 460 MB zip / 3.1 GB.
+- Licence ([API guide](https://fdc.nal.usda.gov/api-guide/)): "USDA FoodData Central data are in the public domain and they are not copyrighted. They are published under CC0 1.0 Universal (CC0 1.0)". No permission needed; USDA *requests* listing FoodData Central as source and notifying them. Suggested citation: "U.S. Department of Agriculture, Agricultural Research Service. FoodData Central, 2019. fdc.nal.usda.gov."
+- Embed: yes (CC0). Ship a derived list (SR Legacy `description` + `food_category`) not the raw dump.
+- CSV structure (UNVERIFIED — the field-description PDF could not be text-extracted; from memory): `food.csv` (fdc_id, data_type, description, food_category_id, publication_date), `food_category.csv` (id, code, description), `nutrient.csv` (id, name, unit_name, nutrient_nbr, rank), `food_nutrient.csv` (id, fdc_id, nutrient_id, amount, …), `food_portion.csv`, `measure_unit.csv`, `sr_legacy_food.csv` (fdc_id, NDB_number), `foundation_food.csv`.
+
+### Dishes, cuisines, drinks
+
+| Source | Licence | Embed? | Count | Notes |
+|---|---|---|---|---|
+| Wikidata dishes (`Q746549`) | CC0 | Yes | 7 480 with English label | SPARQL `?i wdt:P31/wdt:P279* wd:Q746549` |
+| Wikidata cuisines (`Q1968435`) | CC0 | Yes | 219 | |
+| Wikidata cocktails (`Q134768`) | CC0 | Yes | 305 | |
+| Wikidata beer (`Q44` subclass tree) / wine (`Q282`) | CC0 | Yes | 420 / 3 025 | Wine tree is mostly appellations/brands; filter by P279 depth |
+| IBA official cocktails | IBA site "© IBA 2026 – All rights reserved"; Wikipedia list CC BY-SA 4.0 | Names only are facts (102 cocktails: 34 Unforgettables, 34 Contemporary Classics, 34 New Era, 2024 revision); do not copy recipes/descriptions | 102 | Safest: take the names from Wikidata (`P31 Q134768`) rather than the IBA page |
+| [Open Brewery DB](https://github.com/openbrewerydb/openbrewerydb) | MIT (LICENSE, © 2025 Open Brewery DB) | Yes | 11 931 breweries (US 8 308, DE 1 445, AU 514, BE 478, CA 283, NZ 243) | `breweries.csv`: id (UUID), name, brewery_type (micro 5 906, brewpub 3 947, closed 644, planning 635, regional 240, contract 208, large 137, proprietor, taproom, bar, nano, cidery, beergarden), address_1..3, city, state_province, postal_code, country, phone, website_url, longitude, latitude. Real businesses — fine for names, but generating "fake" data with real addresses/phones may be undesirable |
+| Wikipedia lists (cuisines, dishes, IBA) | CC BY-SA 4.0 | No verbatim copies | | |
+
+## Summary of safe picks
+
+- Countries: `datasets/country-codes` (PDDL) or `annexare/Countries` (MIT); flag emoji derived from alpha-2.
+- Languages: `langs` / `iso-639-1` (MIT, 639-1 only).
+- Subdivisions: US-only list from Census (public domain); skip world list.
+- Commerce words: faker `en` lists (MIT). Check digits: implement from the rules above; GS1 prefix table: hand-write the few ranges needed, use `952`/`2xx` for generated barcodes; ISBN range file: do not embed.
+- Books: Gutenberg `pg_catalog.csv` (CC0), filtered.
+- Music: MusicBrainz genre names (CC0, 2 202), ID3v1 list (192), Wikidata instruments (CC0).
+- Food: USDA FDC SR Legacy/Foundation descriptions (CC0, cite USDA), Wikidata dishes/cuisines (CC0), Open Brewery DB (MIT).
+- Avoid: mledoze (ODbL), lukes (CC BY-SA), stefangabos (CC BY-SA/LGPL), Debian iso-codes (LGPL), SIL 639-3 table (no-redistribution clause), GS1/ISBN-International pages (all rights reserved), Wikipedia tables verbatim (CC BY-SA), MusicBrainz tag associations (CC BY-NC-SA).

docs/research/research-geo-se.md

@@ -0,0 +1,114 @@
+# Open data sources for real Swedish geographic and postal-address data
+
+Researched 2026-09-16. Goal: Sweden → län → kommun → tätort/postort → street, with real postal codes, embeddable in an MIT-licensed library.
+
+## TL;DR
+
+| Need | Best source | Licence | Embed in MIT repo? |
+|---|---|---|---|
+| Län + kommun names/codes | SCB `kommunlankod-2026.xlsx` | CC0 1.0 | Yes, no attribution needed |
+| Population per kommun (weighting) | SCB PxWeb API `BE0101A/BefolkningNy` | CC0 1.0 | Yes |
+| Tätorter (2,017) with code, kommun, län, population | SCB open geodata WFS/GeoPackage `stat:Tatorter_2023` + PxWeb `MI0810A/LandarealTatortN` | CC0 1.0 | Yes |
+| Streets + house numbers + postnummer + postort + kommun + coordinates (~3.9M points) | Lantmäteriet *Belägenhetsadress Nedladdning, vektor* (STAC, GeoPackage per kommun) | CC BY 4.0 **plus** personal-data terms, purpose review, Basic-auth download | **Not as-is** — raw redistribution is bound by the personal-data terms; a derived aggregate (street ↔ postnummer ↔ postort, no house numbers) is the realistic path but needs a stated purpose approved by Lantmäteriet. Legal review needed. |
+| postnummer → postort (list only) | GeoNames `SE.zip` (18,887 rows, 1,780 postorter) | CC BY 4.0 | Yes with attribution — but source is an undated third-party file, likely stale |
+| Street names per kommun/postort without Lantmäteriet | Trafikverket NVDB *Gatunamn* (Lastkajen) | CC0 1.0 | Yes, but no postnummer; needs account to download |
+| Street names | OpenStreetMap | ODbL 1.0 | Effectively no (share-alike on the extracted database) |
+| Municipal address files | OpenAddresses `sources/se/*` (16 kommuner) | Mixed: CC0 (Göteborg, Malmö), CC BY 4.0 (Helsingborg), unspecified (Stockholm 2016 dump, Uppsala) | Per source |
+| Postal code register (authoritative) | PostNord → Postnummerservice Norden AB / Geposit AB | Commercial; "resale/sublicensing not permitted" | No |
+
+## 1. Lantmäteriet
+
+Since 2025-02-03 the EU High Value Datasets (HVD) products are fee-free but Lantmäteriet explicitly says they are *not* "öppna data" because conditions attach ([Avgiftsfria produkter](https://www.lantmateriet.se/sv/geodata/vara-produkter/avgiftsfria-produkter/)). Two tiers exist:
+
+- **Öppna data** — licence "Creative Commons, CC0" ([Öppna data](https://www.lantmateriet.se/sv/geodata/vara-produkter/avgiftsfria-produkter/oppna-data/)). Verified on Geotorget product pages for *Topografi 50/100/250/1M Nedladdning, vektor* ("Villkor: Creative Commons, CC0", "Juridisk prövning: Nej", GeoPackage, download in Geotorget or via the *Geotorget Nedladdning* API).
+- **Värdefulla datamängder (HVD)** — licence **CC BY 4.0** per the terms document *Användningsvillkor för värdefulla datamängder* (DNR LM2025/009266 v1.0, 2025-02-01, [PDF](https://www.lantmateriet.se/globalassets/geodata/geodataprodukter/anvandningsvillkor_for_vardefulla_datamangder.pdf)). Products with personal data use *Användningsvillkor för värdefulla datamängder som innehåller personuppgifter* (DNR LM2025/009269 v1.1, 2025-03-04, [PDF](https://www.lantmateriet.se/globalassets/geodata/geodataprodukter/anvandningsvillkor_for_vardefulla_datamangder_pu.pdf)).
+
+Per-product status (Geotorget product pages, rendered 2026-09-16):
+
+| Product | Fee | Terms | Legal review | Access | Format |
+|---|---|---|---|---|---|
+| [Belägenhetsadress Nedladdning, vektor](https://geotorget.lantmateriet.se/geodataprodukter/belagenhetsadress-nedladdning-vektor-api) | 0 | HVD terms **with personal data** | **Yes** | STAC API `https://api.lantmateriet.se/stac-vektor/v1`, collection `belagenhetsadresser` | GeoPackage, one zip per kommun |
+| [Belägenhetsadress Direkt](https://geotorget.lantmateriet.se/geodataprodukter/belagenhetsadress-direkt-api) | 0 | HVD terms with personal data | Yes | REST `https://api.lantmateriet.se/distribution/produkter/belagenhetsadress/v4.2` (OAuth2 via API portal) | JSON / XML |
+| [Belägenhetsadress Nedladdning, Inspire](https://www.lantmateriet.se/sv/geodata/vara-produkter/produktlista/belagenhetsadress-nedladdning-inspire/) | 0 | HVD terms with personal data | Yes (3–5 working days extra for security review) | Atom feed, login | GML; SWEREF 99 TM / ETRS89; semi-annual |
+| [Ortnamn Nedladdning, vektor](https://geotorget.lantmateriet.se/geodataprodukter/ortnamn-nedladdning-vektor-api) | 0 | HVD terms (no personal data) | No | STAC collection `ortnamn`, one national zip (58.5 MB) | GeoPackage |
+| [Ortnamn Direkt](https://geotorget.lantmateriet.se/geodataprodukter/ortnamn-direkt-api) | 0 | HVD terms | No | REST `…/distribution/produkter/ortnamn/v2.2` | JSON / XML |
+| Kommun, län och rike (administrativ indelning) | 0 | STAC says `CC-BY-4.0`; a Lantmäteriet HVD presentation (GISS, 2025-03-04) tables it as **CC0** ("CC0 används som licens för tjänster som är gemensamma för HVD och NGP … Kommun, Län och Rike") — **conflicting, unverified which prevails** | No | STAC collection `kommun-lan-rike` (zip ~6 MB, yearly + "aktuell"); OGC API Features `https://api.lantmateriet.se/ogc-features/v1/administrativ-indelning` (collections `kommuner`, `lan`, `rike`, plus `-2025`/`-2026`; items endpoint returns 401) | GeoPackage / GeoJSON |
+| [Topografi 10 Nedladdning, vektor](https://geotorget.lantmateriet.se/geodataprodukter/topografi-10-nedladdning-vektor) | 0 | HVD terms with personal data | **Yes** | Geotorget download / API | GeoPackage |
+| Topografi 50 / 100 / 250 / 1M Nedladdning, vektor | 0 | **CC0** | No | Geotorget download / *Geotorget Nedladdning* API | GeoPackage |
+
+Notes:
+- *Administrativ indelning* is not a standalone Geotorget product; it is the STAC collection `kommun-lan-rike`, the OGC API above, an [Inspire download](https://www.lantmateriet.se/sv/geodata/vara-produkter/produktlista/administrativ-indelning-nedladdning-inspire/) (Atom/WCS), or a layer in Topografi 100/250/1M (CC0). Topografi 1M also carries "tätortsgränser och tätortssymboler".
+- The STAC catalog root and collection/item listings are public (HTTP 200, no auth); it self-describes: "avgiftsfria och får användas enligt creative commons licens CC BY 4.0. För vissa informationsmängder kommer din användning att prövas juridiskt … och du behöver då godkänna särskilda användningsvillkor." Asset downloads on `dl1.lantmateriet.se` return **401 Basic** — credentials are the Geotorget account (private person) or a system account (organisation). Private accounts are "endast för privat bruk" ([konto-privatperson](https://geotorget.lantmateriet.se/konto-privatperson)); an MIT library is not private use, so apply as an organisation.
+- **Record count.** 290 STAC items (one per kommun), total 324 MB zipped. Smallest Bjurholm 143 KB, largest Stockholm 9.1 MB, Göteborg 8.9 MB, Malmö 4.9 MB. Fastighetsregistret held 3,882,361 valid addresses in 2022 (2,886,796 street addresses + 964,737 rural) per [SCB/LM presentation](https://kartografiska.se/wp-content/uploads/5C_Foretagsadresser_SCB_LM.pdf); the OSM import thread cites 3.7 M. So ~3.9 M address points, not 4–5 M.
+- **Fields** (JSON schema `belagenhetsadress-4.2.2.json`, matches *Nationell specifikation Adress* v1.0.1): `kommunkod`, `kommunnamn`, `kommundel` (geografisk kommundel), `adressomrade` + `adressomradestyp` (gatuadressområde / metertalsadressområde / byadressområde), `gardsadressomrade`, `adressplatsnummer`, `bokstavstillagg`, `lagestillagg`/`lagestillaggsnummer`, `adressplatstyp`, `adressplatsbeteckning`, `postnummer`, `postort` (both "beslutas av PostNord AB", only on current addresses), `popularnamn`, `distriktskod`/`distriktsnamn`, `adressattAnlaggning`, `anmarkningstyp`/`anmarkningstext`, `registerenhetsreferens` (property link), `objektidentitet`, `objektstatus`, `versionGiltigFran`, `geometri` (SWEREF 99 TM point). No län field — derive from kommunkod (first two digits).
+- Lantmäteriet treats addresses as personal data via Fastighetsregisterlagen § 6: usage only for the purpose approved in the application, EU/EEA-only storage, spot checks, revocation (terms PU §§ 3.3–4.2). The OSM community reports Lantmäteriet refused ODbL compatibility and OpenAddresses ([#7657](https://github.com/openaddresses/openaddresses/issues/7657), [#7608](https://github.com/openaddresses/openaddresses/issues/7608)) has not been able to include it; the OSM thread ([127427](https://community.openstreetmap.org/t/lantmateriet-belagenhetsadress/127427), 91 posts, last 2026-07-25) says one mapper got approval for a full 3.7 M import but no formal import has happened.
+- **Attribution text required (CC BY 4.0 + terms § 3.1):** product name as data source, "©Lantmäteriet", whether the data was processed, and that CC BY 4.0 applies — may be placed in accompanying documentation/metadata.
+- Ortnamn: name types include `Tätort` ("minst 200 invånare"), Bebyggelse, Trakt, Kyrka, Anläggning, nature names; Ortnamnsregistret holds ~980,000 names (Lantmäteriet, figure not re-verified). Not delivered as CSV.
+
+## 2. SCB (Statistics Sweden)
+
+- **Licence.** Statistics and geodata published as open data: "Creative Commons 0 1.0 Universal, CC0" ([användningsvillkor](https://www.scb.se/om-scb/om-scb.se-och-anvandningsvillkor)); since 2021-07-01. Optional credit "Källa: SCB" / "Source: Statistics Sweden"; do not cite SCB as source for data you have processed. API limit 10 requests / 10 s per IP.
+- **Län and kommuner with codes.** 21 län, 290 kommuner ([lan-och-kommuner](https://www.scb.se/hitta-statistik/regional-statistik-och-kartor/regionala-indelningar/lan-och-kommuner/)). Download: `https://www.scb.se/contentassets/7a89e48960f741e08918e489ea36354a/kommunlankod-2026.xlsx` (18 KB, HTTP 200). Codes: län 2 digits (01 Stockholm … 25 Norrbotten), kommun 4 digits = län + 2 (0180 Stockholm).
+- **Population per kommun.** PxWeb API `https://api.scb.se/OV0104/v1/doris/sv/ssd/START/BE/BE0101/BE0101A/BefolkningNy` — region dimension has 312 values (riket, 21 län, 290 kommuner), years 1968–2024 (verified 2026-09-16). Also `BE0101C/BefArealTathetKon` (density, 1991–2025).
+- **Tätorter.** 2,017 tätorter in 2023 (WFS `resultType=hits` on `stat:Tatorter_2023` → `numberOfFeatures="2017"`; SCB also states 2,017). Open geodata: `https://geodata.scb.se/geoserver/stat/wfs` (WFS/WMS) and GeoPackage download ([statistiska tätorter](https://www.scb.se/vara-tjanster/oppna-data/oppna-geodata/statistiska-tatorter/)). Attributes: `tatortskod`, `tatort`, `kommun`, `kommunnamn`, `lan`, `lannamn`, `area_ha`, `bef` (population), `ar`. Population per tätort also in PxWeb `MI0810/MI0810A/LandarealTatortN` (2,827 region codes, every 5 years to 2023). Also `Smaorter_2023`, `DeSO_2025`, `RegSO_2025`.
+
+## 3. Postal codes
+
+- **Owner.** PostNord Sverige owns and operates the postnummer system; PTS is supervisory authority; a Postnummerråd approves changes 3–4 times a year (sv.wikipedia). Register management is delegated to **Postnummerservice Norden AB**, machine access via **Geposit AB**. Postnummerservice's [terms](https://postnummerservice.se/en/information/kopvillkor): "free use within your own organization", "resale of data is not permitted", "sublicensing to third parties is not permitted" — unusable for an open library.
+- **No open list from PostNord, Lantmäteriet or Digg.** dataportal.se search for "postnummer" (2026-09-16) yields only Postnummerservice's own commercial catalogue entry ("Svenska postnummer och postorter", published 2012), Helsingborg's municipal postnummer polygons, and Lantmäteriet's Inspire address entry. The old dataportal community called the postnummer system "slarvats bort till ett privat bolag". GitHub CSVs (`zegl/sweden-zipcode`, `lapplandi/sveriges-postnummer`, `beshrkayali/sverige_postnummer`) state no licence or provenance — do not use.
+- **Lantmäteriet addresses carry `postnummer` + `postort`** on every current address point (set by PostNord), so the address product is the only lawful open-ish route to a *complete, current* postnummer ↔ postort ↔ street mapping.
+- **GeoNames `SE.zip`** ([readme](https://download.geonames.org/export/zip/readme.txt)): CC BY 4.0 ("a link on your website to www.geonames.org is ok"). Downloaded and counted: **18,887 rows, 18,887 distinct postnummer, 1,780 postorter, 22 admin1 values (21 län + blank), 285 admin2 (kommun code) values; 3,055 rows lack kommun; accuracy: 16,115 rows = 4 (centroid of postal code area), 389 = 1, 2,383 blank**. Fields: country, postal code (`111 64` with space), place name, admin1 name/code (län), admin2 name/code (kommun 4-digit), admin3, lat, lon, accuracy. The GeoNames sources page lists Sweden's source as `www.pellesoft.se/upload/program/prg00745.zip` with **no date** (URL now dead) — a hobbyist upload, so freshness is unknown; the row count (18.9 k) exceeds today's ~17 k total codes, indicating retired codes are still present. The zip is rebuilt daily (Last-Modified 2026-09-16) but that says nothing about the underlying data. 96 KB zipped.
+- **Counts.** Postnummerservice: "~17 000 postal codes", "10 500 deliverable postal codes", "1 743 postal cities"; Kartanalys: 10,839 geographic postnummer, 1,740 postorter; Wikipedia: >16,100 areas (2008). Changes quarterly.
+- **Structure** (sv/en Wikipedia): 5 digits written `NNN NN`. Lower numbers further south, except 1xx xx = Stockholm. First digit/tens: 10–19 Stockholm, 20–29 Skåne (Malmö 20–21), 30–39 southern Sweden, 40–49 Göteborg area (Göteborg 40–41), 50–59, 60–69, 70–79, 80–89, 90–99 progressively north to Norrbotten (98x). Postorter come in 2-, 3- or 5-position sizes; in two-position places the 3rd digit (in three-position the 4th) marks delivery type: 0 = boxes/postal, 1 = boxes/business, 2–4 and 6–7 = street delivery, 5 = rural (lantbrevbäring), 8 = reply mail, 9 = competitions/temporary — with exceptions in big cities. Box codes are therefore not street-deliverable and should be excluded when generating street addresses.
+
+## 4. Streets
+
+- **Lantmäteriet** — see § 1; `adressomrade` with `adressomradestyp = gatuadressområde` is the street name.
+- **Trafikverket NVDB** — CC0 1.0 Universal since 2017-04-03 ([trafikverket.se](https://www.trafikverket.se/e-tjanster/hamta-data-fran-trafikverket/) links `creativecommons.org/publicdomain/zero/1.0/deed.sv`; [OSM wiki](https://wiki.openstreetmap.org/wiki/Sweden/trafikverket)). Data product **Gatunamn** = "det officiellt adressbildande namnet på gatan" (municipal decision); Trafikverket is cleaning it in 2026 so it holds only names used in street addresses ([Vägnamn i NVDB](https://www.nvdb.se/sv/aktuellt/nyhetsarkiv/2026/vagnamn-i-nvdb/)). Download via Lastkajen (registration with e-mail + accept licence; Shapefile/GeoPackage), or the Datautbytesportal API. Gives street ↔ kommun (and geometry) but **no postnummer/postort**. Record count not published on the pages checked.
+- **OpenStreetMap** — ODbL 1.0. A list of street names extracted from OSM is a Derivative Database: "Where you make our data or any Derivative Database available to others, it must continue to be licensed under the ODbL" ([OSMF FAQ](https://osmfoundation.org/wiki/Licence/Licence_and_Legal_FAQ)); only *Produced Works* (maps etc.) escape share-alike. Embedding an OSM-derived JSON in an MIT repo therefore forces that data file to ODbL with attribution "© OpenStreetMap contributors" + link to openstreetmap.org/copyright, and downstream consumers inherit share-alike on the data. Not recommended.
+- **OpenAddresses** — 16 Swedish sources, all municipal (`sources/se/municipality_of_*.json`: Alingsås, Gislaved, Göteborg, Helsingborg, Höganäs, Kalmar, Kristinehamn, Malmö, Nacka, Sävsjö, Stockholm, Uppsala, Västerås, Vaxholm, Växjö, Österåker). OA does not relicense; licence is per source. Checked: Göteborg **CC0 1.0** (attribution "Göteborgs stad", shapefile, `gatunamn`), Malmö **CC0 1.0** (CSV with `ADRESSOMR`, `ADRESSPLAT`, `POSTNR`, `POSTORT`, cached 2024-11), Helsingborg **CC BY 4.0** ("Helsingborgs stad", GeoJSON with Gatunamn/Postnummer/Stad), Uppsala attribution required, licence name not stated (ArcGIS REST), Stockholm **no licence field**, data is a 2016 cache of a defunct WS. Usable for a few big cities only; no national coverage.
+- **Unique street-name counts.** SCB (Lägenhetsregistret 2019, [artikel](https://www.scb.se/hitta-statistik/artiklar/2021/pa-ringvagen-bor-det-flest/)): "Antalet gator eller vägar med en eller flera adresser är 399 175 i hela landet" — this is streets counted per place, not unique names. Most frequent: Ringvägen 205, Skogsvägen 204, Björkvägen 200, Skolgatan 196, Storgatan 180 (in 180 kommuner, 55,360 residents). A national unique-name count was not found; expect well under 399 k (probably 100–200 k) — **unverified**.
+
+## 5. Sizes
+
+| Level | Count | Source |
+|---|---|---|
+| Län | 21 | SCB |
+| Kommuner | 290 | SCB; 290 STAC items |
+| Tätorter (2023) | 2,017 | SCB WFS hits |
+| Postorter | ~1,740–1,780 | Postnummerservice 1,743; Kartanalys 1,740; GeoNames 1,780 |
+| Postnummer | ~17,000 total, ~10,500–10,840 deliverable/geographic | Postnummerservice, Kartanalys |
+| Streets with addresses (per place) | 399,175 (2019) | SCB |
+| Unique street names | not published; likely 100–200 k | unverified |
+| Address points | ~3.9 M (3,882,361 in 2022) | Lantmäteriet/SCB |
+| Lantmäteriet address GeoPackages | 324 MB zipped, 290 files | STAC |
+| GeoNames SE.zip | 96 KB (18,887 rows) | download |
+| Ortnamn national GeoPackage | 58.5 MB zipped; ~980 k names | STAC; Lantmäteriet |
+
+Embedding suggestion (size-driven, before licence): län + kommun + population (~20 KB) and tätorter with population (~100 KB) are trivially embeddable under CC0. A postnummer→postort table is ~300 KB raw. A national (street, postort, postnummer-range) table at ~400 k rows is roughly 15–25 MB JSON, ~3–5 MB gzipped — optional download pack territory; a default pack could carry the top N streets per kommun (e.g. 20 × 290 ≈ 6 k rows, <300 KB).
+
+## 6. Licence compatibility with an MIT repo
+
+| Source | Licence | Embed? | Required notice |
+|---|---|---|---|
+| SCB codes, population, tätorter | CC0 1.0 | Yes | None; optional "Källa: SCB" |
+| Lantmäteriet Topografi 50/100/250/1M (admin boundaries, tätort symbols) | CC0 | Yes | None (courtesy credit suggested) |
+| Lantmäteriet Ortnamn | CC BY 4.0 + HVD terms (no personal data, no review) | Yes, data file stays CC BY 4.0 inside the MIT repo | "Källa: Ortnamn Nedladdning, vektor, ©Lantmäteriet, bearbetad, CC BY 4.0" (product name, ©Lantmäteriet, processed-flag, CC BY 4.0) |
+| Lantmäteriet kommun-län-rike | CC BY 4.0 per STAC (CC0 per LM slide — unresolved) | Yes | Same attribution form as above unless CC0 is confirmed |
+| Lantmäteriet Belägenhetsadress | CC BY 4.0 + personal-data terms, purpose approval, EU-only storage, revocable | **Only with an approved application** whose stated purpose covers deriving and publishing an aggregated street/postnummer/postort list under CC BY 4.0. Raw points (house numbers, property links) should not be redistributed. Get written confirmation from geodatasupport@lm.se; treat as unresolved until then. | Product name, ©Lantmäteriet, "bearbetad", CC BY 4.0 |
+| GeoNames SE | CC BY 4.0 | Yes | Link/credit to www.geonames.org; note staleness |
+| Trafikverket NVDB Gatunamn | CC0 1.0 | Yes | None |
+| OpenStreetMap | ODbL 1.0 | Not practical (share-alike on the data) | "© OpenStreetMap contributors" + ODbL if ever used |
+| OpenAddresses Göteborg, Malmö | CC0 1.0 | Yes | None (Göteborg asks credit "Göteborgs stad") |
+| OpenAddresses Helsingborg | CC BY 4.0 | Yes | "Helsingborgs stad" |
+| OpenAddresses Stockholm/Uppsala | unspecified | No | — |
+| PostNord/Postnummerservice files | commercial, no redistribution | No | — |
+
+CC BY 4.0 and CC0 data files can sit in an MIT repo; the MIT licence covers the code, and a `DATA-LICENSES.md`/NOTICE lists each dataset with its licence and the attribution above. CC BY 4.0 § 2(a)(5)(B) (no technical protection measures) is irrelevant for a plain JSON file.
+
+## Could not verify
+- Whether the kommun-län-rike collection is CC0 or CC BY 4.0 (sources disagree).
+- Whether Lantmäteriet will approve an application whose purpose is publishing a derived open dataset; OSM/OpenAddresses experience suggests refusal is likely.
+- GeoNames Sweden source date; assume stale.
+- National count of unique street names; Gatunamn (NVDB) record count.
+- Exact number of active postnummer today (three sources: ~17 k total / 10.5–10.8 k deliverable).

docs/research/research-geo-world.md

@@ -0,0 +1,208 @@
+# Open geographic / postal-address data outside Sweden — for fejkdata (MIT)
+
+Researched 2026-09-16. "Embeddable in MIT repo" below means: the data licence permits redistribution and commercial use with at most an attribution notice, so the data can ship inside an MIT-licensed package with a separate `DATA-LICENSES` file. Anything share-alike (ODbL, CC BY-SA) is flagged as **not** embeddable. Items marked *(unverified)* were not confirmed against a primary source.
+
+## 0. Summary table
+
+| Country | Official open address register | Licence | Addresses | Postcode in register | Postcode data open? | Embed in MIT? |
+|---|---|---|---|---|---|---|
+| US | DOT National Address Database (NAD) | US public domain | ~80 M (Jun 2026) | ZIP (optional field; fill rate *unverified*) | ZCTA yes (public domain); USPS publishes no open ZIP file | Yes |
+| Norway | Kartverket Matrikkelen – Adresse | CC BY 4.0 | 2.58 M (Jun 2024) | Yes (postnummer + poststed) | Yes: Bring postnummerregister, NLOD 2.0 | Yes |
+| Denmark | Danmarks Adresseregister (DAR) via DAWA / Datafordeler | CC BY 4.0 | ~3.8 M | Yes | Yes (in DAR; 1,089 postnumre) | Yes |
+| Finland | DVV building addresses (final release Feb 2025) + Posti PCF/BAF | CC BY 4.0 / Posti terms (free; redistribute with terms attached) | ~5 M buildings | Yes | Yes (Posti PCF/BAF) | Yes (Posti: attach terms) |
+| UK (GB) | No open *address* register. OS Open UPRN (points only), OS Open Names (streets/places/postcodes), Code-Point Open (postcodes) | OGL v3 | 40 M UPRNs (no addresses); 870 k roads; 1.7 M postcodes | n/a | Yes for GB (Code-Point Open, OGL); Northern Ireland restricted | Yes (GB only; no house numbers) |
+| Germany | No national register. 15 Länder publish Hauskoordinaten/Gebäudereferenzen (dl-de/by-2-0, dl-de/zero-2-0); Bavaria missing | mixed, mostly permissive | ~19 M points across Länder | Yes in Hauskoordinaten | **No** — Deutsche Post DATAFACTORY is proprietary; open PLZ lists are OSM-derived (ODbL) | Partly (Länder files yes; a national PLZ list is the problem) |
+| Netherlands | BAG (Kadaster/PDOK) | CC0 1.0 | ~9.9 M | Yes | Yes (in BAG) | Yes |
+| France | Base Adresse Nationale (BAN) | Licence Ouverte / Etalab 2.0 | >25 M | Yes | Yes (BAN; La Poste "base officielle des codes postaux", open licence) | Yes |
+| Spain | Catastro INSPIRE Addresses (AD) + INE callejero | Catastro own licence (free, attribution) / INE legal notice | ~15.7 M points (mailwoman count) | Yes (AD:PostCode) | Correos' postcode layer closed since 2017; postcodes still appear per address in Catastro AD | Probably (read Catastro licence PDF first) |
+| Australia | G-NAF (Geoscape via data.gov.au) | EULA based on CC BY 4.0 + "no mail-out" restriction | 15.95 M (Aug 2026) | Yes | In G-NAF; Australia Post file is non-commercial only | Yes (with EULA notice) |
+| Canada | StatCan National Address Register (NAR) 2024; ODA v1 (2021) | StatCan Open Licence / OGL-Canada | 17.1 M / ~10 M | NAR: MAIL_POSTAL_CODE "for most addresses" | Canada Post's postal code file is **proprietary**; only what NAR/ODA carry | Yes (NAR); postcode completeness unverified |
+
+## 1. Per-country detail
+
+### United States
+
+**Official address register — National Address Database (NAD), US DOT**
+- Licence: US public domain (`http://www.usa.gov/publicdomain/label/1.0/`), stated on data.gov: https://catalog.data.gov/dataset/national-address-database-nad-text-file. No attribution required. Embeddable.
+- Size: ~80 M records; current release compiled 2026-06-30, dataset updated 2026-09-03. Sources: https://data.transportation.gov/dataset/National-Address-Database-NAD-Text-File/fc2s-wawr/about_data, https://www.placekey.io/datasets/national-address-database (80 M; "24 fields" in the text export).
+- Coverage: aggregated from state/local providers; a mix of fully, partially and non-participating states (placekey page; the DOT page https://www.transportation.gov/gis/national-address-database blocks scrapers, so the coverage map was not read).
+- Formats: zipped flat text and File Geodatabase. The text file is ordered by spatial cluster, not by OID.
+- Fields (NAD schema https://www.transportation.gov/sites/dot.gov/files/2023-07/NAD_Schema_202304.pdf, blocked for bots; list taken from the FGDC content-requirements doc https://www.fgdc.gov/organization/working-groups-subcommittees/address-sc/220810-nad-content-requirements-approved.pdf): address number parts (`AddNum_Pre`, `Add_Number`, `AddNum_Suf`), street name parts (`St_PreDir`, `St_PreTyp`, `St_Name`, `St_PosTyp`, `St_PosDir`, `StNam_Full`), sub-address (`Building`, `Floor`, `Unit`, `Room`), place names (`Inc_Muni`, `Post_City`, `Census_Plc`, `Uninc_Comm`), `County`, `State`, `Zip_Code`, `Plus_4`, `Longitude`, `Latitude`, `NatGrid`, `Addr_Type`, `Placement`, `NAD_Source`, `DataSet_ID`. FGDC: "A Zip Code is recommended but not mandatory"; a complete place name and state are mandatory. Expect some records without ZIP.
+- Census Address Count Listing files give housing-unit counts per block, not addresses: https://census.gov/geographies/reference-files/2025/geo/addcountlisting.html. The Census MAF itself is not public.
+
+**Administrative hierarchy — Census Bureau (public domain)**
+- States: 50 + DC + PR + 4 island areas = 56 FIPS state codes ("States & Equivalents: 56" in the 2020 tallies). ANSI list: https://www.census.gov/library/reference/code-lists/ansi/ansi-codes-for-states.html (also carries FM, MH, PW, UM).
+- Counties: 3,144 in the 50 states + DC; 3,234 including PR and island areas (2020 tallies https://www.census.gov/geographies/reference-files/time-series/geo/tallies.html; https://en.wikipedia.org/wiki/List_of_United_States_counties_and_county_equivalents). Connecticut switched to 9 planning regions as county equivalents in 2022 — use a current vintage.
+- Places: 19,734 incorporated places + 12,454 CDPs (2020 tallies, incl. PR/island areas); "approximately 19,500" incorporated places in the 50 states (https://www.census.gov/library/stories/2020/05/america-a-nation-of-small-towns.html).
+- TIGER/Line 2025 shapefiles and the pipe-delimited **Gazetteer files** (2026 vintage: states, counties, county subdivisions, places, ZCTAs; GEOID, name, area, centroid lat/lon; **no population**): https://www.census.gov/geographies/reference-files/time-series/geo/gazetteer-files.html. Gazetteer excludes island areas, includes PR.
+- Population per place: **SUB-EST2025** CSVs (incorporated places + MCDs, April 2020 base through July 2025, released May 2026): https://www.census.gov/data/tables/time-series/demo/popest/2020s-total-cities-and-towns.html. Public domain. The right file for population-weighted draws.
+
+**Postal codes**
+- USPS publishes no open ZIP list file; counts quoted 41,541–41,695 active ZIPs (https://facts.usps.com/42000-zip-codes/ says 41,554). Commercial "USPS-licensed" databases exist.
+- Census **ZCTA**: 33,791 ZCTAs (2020), public domain, in TIGER + Gazetteer + relationship files (ZCTA↔place, ZCTA↔county): https://www.census.gov/programs-surveys/geography/technical-documentation/records-layout/2020-zcta-record-layout.html. PO-box-only ZIPs have no ZCTA (https://en.wikipedia.org/wiki/ZIP_Code_Tabulation_Area).
+- GeoNames `US.txt` (CC BY 4.0): ZIP → place name, state, county, lat/lon. Count not stated on the site *(≈41 k, unverified)*.
+
+**Street names**
+- TIGER/Line `featnames` relationship files (per county) hold every feature name with prefix/suffix parts; joined to `edges` and `addrfeat` (address ranges). Public domain. No published count of unique street names *(unverified; must be computed)*.
+- NAD `St_Name`/`StNam_Full` distinct per `Post_City` is the cheaper source for "real streets in Denver".
+
+**Sizes for the embed-vs-download decision (US)**
+
+| Level | Count | Source |
+|---|---|---|
+| States + DC + PR + island areas | 56 | Census tallies 2020 |
+| Counties / equivalents | 3,144 (50+DC) / 3,234 (with PR + islands) | Census tallies, Wikipedia |
+| Incorporated places | 19,734 (all) / ~19,500 (50+DC) | Census tallies |
+| CDPs | 12,454 | Census tallies |
+| ZCTAs / USPS ZIPs | 33,791 / ~41,500 | Census tallies / USPS facts |
+| Address points (NAD) | ~80 M | DOT |
+| OpenAddresses US collections | NE 2.35 GB + South 7.11 GB + West 8.54 GB + Midwest 4.86 GB (zipped CSV); 2,381 US sources | https://batch.openaddresses.io/ (2026-09) |
+| Unique street names | unknown; compute from NAD/TIGER | — |
+
+### Norway
+- **Matrikkelen – Adresse** (Kartverket). Licence CC BY 4.0 (https://data.norge.no/en/datasets/8f1151bf-ad63-3e47-adbb-143f876776ee/matrikkelen-adresse; Geonorge register https://register.geonorge.no/inspire-statusregister/matrikkelen-adresse/73bc0329-faac-419b-80ff-d113c0ffe6a0). Formats: CSV, SOSI, GML, SQL/PostGIS, FGDB, Atom feed, REST API `https://api.kartverket.no/adresser/v1/`. Files per kommune (daily) and fylke/country (weekly).
+- Counts (Kartverket "Antall vegadresser og matrikkeladresser", extract 2024-06-01, https://www.kartverket.no/datakvalitet/rapporter/adresse/Antall%20vegadresser%20og%20matrikkeladresser.xlsx): **2,577,825 addresses** — 2,536,663 vegadresser (street addresses) + 41,162 matrikkeladresser (cadastral, no street name).
+- Fields (CSV): adressenavn (street), nummer, bokstav, postnummer, poststed, kommunenummer, kommunenavn, grunnkrets, tettsted, coordinates (EPSG:25833 / 4258); bruksenhetsnummer only in non-CSV formats. Fylke = first two digits of kommunenummer.
+- Hierarchy: 15 fylker + Svalbard, 357 kommuner (2024). Kartverket/SSB code lists are open *(SSB Klass API licence NLOD, unverified)*.
+- Postal codes: Bring/Posten **Postnummerregister** — postnummer, poststed, kommunenummer, kommunenavn, kategori; TAB text + xlsx; **NLOD 2.0** per data.norge.no (https://data.norge.no/en/datasets/5e6847ba-156d-4e14-85d3-8d7f8b727523/postnummer-i-norge). Updated annually (Sept/Oct). Embeddable.
+- OpenAddresses: 17 `no/` sources.
+
+### Denmark
+- **Danmarks Adresseregister (DAR)**, served by **DAWA** (`https://api.dataforsyningen.dk/`) and Datafordeler. Licence **CC BY 4.0**, credit "Klimadatastyrelsen" (https://datafordeler.dk/vejledning/brugervilkaar/danmarks-adresseregister-dar/). Embeddable.
+- Counts: ~3.8 M adresser (Klimadatastyrelsen DAWA report); DAWA today: **1,089 postnumre, 53,976 vejnavne, 99 kommuner, 5 regioner** (queried 2026-09-16 via `?format=csv`).
+- Fields (DAWA `adresser`): vejnavn, husnr, etage, dør, postnr, postnrnavn, supplerende bynavn, kommunekode/-navn, regionskode/-navn, sogn, koordinater (ETRS89/WGS84), DAR ids. Downloads: CSV, JSON, GeoJSON, per kommune or whole country; `adgangsadresser` = access address (building entrance, no floor/door). Docs: https://dawadocs.dataforsyningen.dk/dok/adresser.
+- Hierarchy: 5 regioner → 98 kommuner (99 rows incl. Christiansø) → postnumre/byer. All in DAWA.
+- Postcodes: in DAR (source PostNord, updated on change), same licence.
+- OpenAddresses: `dk/countrywide` (1 source).
+
+### Finland
+- No single address-register download today. Sources:
+  - **DVV (Digital and Population Data Services Agency) building address open data** — CC BY 4.0; final official release 2025-02-14, distribution ended 2025-03-14; archived copies (CSV/GeoPackage, ~5 M buildings: street fi/sv, house number, postal code, municipality, WGS84) at https://markuskainu.fi/posts/2025-02-03-dvv-rakennusten-osoitetiedot/. Successor: **SYKE/Ryhti building addresses** via OGC API Features, CC BY 4.0 (https://avoindata.suomi.fi/data/fi/dataset/syke-rakennusten-osoitteet-ogc-api-features) — page blocked the fetch; verify fields and bulk download before relying on it.
+  - **NLS (Maanmittauslaitos)** topographic database address points — CC BY 4.0 (https://www.maanmittauslaitos.fi/en/opendata-licence-cc40); address points computed from road-name data.
+  - **Digiroad** (Fintraffic since 2026-01-01) — CC BY 4.0; road names + address number ranges (https://vayla.fi/en/transport-network/data/digiroad).
+- **Posti postal code files** (https://www.posti.fi/en/for-businesses/customer-support/postal-code-services; directory https://www.posti.fi/webpcode/): **PCF** (postal code, names fi/sv + abbreviations, type code, region code/name, municipality code/name fi/sv, language code; daily), **BAF** (every street per postal code with odd/even building-number ranges + municipality; weekly; excludes Åland), **POM** (monthly changes). Fixed-width `.dat`, Unix LF. Terms (service description 2015-01-01): "The files are freely downloadable … The data can be disclosed to third parties but it must be ensured that the recipient of the data is aware of the terms of use for the service as well as the download date." No licence name, no fee, no commercial restriction stated. Embeddable if the terms PDF + download date ship with the data.
+- Hierarchy: Statistics Finland classifications, CC BY 4.0: Municipalities 2026 https://stat.fi/en/luokitukset/kunta/kunta_1_20260101, Regions 2026 https://stat.fi/en/luokitukset/maakunta/maakunta_1_20260101. Posti PCF also carries municipality + region codes.
+- OpenAddresses: `fi/countrywide-fi`, `fi/countrywide-sv` + 38 municipal sources.
+
+### United Kingdom (Great Britain; NI separate)
+- No open full-address register (AddressBase is paid). Open pieces, all **OGL v3**, attribution "Contains OS data © Crown copyright and database right [year]" (https://wiki.openstreetmap.org/wiki/Ordnance_Survey_OpenData_Licence; OS moved to OGL v3 in 2015):
+  - **OS Open Names** — "over 870 000 named and numbered roads, nearly 44 000 settlements and over 1.6 million postcodes"; 34 attributes incl. `name1`, `name2` (Welsh/Gaelic), `type`, `local_type` (populatedPlace: City/Town/Village/Hamlet/Suburban Area/Other Settlement; transportNetwork: Named Road, Numbered Road, Section Of …; other: Postcode), `postcode_district`, `populated_place`, `district_borough`, `county_unitary`, `region`, `country`, `GEOMETRY_X/Y`. CSV/GML/GeoPackage, quarterly. https://docs.os.uk/os-downloads/products/addresses-and-names-portfolio/os-open-names. Gives street → settlement → district → county → region → country, but **no house numbers** and only the postcode district per road.
+  - **Code-Point Open** — ~1.7 M GB postcode units; fields Postcode, Positional_quality_indicator, Eastings, Northings, Country_code, NHS codes, Admin_county_code, Admin_district_code, Admin_ward_code. No street or locality names. GB only; NI excluded. https://docs.os.uk/os-downloads/products/areas-and-zones-portfolio/code-point-open.
+  - **OS Open UPRN** — ~40 M UPRN + coordinates, no addresses. https://www.ordnancesurvey.co.uk/products/os-open-uprn.
+  - **ONS Postcode Directory (ONSPD)** — postcode → all admin geographies; OGL, but requires Royal Mail attribution ("Contains Royal Mail data © Royal Mail copyright and database right [year]") and NI (BT) postcodes are end-user-licence only (no commercial redistribution): https://www.ons.gov.uk/methodology/geography/licences. Drop NI rows.
+- Hierarchy codes: ONS GSS codes (countries, regions, counties/UAs, districts, wards) via the ONS Open Geography Portal, OGL. Population: ONS mid-year estimates (OGL) at local-authority level; settlement population must come from GeoNames/Wikidata.
+- OpenAddresses: 0 `gb/` sources.
+- Practical: a deliverable GB address needs street + number + postcode unit; open data gives street + postcode *district* only. "Real street in a real town with a plausible district" is achievable; a real full postcode for that street is not.
+
+### Germany
+- **No national open address register.** BKG sells "Georeferenzierte Adressdaten" commercially. Open data is per Land: Hauskoordinaten / Gebäudereferenzen from 15 Länder under dl-de/by-2-0, dl-de/zero-2-0 or CC BY 4.0; **Bavaria missing** (~13 M people) — https://github.com/sister-software/mailwoman/issues/2300 (19.3 M permissive points vs 20.3 M in an ODbL build). Example: NRW "Gebäudereferenzen NW", dl-de/zero-2-0, semi-annual, https://open.nrw/dataset/172a0ba8-d470-47c1-ac89-b85f8190ac7e. Thuringia is dl-de/by-2-0. OpenAddresses lists 14 `de/*/statewide` sources (bb, bw, hb, he, hh, mv, ni, nw, rp, sh, sl, sn, st, th) — no `by`; Berlin appears as a Geoportal source *(check)*.
+- Hierarchy: **BKG VG250** (Land → Regierungsbezirk → Kreis → Verwaltungsgemeinschaft → Gemeinde with AGS keys; VG250-EW adds population), dl-de/by-2-0: https://gdz.bkg.bund.de/index.php/default/open-data/verwaltungsgebiete-1-250-000-stand-01-01-vg250-01-01.html. **Destatis Gemeindeverzeichnis GV100/GV-ISys** (all Gemeinden with AGS, population, area, PLZ of the Verwaltungssitz), quarterly ASCII/Excel: https://www.destatis.de/DE/Themen/Laender-Regionen/Regionales/Gemeindeverzeichnis/_inhalt.html — licence dl-de/by-2-0 *(not stated on that page; Destatis default, unverified)*. 16 Länder, ~400 Kreise, ~10,800 Gemeinden *(approx.)*.
+- **Postal codes are the blocker**: Deutsche Post **DATAFACTORY** is proprietary and licensed for internal use only (https://www.deutschepost.de/de/d/deutsche-post-direkt/datafactory.html; a FragDenStaat request for the data was refused). Open PLZ lists (OpenPLZ API, yetzt/postleitzahlen, Geofabrik postcode polygons) are **OSM-derived → ODbL** (OpenPLZ sources page https://www.openplzapi.org/en/sources/: streets and PLZ from OSM, municipalities from Destatis GV100). GeoNames `DE.txt` provenance not stated *(unverified)*. The Länder Hauskoordinaten carry PLZ per address, so a PLZ↔Ort↔Straße table can be derived legitimately for the 15 covered Länder; GV100 gives one PLZ per Gemeinde seat.
+- OpenAddresses: 32 `de/` sources.
+
+### Netherlands
+- **BAG** (Basisregistratie Adressen en Gebouwen), Kadaster. Licence **CC0 1.0** (https://data.overheid.nl/en/dataset/basisregistratie-adressen-en-gebouwen--bag-; OSM wiki: "released under a Public Domain license"). Embeddable, no attribution.
+- ~9.9 M addresses (mailwoman count). Object types: panden, verblijfsobjecten, **nummeraanduidingen** (huisnummer + huisletter + toevoeging + postcode), **openbare ruimten** (streets), **woonplaatsen**. Fields yield street, number, postcode (4 digits + 2 letters), woonplaats, gemeente (CBS code), RD/WGS84 point.
+- Download: BAG Extract XML (monthly full + mutations) and BAG GeoPackage via PDOK Atom `https://service.pdok.nl/lv/bag/atom/bag.xml`; REST API `https://api.pdok.nl/lv/bag/`. data.overheid.nl mentions distribution fees for some Kadaster products; the PDOK extract/GeoPackage is free.
+- Hierarchy: 12 provincies → 342 gemeenten (2025) → ~2,500 woonplaatsen. CBS gebiedsindelingen (CC BY 4.0); CBS "Kerncijfers wijken en buurten" for population.
+- Postcodes: in BAG since 2012 (CC0). PostNL's product is proprietary but unnecessary.
+- OpenAddresses: `nl/countrywide` (1 source).
+
+### France
+- **Base Adresse Nationale (BAN)** — **Licence Ouverte / Etalab 2.0** since 2020-01-01 (https://fr.wikipedia.org/wiki/Base_adresse_nationale; https://adresse.data.gouv.fr/outils/telechargements). Embeddable with attribution.
+- >25 M addresses. Downloads: CSV national / per département / per commune, "CSV with BAN ids", addok format; daily updates with weekly/monthly archives; also MVT/WFS/WMS. CSV fields `id, id_fantoir, numero, rep, nom_voie, code_postal, code_insee, nom_commune, code_insee_ancienne_commune, nom_ancienne_commune, x, y, lon, lat, type_position, alias, nom_ld, libelle_acheminement, nom_afnor, source_position, source_nom_voie, certification_commune, cad_parcelles` *(from BAN docs, not re-verified this session)*.
+- Hierarchy: INSEE **Code Officiel Géographique** (18 régions, 101 départements, ~34,900 communes), Licence Ouverte; INSEE populations légales per commune (Licence Ouverte).
+- Postal codes: in BAN; also La Poste **"Base officielle des codes postaux"** (commune → code postal → libellé d'acheminement, INSEE code), open licence, https://datanova.laposte.fr/datasets/laposte-hexasmal and https://www.data.gouv.fr/datasets/base-officielle-des-codes-postaux.
+- OpenAddresses: 107 `fr/` sources (one per département, fed weekly from BAN).
+
+### Spain
+- **Catastro INSPIRE Addresses (AD)** — per-municipality GML via Atom `http://www.catastro.minhap.es/INSPIRE/Addresses/ES.SDGC.AD.atom.xml`, refreshed ~6-monthly; WFS `http://ovc.catastro.meh.es/INSPIRE/wfsAD.aspx`. Fields: ThoroughfareName, address number, `AD:PostCode` (5 digits), AdminUnitName (municipio, provincia), point at building entrance or parcel centroid. Coverage: 95% of Catastro territory, **excludes País Vasco and Navarra** (own cadastres; Navarra has OpenAddresses source `es/nc/statewide`). Licence: "licencia de cesión de derechos que se obtendrá de manera automática", free, attribution "© Dirección General del Catastro" (https://www.catastro.hacienda.gob.es/webinspire/documentos/Conjuntos%20de%20datos.pdf; https://www.catastro.hacienda.gob.es/webinspire/index.html). mailwoman labels it CC BY 4.0 (15.66 M points); the Catastro PDF does not say "CC BY" — **read the "Descripción de la licencia" PDF before embedding**.
+- **INE Callejero del Censo Electoral** — VIAS, TRAMOS, PSEUDOVIAS, UNIDADES POBLACIONALES; ASCII ZIP, national + per province, Jan/Jul each year, `https://www.ine.es/prodyser/callejero/caj_esp/caj_esp_MMYYYY.zip` (https://datos.gob.es/en/catalogo/ea0010587-callejero-de-censo-electoral). Licence = INE aviso legal (free reuse with source citation *(unverified; page blocked)*). Whether TRAMOS carry código postal: *(unverified)*.
+- **CartoCiudad (IGN/CNIG)** — addresses, postcodes, toponyms, admin units; CC BY 4.0 (https://datos.gob.es/es/aplicaciones/direcciones-postales-de-cartociudad-espana). Its postcode polygon layer came from Correos and was **withdrawn in 2017** (Correos now sells it, ~€6k/yr) — https://www.nosolosig.com/articulos/asi-hice-el-mapa-de-los-codigos-postales-de-espana-con-sig-y-datos-abiertos. Per-address postcodes remain in Catastro AD and CartoCiudad addresses.
+- Hierarchy: 17 CCAA + 2 ciudades autónomas → 50 provincias → 8,132 municipios; INE padrón gives population per municipio; INE codes are open.
+- OpenAddresses: `es/countrywide` (Catastro) + `es/nc/statewide` + 4 others.
+
+### Australia
+- **G-NAF** (Geoscape via data.gov.au, https://data.gov.au/data/dataset/geocoded-national-address-file-g-naf). August 2026: **15,949,543 addresses** (15,108,510 principal). Quarterly. PSV tables (~5 GB unpacked, many tables to join) + **G-NAF Core** single simplified table; GDA94 or GDA2020.
+- Licence: **G-NAF End User Licence Agreement, "based on CC BY 4.0"**, plus: must not be used to compile addresses for sending mail unless each is verified against a secondary source; must follow the Australian Privacy Principles. Preferred attribution: "Incorporates or developed using G-NAF © Geoscape Australia licensed by the Commonwealth of Australia under the Open Geo-coded National Address File (G-NAF) End User Licence Agreement." Embeddable with the EULA text and that notice; the mail restriction is irrelevant to fake data but must be passed on.
+- Fields: number first/last with prefix/suffix, flat/level, street name + type + suffix, locality, state, postcode, lat/lon, mesh block; LGA via ABS codes.
+- Hierarchy: 8 states/territories → LGAs (~560) → localities (~15,000 in the G-NAF locality table). **ABS ASGS Edition 3** boundaries/allocation files, CC BY 4.0: https://www.abs.gov.au/statistics/standards/australian-statistical-geography-standard-asgs/edition-3-july-2021-june-2026/access-and-downloads/digital-boundary-files. Population: ABS ERP by SA2/LGA (CC BY 4.0).
+- Postcodes: in G-NAF. Australia Post's postcode file is free only as a **non-commercial PDF**; CSV products are paid (https://auspost.com.au/business/services/data-services/address-data/postcode-data). ABS "Postal Areas" (2,644 POAs, CC BY 4.0) are mesh-block approximations.
+- OpenAddresses: `au/countrywide` + 7 statewide + councils (58 sources).
+
+### Canada
+- **National Address Register (NAR)**, Statistics Canada — December 2024 release, **17.1 M** records, all 13 provinces/territories; two CSVs per province (addresses, locations). Licence **Statistics Canada Open Licence** (use, reproduce, distribute, sell, sublicense; attribution "Adapted from Statistics Canada, National Address Register, 2024. This does not constitute an endorsement by Statistics Canada of this product."): https://www.statcan.gc.ca/en/reference/licence; user guide https://www150.statcan.gc.ca/n1/pub/46-26-0002/462600022024002-eng.htm. Fields: `CIVIC_NO, CIVIC_NO_SUFFIX, OFFICIAL_STREET_NAME/TYPE/DIR, APT_NO_LABEL, MAIL_STREET_NAME/TYPE/DIR, MAIL_MUN_NAME, MAIL_PROV_ABVN, MAIL_POSTAL_CODE, CSD_CODE, CSD_ENG_NAME, CSD_FRE_NAME, PROV_CODE, REPPOINT_LATITUDE/LONGITUDE, BG_X/Y, BU_USE`. Postal code present "for most addresses" — completeness per province unverified. Embeddable.
+- **Open Database of Addresses (ODA) v1** (2021), ~10 M records from 99 datasets, OGL-Canada, CSV per province, postal code where the provider had it: https://www.statcan.gc.ca/en/lode/databases/oda. Older; prefer NAR.
+- Hierarchy: StatCan Standard Geographical Classification 2021 — 13 provinces/territories → 293 census divisions → ~5,160 census subdivisions (municipalities); population per CSD from Census 2021. StatCan Open Licence.
+- **Postal codes are proprietary**: Canada Post asserts copyright, sells the file, and sued Geocoder.ca in 2012 (https://opennorth.ca/resources/open-postal-code-data/). StatCan's PCCF is licensed from Canada Post and not redistributable. GeoNames `CA.txt` = **FSA (first 3 chars) only**. Only NAR/ODA `MAIL_POSTAL_CODE` gives open per-address postal codes.
+- OpenAddresses: `ca/countrywide` + 221 sources; Canada collection 1.25 GB.
+
+## 2. Cross-country hierarchy sources
+
+| Source | Licence | Content | Notes |
+|---|---|---|---|
+| Debian **iso-codes** (`iso_3166-1.json`, `iso_3166-2.json`; XML deprecated) | LGPL-2.1 (https://salsa.debian.org/iso-codes-team/iso-codes) | All ISO 3166-1 countries and 3166-2 subdivisions with codes, types, parents; 90+ translations | LGPL on a data file is awkward inside MIT; ship it as a separately-licensed data file or consume it only at build time. ISO's own OBP is not open. |
+| **GeoNames** `admin1CodesASCII.txt`, `admin2Codes.txt` | CC BY 4.0 | admin1/admin2 names + geonameids per country | admin1 codes follow ISO 3166-2 for some countries and FIPS/numeric codes for others. |
+| `amckenna41/iso3166-2` (PyPI) | MIT (code); data compiled from Wikipedia/RestCountries | 250 countries, >5,000 subdivisions with names, lat/lon | Convenient; provenance mixed. |
+| Per-country official lists (section 1) | see country | authoritative codes + population | Preferred for the 11 target countries. |
+
+## 3. Localities with population
+
+| Source | Licence | Size | Fit |
+|---|---|---|---|
+| **GeoNames** `cities500/1000/5000/15000` + `allCountries` (https://download.geonames.org/export/dump/) | CC BY 4.0 (attribution to geonames.org) | cities500 ≈185 k, cities1000 ≈130 k, cities5000 ≈50 k, cities15000 ≈25 k; fields geonameid, name, asciiname, alternatenames, lat/lon, feature class/code (PPL, PPLA…), country, admin1–4 codes, population, timezone | Best single cross-country locality+population source; population uneven (many zeros). Embeddable with attribution. |
+| **Census SUB-EST2025** (US) | public domain | ~19.5 k places | Authoritative US population. |
+| **Wikidata** (P1082 population, P131 located-in, P281 postal code) | CC0 | all | No attribution; needs a SPARQL/dump pipeline; quality varies. |
+| **Who's On First** (https://www.whosonfirst.org/docs/licenses/) | CC0 for WOF's own work; 312 sources ranging CC0…CC BY 3.0/4.0, OGL | localities, regions, counties with hierarchy + population where sourced | Attribution list is per source; no share-alike found. Heavy (one GeoJSON per record). |
+| **dr5hn/countries-states-cities-database** (https://github.com/dr5hn/countries-states-cities-database) | **ODbL 1.0** | 250 countries, 5,299 states, 153,765 cities, 844,248 postcodes (125 countries) | **Share-alike → not embeddable in MIT.** Community-maintained; README: data "may contain errors or lag behind geopolitical changes"; issue tracker shows large re-parenting fixes (8,727 French cities). Not a source of truth. |
+
+## 4. Postal codes — GeoNames and per-country openness
+
+- **GeoNames postal** (https://download.geonames.org/export/zip/, CC BY 4.0): TSV fields `country code, postal code, place name, admin name1, admin code1, admin name2, admin code2, admin name3, admin code3, latitude, longitude, accuracy (1–6)`. ~100 countries. Readme caveats: Canada and Netherlands partial codes only (full NL/CA files separate); UK from Royal Mail (© Royal Mail 2022 — attribution needed); Ireland/Malta first letters only; Argentina 5 chars; Brazil only `-000` codes; Chile/China partial. Lat/lon is algorithmic (matched to a GeoNames toponym or averaged from neighbours) — a locality centroid, not the postcode's. Per-country provenance is not documented, so DE's origin is unknown.
+
+| Country | Open? | Best open source |
+|---|---|---|
+| US | Yes (ZCTA, public domain); USPS list not a file | Census ZCTA + GeoNames US.txt |
+| Norway | Yes (NLOD 2.0) | Bring postnummerregister; also in Matrikkelen |
+| Denmark | Yes (CC BY 4.0) | DAWA `postnumre` (1,089) |
+| Finland | Yes (Posti terms) | Posti PCF (codes) + BAF (streets per code) |
+| UK | GB yes (OGL, Code-Point Open); NI restricted | Code-Point Open + OS Open Names |
+| Germany | **No** (Deutsche Post proprietary); OSM lists are ODbL | Derive PLZ↔Ort↔Straße from Länder Hauskoordinaten (15 Länder); GV100 for seat PLZ |
+| Netherlands | Yes (CC0, in BAG) | BAG |
+| France | Yes (Licence Ouverte) | La Poste hexasmal + BAN |
+| Spain | Polygon layer closed since 2017; per-address codes open via Catastro AD (own licence) | Catastro AD `AD:PostCode`; INE callejero |
+| Australia | In G-NAF (EULA/CC BY 4.0); Australia Post file paid/non-commercial | G-NAF locality table |
+| Canada | **No** (Canada Post proprietary; GeoNames FSA only) | NAR `MAIL_POSTAL_CODE` where present |
+
+## 5. OpenAddresses and OpenStreetMap
+
+**OpenAddresses** (https://openaddresses.io/, https://batch.openaddresses.io/)
+- Global collection 48.39 GB (2026-09-07); US in 4 regional collections (table in section 1); Canada 1.25 GB. 3,282 sources (1,586 flagged "need attention"). Sources per country in the batch list: US 2,381, CA 222, FR 107, AU 58, FI 40, DE 32, NO 17, ES 6, DK 1, NL 1, GB 0.
+- Licensing: the repo LICENSE (BSD-3) covers source definitions only. **Each source keeps its own licence**, recorded in the source JSON and `state.txt`; "most sources only require attribution", but several are share-alike (ODbL, CC BY-SA) — e.g. OSM-derived and some municipal sources (https://geocode.earth/docs/reference/data_sources/). For MIT: filter by the per-source licence field, keep PD/CC0/CC BY/OGL/dl-de sources, generate the attribution file from `state.txt`. The old results.openaddresses.io coverage page is frozen at mid-2021.
+- The official registers above are themselves the OA sources for NO/DK/FI/NL/FR/ES/AU (and NAD/state sources for the US); going to the register directly gives cleaner licensing and fresher data. OA's value is one uniform CSV schema (`LON,LAT,NUMBER,STREET,UNIT,CITY,DISTRICT,REGION,POSTCODE,ID,HASH`).
+
+**OpenStreetMap** — ODbL 1.0
+- Any *substantial* extract (systematic, >100 features, or an area with >1,000 inhabitants) is a Derivative Database; distributing it — including embedded in a library — requires offering it under ODbL (share-alike) plus attribution. OSMF guideline: https://osmfoundation.org/wiki/Licence/Community_Guidelines/Substantial_-_Guideline ("village map OK, town map not OK"; repeated small extractions count as one).
+- Consequence: an embedded street/postcode list derived from OSM (OpenPLZ, Geofabrik postcode polygons, yetzt/postleitzahlen, OSM-derived dr5hn rows) cannot ship inside an MIT package without that data file being ODbL and tracked as a separate database; downstream users mixing it into their databases inherit share-alike. Keep OSM out of the default; at most an *optional* download pack clearly labelled ODbL.
+
+## 6. Embed vs. optional download — sizes and a recommendation
+
+**US (ships by default with en_US)**
+- Embed (small, public domain): 56 states (+ USPS abbreviations, FIPS), 3,234 counties (GEOID, name, state), ~19.5 k incorporated places with population + county + lat/lon (SUB-EST2025 joined to the Gazetteer), 33,791 ZCTAs with the ZCTA→place/county relationship, and a compact **street-name pool per place** derived from NAD (distinct `St_Name`+`St_PosTyp` per `Post_City`/`State`, top-N by frequency). Places+ZCTAs ≈ 2–4 MB gzipped; a street pool for all ~19.5 k places would run to tens of MB — cap to the ~1,000 largest places or ~20 streets per place for the default pack.
+- Download pack: full NAD (~80 M rows, multi-GB), TIGER `featnames`/`addrfeat` per county, OA US collections (~23 GB zipped). Real house numbers + ZIP+4 come only from here.
+- Deliverability caveat: NAD ZIP is optional, and place naming mixes `Post_City` (USPS city) with `Inc_Muni`; use `Post_City` + `Zip_Code` for mailing realism.
+
+**Other countries** — same shape: embed hierarchy + localities-with-population + postcode↔locality table (a few MB per country: NO's 2.6 M addresses collapse to ~4.5 k postcodes and ~90 k street names; DK 1,089 postcodes / 54 k street names; NL/FR/AU similar), with street pools for the largest localities; keep full address points as download packs fetched from the official Atom/CSV endpoints at runtime, never vendored (G-NAF ~5 GB, BAN CSV ~2 GB, BAG ~2 GB, NAR ~1 GB).
+
+**Licence file needed in the repo** (`DATA-LICENSES.md`): CC BY 4.0 (GeoNames, Kartverket, Klimadatastyrelsen, Statistics Finland, DVV/SYKE, NLS, ABS, CartoCiudad), NLOD 2.0 (Bring), OGL v3 (OS, ONS), Licence Ouverte 2.0 (BAN, La Poste, INSEE), G-NAF EULA notice, StatCan Open Licence notice, dl-de/by-2-0 and dl-de/zero-2-0 (BKG, Länder), Posti terms + download date, Catastro "© Dirección General del Catastro". CC0/public domain (US Census, DOT NAD, BAG, Wikidata) need nothing but are worth listing.
+
+## 7. Open questions / unverified
+- NAD ZIP fill rate and the list of non-participating states (DOT site blocks fetches; read the NAD coverage map manually).
+- Unique US street-name count — compute from NAD once downloaded.
+- Catastro licence text ("Descripción de la licencia" PDF) — confirm redistribution inside a software package.
+- INE aviso legal wording; whether callejero TRAMOS include código postal.
+- Destatis GV100 licence statement; Berlin/Bayern Hauskoordinaten status.
+- SYKE/Ryhti address API fields and bulk-download availability (avoindata.fi blocked the fetch).
+- NAR `MAIL_POSTAL_CODE` completeness per province.
+- GeoNames DE/US postal-code provenance and counts.
+- BAN CSV field list re-verification.

docs/research/research-libraries.md

@@ -0,0 +1,287 @@
+# Fake-data libraries: data-category inventory
+
+Researched 2026-09-16. Sources: official docs and current `master`/`main`/`next` source trees. "Big four" = @faker-js/faker (JS), Faker (Python), gofakeit (Go), Datafaker (Java). Abbreviations in tables: **JS**, **PY**, **GO**, **DF**, **BG** (Bogus), **MM** (Mimesis), **CH** (Chance), **RB** (Ruby faker).
+
+## 1. Category modules and their generators
+
+Legend: names are the library's own identifiers (camelCase = JS/DF/BG, snake_case = PY/MM/RB, PascalCase = GO). Columns list what each library offers under that module; a `—` means the module does not exist there.
+
+### person
+
+| Lib | Generators |
+|---|---|
+| JS | firstName, lastName, middleName, fullName, prefix, suffix, sex, sexType, gender, bio, jobArea, jobDescriptor, jobTitle, jobType, zodiacSign (all name methods take `sex: female|male|generic`) |
+| PY | first_name/_female/_male/_nonbinary, last_name/_female/_male/_nonbinary, name/_female/_male/_nonbinary, prefix/_female/_male/_nonbinary, suffix/_female/_male/_nonbinary, language_name; separate providers: job (job, job_female, job_male), ssn (ssn; per-locale variants: vat_id, itin, ein, nif, cpf, personnummer …), passport (passport_number, passport_dob, passport_owner), profile (profile → job, company, ssn, residence, current_location, blood_group, website, username, name, sex, address, mail, birthdate; simple_profile) |
+| GO | Person (struct), Name, NamePrefix, NameSuffix, FirstName, MiddleName, LastName, Gender, Age, Ethnicity, SSN, EIN, Hobby, SocialMedia, Bio, Contact, Email, Phone, PhoneFormatted, Teams |
+| DF | Name: name, nameWithMiddle, fullName, firstName, femaleFirstName, maleFirstName, lastName, prefix, suffix, title, username; Demographic: race, educationalAttainment, demonym, sex, maritalStatus; Gender; Pronouns; BloodType; Mbti; Zodiac; Relationship; Hobby; Mood; Job; Passport; DrivingLicense; IdNumber (valid/invalid per country: US SSN, SE, ZA, SG FIN/UIN, CN, PT NIF, MX, PL PESEL, KR RRN, GE); CPF/CNPJ (BR); Nigeria; Australia |
+| BG | Name: FirstName, LastName, FullName, Prefix, Suffix, FindName, JobTitle, JobDescriptor, JobArea, JobType; Person object (Gender, FirstName, LastName, FullName, UserName, Avatar, Email, DateOfBirth, Address, Phone, Website, Company); country extensions: Ssn/Ein (US), Personnummer/Samordningsnummer (SE), Cpr (DK), Henkilotunnus (FI), Fodselsnummer (NO), Pesel/Nip/Regon (PL), CodiceFiscale (IT), Cpf/Cnpj (BR), Sin (CA), Nino (GB), Cnp (RO), NationalNumber (BE), Nif/Nipc (PT) |
+| MM | Person: first_name, surname/last_name, patronymic, full_name, name, title, username, password, email, birthdate, gender, gender_symbol, gender_code, sex, height, weight, blood_type, occupation, nationality, university, academic_degree, language, phone_number, telephone, identifier |
+| CH | first, last, name, prefix, suffix, gender, age, birthday, ssn, cpf, cf (IT codice fiscale), profession |
+| RB | Name (name, name_with_middle, first_name, male_first_name, female_first_name, neutral_first_name, last_name, prefix, suffix, initials), Gender, Demographic, IDNumber (per-country), Job, Relationship, Blood, DrivingLicence, Avatar, ChileRut, SouthAfrica, NationalHealthService |
+
+### address / location
+
+| Lib | Generators |
+|---|---|
+| JS (location) | buildingNumber, cardinalDirection, city, continent, country, countryCode, county, direction, language, latitude, longitude, nearbyGPSCoordinate, ordinalDirection, postalAddress, secondaryAddress, state({abbreviated}), street, streetAddress, timeZone, zipCode({state, format}) |
+| PY (address) | address, building_number, city, city_suffix, country, country_code, current_country, current_country_code, postcode, street_address, street_name, street_suffix; locale extras e.g. en_US: city_prefix, secondary_address, administrative_unit, state_abbr, zipcode_plus4, postcode_in_state, military_ship/state/apo/dpo; ja_JP: prefecture, city, town, chome, ban, gou, building_name; fr_FR: department, department_name, department_number, region; es_ES: region/autonomous_community; de_DE: city_with_postcode, street_suffix_short/long; en_GB: county; separate **geo** provider: coordinate, latitude, longitude, latlng, local_latlng(country), location_on_land (real geonames.org places with tz) |
+| GO (address) | Address (struct), City, Country, CountryAbr, State, StateAbr, Street, StreetName, StreetNumber, StreetPrefix, StreetSuffix, Unit, Zip, Latitude, LatitudeInRange, Longitude, LongitudeInRange |
+| DF (Address) | streetName, streetAddressNumber, streetAddress, secondaryAddress, zipCode, postcode, eircode, zipCodePlus4, zipCodeByState, countyByZipCode, streetSuffix, streetPrefix, citySuffix, cityPrefix, city, cityName, state, stateAbbr, latitude, longitude, latLon, lonLat, timeZone, country, countryCode, buildingNumber, fullAddress, mailBox; plus Country (name, code2/3, capital, currency, flag), Nation (nationality, language, capital, flag), Compass, Mountain, Planet, Space, Locality (locale codes) |
+| BG (Address) | ZipCode, City, StreetAddress, CityPrefix, CitySuffix, StreetName, BuildingNumber, StreetSuffix, SecondaryAddress, County, Country, FullAddress, CountryCode, State, StateAbbr, Latitude, Longitude, Direction, CardinalDirection, OrdinalDirection; premium Bogus.Locations: GPS, altitude, depth, geohash |
+| MM (Address) | street_number, street_name, street_suffix, secondary_address, address, state/region/province/federal_subject/prefecture (abbr), postal_code/zip_code, country, country_code (A2/A3/numeric), country_emoji_flag, default_country, city, latitude, longitude, coordinates (DMS option), continent, calling_code/isd_code, iata_code, icao_code |
+| CH | address, street, city, zip, postal (CA), postcode (GB), state, province, country, areacode, phone, latitude, longitude, altitude, depth, coordinates, geohash, locale |
+| RB | city, street_name, street_address, secondary_address, building_number, mail_box, community, zip_code(state_abbreviation:), zip, postcode, time_zone, street_suffix, city_suffix, city_prefix, state, state_abbr, country, country_by_code, country_name_to_code, country_code, country_code_long, latitude, longitude, full_address, full_address_as_hash; Travel: airport, train_station; Locations: australia |
+
+### company
+
+| Lib | Generators |
+|---|---|
+| JS | name, buzzAdjective, buzzNoun, buzzPhrase, buzzVerb, catchPhrase, catchPhraseAdjective, catchPhraseDescriptor, catchPhraseNoun (locale data: adjective, descriptor, noun, legal_entity_type, name_pattern) |
+| PY | company, company_suffix, catch_phrase, bs; locale extras (e.g. it_IT/pl_PL company_vat, nl_NL, ru_RU large_company, etc.) |
+| GO | Company, CompanySuffix, BS, Blurb, BuzzWord, Slogan, Job, JobDescriptor, JobLevel, JobTitle |
+| DF | Company: name, suffix, industry, profession, buzzword, catchPhrase, bs, logo, domainName, url; Business; Brand; IndustrySegments; Marketing; Team; Restaurant; Subscription; Twitter; University; Educator |
+| BG | CompanySuffix, CompanyName, CatchPhrase, Bs |
+| MM | Finance.company, company_type |
+| CH | company, profession |
+| RB | Company (name, suffix, industry, profession, type, catch_phrase, buzzword, bs, logo, ein, duns_number, swedish_organisation_number, czech_organisation_number, french_siren/siret, norwegian/australian/spanish/polish/russian/south_african/brazilian ids …), Business, Marketing, IndustrySegments, Team, University, Educator, Restaurant, Subscription, Construction |
+
+### internet
+
+| Lib | Generators |
+|---|---|
+| JS | displayName, domainName, domainSuffix, domainWord, email, exampleEmail, emoji, httpMethod, httpStatusCode, ip, ipv4, ipv6, jwt, jwtAlgorithm, mac, password, port, protocol, url, userAgent, username |
+| PY | email, safe_email, free_email, company_email, ascii_* variants, free_email_domain, domain_name, domain_word, safe_domain_name, tld, hostname, dga, http_method, http_status_code, iana_id, image_url, ipv4, ipv4_network_class, ipv4_private, ipv4_public, ipv6, mac_address, nic_handle(s), port_number, ripe_id, slug, uri, uri_extension, uri_page, uri_path, url, user_name; user_agent provider (chrome, firefox, safari, opera, internet_explorer, platform tokens); emoji provider |
+| GO | URL, UrlSlug, DomainName, DomainSuffix, IPv4Address, IPv6Address, MacAddress, HTTPStatusCode(Simple), HTTPMethod, HTTPVersion, LogLevel, UserAgent (+Chrome/Firefox/Opera/Safari/API), Username, Password, Emoji (+22 category fns), InputName, Svg |
+| DF | Internet: username, emailAddress(name), safeEmailAddress, emailSubject, domainName/Word/Suffix, url, webdomain, image, httpMethod, password(opts), port, macAddress, ipV4Address, privateIpV4Address, publicIpV4Address, ipV4Cidr, ipV6Address, ipV6Cidr, slug, uuidv3/4/7, userAgent, botUserAgent; Http; Domain; Sip; Emoji; SlackEmoji; Credentials; Aws; Azure; Hashing; Fingerprint; Computer; Device; Camera; Drone; Robin |
+| BG | Avatar, Email, ExampleEmail, UserName, UserNameUnicode, DomainName, DomainWord, DomainSuffix, Ip, Port, IpAddress, IpEndPoint, Ipv6, UserAgent, Mac, Password, Color, Protocol, Url, UrlWithPath, UrlRootedPath |
+| MM | Internet: content_type, dsn, http_status_message/code, http_method, ip_v4/v6 (+object, cidr, with_port, special), cloud_region, asn, mac_address, hostname, url, uri, query_string/parameters, tld, user_agent, port, path, slug, public_dns, http_request/response_headers |
+| CH | avatar, color, domain, email, fbid, google_analytics, hashtag, ip, ipv6, klout, tld, twitter, url |
+| RB | Internet (email, username, password, domain_name, ip_v4/v6, mac_address, url, slug, user_agent, uuid, bot_user_agent …), Internet::HTTP, Omniauth, Stripe, X (Twitter), Avatar, Placeholdit, LoremFlickr |
+
+### finance / payment / bank
+
+| Lib | Generators |
+|---|---|
+| JS | accountName, accountNumber, amount, bic, bitcoinAddress, creditCardCVV, creditCardIssuer, creditCardNumber, currency, currencyCode, currencyName, currencyNumericCode, currencySymbol, ethereumAddress, iban, litecoinAddress, pin, routingNumber, transactionDescription, transactionType |
+| PY | bank: aba, bank, bank_country, bban, iban, swift, swift8, swift11; credit_card: credit_card_number, credit_card_provider, credit_card_expire, credit_card_security_code, credit_card_full; currency: currency, currency_code, currency_name, currency_symbol, cryptocurrency(_code/_name), pricetag |
+| GO | Price, CreditCard, CreditCardCvv, CreditCardExp, CreditCardNumber, CreditCardType, Currency, CurrencyLong, CurrencyShort, AchRouting, AchAccount, BitcoinAddress, BitcoinPrivateKey, BankName, BankType, Cusip, Isin |
+| DF | Finance: creditCard(type), bic, iban(country), ibanSupportedCountries, usRoutingNumber; Money: currency, currencyCode, currencyNumericCode, currencySymbol; Currency; Stock; FinancialTerms; CryptoCoin; Coin; Business |
+| BG | Account, AccountName, Amount, TransactionType, Currency, CreditCardNumber, CreditCardCvv, BitcoinAddress, EthereumAddress, RoutingNumber, Bic, Iban; GB SortCode, VatNumber |
+| MM | Finance: bank, currency_iso_code, currency_symbol, cryptocurrency_iso_code/_symbol, price, price_in_btc, stock_ticker, stock_name, stock_exchange; Payment: cid, bitcoin_address, ethereum_address, credit_card_network/number/expiration_date, cvv, credit_card_owner |
+| CH | cc, cc_type, currency, currency_pair, dollar, euro, exp, exp_month, exp_year |
+| RB | Finance (credit_card, vat_number, ticker, stock_market), Bank (name, swift_bic, iban, account_number, routing_number, bsb_number), Currency, Crypto, Blockchain (bitcoin, ethereum, tezos, aeternity), Invoice, Stripe, Coin |
+
+### commerce / product
+
+| Lib | Generators |
+|---|---|
+| JS | department, isbn, price, product, productAdjective, productDescription, productMaterial, productName, upc |
+| PY | barcode: ean, ean8, ean13, localized_ean*; isbn: isbn10, isbn13; sbn; (no product/commerce provider in core; community Ecommerce provider) |
+| GO | Product (struct), ProductName, ProductDescription, ProductCategory, ProductFeature, ProductMaterial, ProductUPC, ProductAudience, ProductDimension, ProductUseCase, ProductBenefit, ProductSuffix, ProductISBN |
+| DF | Commerce: department, productName, material, brand, vendor, price(min,max), promotionCode; Barcode; Code (isbn10/13, gtin8/13, ean, asin, imei); Appliance; GarmentSize; Size; Tire; House; ElectricalComponents |
+| BG | Department, Price, Categories, ProductName, Color, Product, ProductAdjective, ProductMaterial, Ean8, Ean13 |
+| MM | Code: locale_code, issn, isbn, ean, imei, pin |
+| CH | — |
+| RB | Commerce (color, department, material, product_name, price, promotion_code, brand, vendor), Barcode, Code (isbn, ean, asin, imei, npi, nric, rut, sin), Appliance, Device, Camera, House |
+
+### vehicle / transport
+
+| Lib | Generators |
+|---|---|
+| JS | bicycle, color, fuel, manufacturer, model, type, vehicle, vin, vrm; airline: aircraftType, airline, airplane, airport, flightNumber, recordLocator, seat |
+| PY | automotive: license_plate, vin (per-locale plate formats) |
+| GO | Car (struct), CarMaker, CarModel, CarType, CarFuelType, CarTransmissionType; Airline*: AircraftType, Airplane, Airport, AirportIATA, FlightNumber, RecordLocator, Seat |
+| DF | Vehicle: vin, manufacturer, make, model(make), makeAndModel, style, color, upholstery(+Color/Fabric), transmission, driveType, fuelType, carType, engine, carOptions, standardSpecs, doors, licensePlate(state); Aviation; Transport; DrivingLicense |
+| BG | Vin, Manufacturer, Model, Type, Fuel; GbRegistrationPlate |
+| MM | Transport: manufacturer, car, airplane, vehicle_registration_code(locale) |
+| CH | — |
+| RB | Vehicle (vin, manufacture, make, model, make_and_model, style, color, transmission, drive_type, fuel_type, car_type, engine, car_options, standard_specs, doors, door, year, mileage, license_plate, singapore_license_plate, version), Drone, Travel::Airport, Travel::TrainStation |
+
+### lorem / text / words
+
+| Lib | Generators |
+|---|---|
+| JS | lorem: lines, paragraph(s), sentence(s), slug, text, word(s); word: adjective, adverb, conjunction, interjection, noun, preposition, sample, verb, words; hacker: abbreviation, adjective, ingverb, noun, phrase, verb; string: alpha, alphanumeric, binary, fromCharacters, hexadecimal, nanoid, numeric, octal, sample, symbol, ulid, uuid |
+| PY | lorem: word(s), sentence(s), paragraph(s), text(s), get_words_list; misc: password, md5, sha1, sha256, binary, boolean, null_boolean, csv/dsv/psv/tsv, fixed_width, json, json_bytes, image, tar, zip, uuid4 |
+| GO | Word families: Noun* (Common/Concrete/Abstract/Collective*/Countable/Uncountable/Proper/Determiner), Verb* (Action/Linking/Helping/Transitive/Intransitive), Adverb* (Manner/Degree/Place/Time*/Frequency*), Preposition* (Simple/Double/Compound), Adjective* (8 kinds), Pronoun* (8 kinds), Connective* (6 kinds), Word, Interjection, Sentence, Paragraph, LoremIpsum{Word,Sentence,Paragraph}, Question, Quote, Phrase{,Noun,Verb,Adverb,Preposition}, Comment; Hacker*, Hipster{Word,Sentence,Paragraph}; Letter(N), Digit(N), Vowel, Lexify, Numerify, RandomString, ShuffleStrings |
+| DF | Lorem (word(s), sentence(s), paragraph(s), characters, fixedString, maxLengthSentence, supplemental), Text (character, upper/lowercase, text with symbol rules), Verb, Word, Hacker, Hipster, Shakespeare, Yoda, Joke, ChuckNorris, Matz, NatoPhoneticAlphabet, FunnyName, Marketing |
+| BG | Lorem: Word(s), Letter, Sentence(s), Paragraph(s), Text, Lines, Slug; Hacker: Abbreviation, Adjective, Noun, Verb, IngVerb, Phrase; Rant: Review(s) |
+| MM | Text: alphabet, level, text, sentence, title, words, word, quote, color, hex_color, rgb_color, answer, emoji |
+| CH | paragraph, sentence, syllable, word, character, letter, string |
+| RB | Lorem (word(s), character(s), sentence(s), paragraph(s), question(s), paragraph_by_chars, multibyte), Markdown, Hipster, Hacker, Adjective, Verbs, Quote, Quotes::Shakespeare/Chiquito/Rajnikanth, ChuckNorris, Emotion, Source (code snippets), Html, Json, Types, Alphanumeric, String, Boolean, NatoPhoneticAlphabet |
+
+### date / time
+
+| Lib | Generators |
+|---|---|
+| JS | anytime, between, betweens, birthdate({mode: age|year}), future, past, recent, soon, month, weekday, timeZone (locale data: month, weekday) |
+| PY | date_time: date, date_time, date_object, time, time_object, date_between(_dates), date_time_between(_dates), date_this_century/decade/month/year, date_time_this_*, date_time_ad, date_of_birth(minimum_age, maximum_age), future_date(time), past_date(time), iso8601, unix_time, time_delta, time_series, timezone, pytimezone, am_pm, century, year, month, month_name, day_of_month, day_of_week |
+| GO | Date, PastDate, FutureDate, DateRange, NanoSecond, Second, Minute, Hour, Month, MonthString, Day, WeekDay, Year, TimeZone, TimeZoneAbv, TimeZoneFull, TimeZoneOffset, TimeZoneRegion |
+| DF | DateAndTime: future, past, between, birthday, birthdayLocalDate, duration, period; Time; TimeAndDate |
+| BG | Past, PastOffset, Soon, SoonOffset, Future, FutureOffset, Between, BetweenOffset, Recent, RecentOffset, Timespan, Month, Weekday |
+| MM | Datetime: date, datetime, time, timestamp, formatted_*, week_date, day_of_week, month, year, day_of_month, timezone, gmt_offset, periodicity, future/past_date(time), duration, bulk_create_datetimes |
+| CH | ampm, date, hammertime, hour, millisecond, minute, month, second, timestamp, timezone, weekday, year |
+| RB | Date (between, between_except, forward, backward, birthday, in_date_period, on_day_of_week_between), Time (between, between_dates, forward, backward) |
+
+### phone
+
+| Lib | Generators |
+|---|---|
+| JS | number({style: human|national|international}), imei |
+| PY | phone_number, country_calling_code, msisdn (per-locale formats) |
+| GO | Phone, PhoneFormatted |
+| DF | PhoneNumber: cellPhone, cellPhoneInternational, phoneNumber, phoneNumberInternational, phoneNumberNational, extension, subscriberNumber |
+| BG | PhoneNumber, PhoneNumberFormat |
+| MM | Person.phone_number/telephone (mask), Address.calling_code |
+| CH | phone, areacode |
+| RB | PhoneNumber (phone_number, cell_phone, country_code, phone_number_with_country_code, cell_phone_with_country_code, cell_phone_in_e164, area_code, exchange_code, subscriber_number, extension) |
+
+### science / medical
+
+| Lib | Generators |
+|---|---|
+| JS | chemicalElement, unit; medical locale data (en only) |
+| PY | — (community: Healthcare, Biology, Geoscience, Scientific) |
+| GO | — |
+| DF | Science: element, elementSymbol, unit, scientist, tool, quark, leptons, bosons; Medical, Medication, Disease, MedicalProcedure, CareProvider, Observation, BloodType, Measurement, Weather, Planet, Space, Mountain, Cannabis, LargeLanguageModel |
+| BG | premium Bogus.Healthcare |
+| MM | Science: rna_sequence, dna_sequence |
+| CH | — |
+| RB | Science (element, element_symbol, element_state, element_subcategory, scientist, modifier, tool), Space, Measurement, Cannabis, Medical (NationalHealthService) |
+
+### music / animal / food / books / entertainment
+
+| Lib | Generators |
+|---|---|
+| JS | music: album, artist, genre, songName; animal: bear, bird, cat, cetacean, cow, crocodilia, dog, fish, horse, insect, lion, petName, rabbit, rodent, snake, type; food: adjective, description, dish, ethnicCategory, fruit, ingredient, meat, spice, vegetable; book: author, format, genre, publisher, series, title |
+| PY | — (core has none; community Music, Sci Fi) |
+| GO | Song, SongName, SongArtist, SongGenre; PetName, Animal, AnimalType, FarmAnimal, Cat, Dog, Bird; Fruit, Vegetable, Breakfast, Lunch, Dinner, Snack, Dessert, Drink; Beer{Alcohol,Blg,Hop,Ibu,Malt,Name,Style,Yeast}; Book, BookTitle, BookAuthor, BookGenre; Movie, MovieName, MovieGenre; Celebrity{Actor,Business,Sport}; Minecraft (18 fns) |
+| DF | Music: instrument, key, chord, genre; RockBand; Artist; Kpop; Hololive; Animal: name, scientificName, genus, species; Cat, Dog, Horse; Food: ingredient, allergen, spice, dish, fruit, vegetable, sushi, measurement; Apple, Beer, Cheese, Coffee, Dessert, IceCream, Tea; Book, Movie, Show, OscarMovie; ~75 entertainment franchises, ~31 videogames, 9 sports |
+| BG | Music (premium Hollywood: movies, TV, actors) |
+| MM | Food: vegetable, fruit, dish, spices, drink |
+| CH | animal (with type), tv, radio, rpg, dice |
+| RB | Music (+11 bands), Creature (animal, bird, cat, dog, horse), Food, Beer, Coffee, Tea, Dessert, Book, Books (4 series), Movie/Movies (17), TvShows (39), Games (26), JapaneseMedia (10), Sports (6), Fantasy::Tolkien, Religion::Bible, Superhero, DcComics, Kpop, Ancient, GreekPhilosophers, Cosmere |
+
+### color / image / system / database / misc
+
+| Lib | Generators |
+|---|---|
+| JS | color: cmyk, colorByCSSColorSpace, cssSupportedFunction, cssSupportedSpace, hsl, human, hwb, lab, lch, rgb, space; image: avatar, avatarGitHub, dataUri, personPortrait, url, urlLoremFlickr, urlPicsumPhotos; system: commonFileExt/Name/Type, cron, directoryPath, fileExt/Name/Path/Type, mimeType, networkInterface, semver; database: collation, column, engine, mongodbObjectId, type; git: branch, commitDate, commitEntry, commitMessage, commitSha; number: bigInt, binary, float, hex, int, octal, romanNumeral; datatype.boolean; science.unit |
+| PY | color: color, color_name, safe_color_name, hex_color, safe_hex_color, rgb_color, rgb_css_color, color_hsl/hsv/rgb/rgb_float; file: file_extension, file_name, file_path, mime_type, unix_device, unix_partition; python: pybool, pydecimal, pydict, pyfloat, pyint, pyiterable, pylist, pyobject, pyset, pystr, pystr_format, pystruct, pytuple, enum; doi; emoji; user_agent |
+| GO | Color, HexColor, RGBColor, HSLColor, SafeColor, NiceColors; Image, ImageJpeg, ImagePng; CSV, JSON, XML, SQL, FileExtension, FileMimeType, Template, Markdown, EmailText, FixedWidth; ID, UUID; Error* (9 kinds); AppName, AppVersion, AppAuthor; Language, LanguageAbbreviation, LanguageBCP, ProgrammingLanguage; School; Gamertag, Dice; Bool, Weighted, FlipACoin; Number/Int*/Uint*/Float* ranges |
+| DF | Color, Image, File, App, ProgrammingLanguage, LanguageCode, Number, Bool, Unique, Options, Photography, Military, OlympicSport, Weather, Construction, Community, Chiquito … (263 providers total) |
+| BG | Images: DataUri, PicsumUrl, PlaceholderUrl, LoremFlickrUrl; System: FileName, DirectoryPath, FilePath, CommonFileName, MimeType, CommonFileType/Ext, FileType, FileExt, Semver, Version, Exception, AndroidId, ApplePushToken, BlackBerryPin; Database: Column, Type, Collation, Engine; Randomizer (numbers, chars, Guid, Hash, Enum, WeightedRandom, Shuffle …) |
+| MM | Hardware: resolution, screen_size, cpu, cpu_frequency, generation, cpu_codename, ram_type, ram_size, ssd_or_hdd, graphics, manufacturer, phone_model; Development: software_license, calver, version, stage, programming_language, os, boolean, system_quality_attribute; File, BinaryFile, Path, Cryptographic (uuid, hash, token, mnemonic), Numeric, Choice |
+| CH | bool, falsy, floating, integer, natural, prime, hex, guid, hash, coin, dice, normal (Gaussian), n, unique, weighted, android_id, apple_token, bb_pin, wp7_anid, wp8_anid2 |
+| RB | Color, File, Computer, Device, ProgrammingLanguage, App, Number, Boolean, Hash/Crypto, Json, Html, Markdown, Military, Compass, Coin, Mountain, Nation, Space, Time, Types, Slack Emoji, Vulnerability identifier |
+
+Module presence summary:
+
+| Module | JS | PY | GO | DF | BG | MM | CH | RB |
+|---|---|---|---|---|---|---|---|---|
+| person | x | x | x | x | x | x | x | x |
+| address | x | x | x | x | x | x | x | x |
+| company | x | x | x | x | x | (finance) | x | x |
+| internet | x | x | x | x | x | x | x | x |
+| finance | x | x | x | x | x | x | x | x |
+| commerce/product | x | barcode only | x | x | x | code only | — | x |
+| vehicle | x | plate/vin | x | x | x | x | — | x |
+| airline | x | — | x | x | — | iata/icao | — | x |
+| lorem/words | x | x | x | x | x | x | x | x |
+| date | x | x | x | x | x | x | x | x |
+| phone | x | x | x | x | x | x | x | x |
+| science | x | — | — | x | — | dna | — | x |
+| medical | data only | — | — | x | premium | — | — | x |
+| music | x | — | x | x | — | — | — | x |
+| animal | x | — | x | x | — | — | x | x |
+| food | x | — | x | x | — | x | — | x |
+| book | x | isbn | x | x | — | — | — | x |
+| color | x | x | x | x | x | text | x | x |
+| image | x | x | x | x | x | binaryfile | avatar | x |
+| system/file | x | x | x | x | x | x | — | x |
+| database | x | — | sql | — | x | dsn | — | — |
+| git | x | — | — | — | — | — | — | — |
+| hacker/hipster | x | — | x | x | x | — | — | x |
+| national IDs | — | ssn per locale | SSN/EIN | IdNumber | ext. pkgs | identifier | ssn/cpf/cf | IDNumber |
+| pop-culture | — | — | minecraft, movie | ~110 | premium | — | tv, rpg | ~85 |
+| hardware | — | — | — | Computer/Device | — | x | — | Computer/Device |
+| weather/space | — | — | — | x | — | — | — | Space |
+
+## 2. Locale coverage and address data structure
+
+### Locale counts
+
+| Lib | Locales | Structure |
+|---|---|---|
+| JS | 77 dirs under `src/locales` (incl. `base`, `en_BORK`, `en_AU_ocker`) | One dir per locale, one file per module per key (`en/location/city_name.ts`). Fallback chain per Faker instance: `[de_CH, de, en, base]`; first locale that defines a key wins. `base` holds locale-independent data (ISO codes, time zones). A key may be explicitly `null` to mean "not applicable" (e.g. no zip codes for HK) so it does not fall through to `en`. https://fakerjs.dev/guide/localization.html |
+| PY | 126 locale codes listed; per provider: address 67, person 86, ssn 56 | Python subclasses per `providers/<provider>/<locale>/__init__.py` overriding tuples/formats of the base. Missing provider for a locale falls back to `en_US`. Coverage is uneven (`am_ET` = phone only). Multi-locale `Faker(['it_IT','en_US'])` with optional weights. https://faker.readthedocs.io/en/master/locales.html |
+| GO | 1 (English/US). No locale API; `Language`/`LanguageBCP` only emit language codes | Static Go maps in `data/*.go`. Open issue #352 "generate data for specific country/language" unanswered. https://github.com/brianvoe/gofakeit/issues/352 |
+| DF | 127 yml files in `src/main/resources` (incl. 49 country-only files like `_SE.yml` and 78 language(-region) files); README says "60+" | One YAML per locale (`sv-SE.yml`) under `sv-SE: faker: address: …`; `en/` split into 264 per-provider files. Locale chain `de_CH → de → en`, key-by-key. Values are templates `#{Name.first_name}`. Custom yml via `faker.addPath(locale, path)`. Country file (`_SE.yml`) layered over language file. |
+| BG | 50 (`data/*.locale.json`; README says 46) | Verbatim copy of faker.js (v5-era) locale JSON, merged with `data_extend/*.locale.json` by a gulp task, shipped as BSON. Falls back to `en`. https://github.com/bchavez/Bogus/wiki/Creating-Locales |
+| MM | 47 (53 dirs incl. `global`, `int`, `bin`, template) | Per locale six JSON files: `address, datetime, finance, food, person, text`. Everything else (internet, payment, code, hardware…) is locale-independent. |
+| CH | 1 (en; `postal` CA and `postcode` GB formats exist, `locale()` only returns codes) | Data inline in `chance.js`. |
+| RB | 58 yml files ("over 40" in README) | One YAML per locale `lib/locales/<code>.yml` (`en` split into `en/*.yml`), I18n gem, key-level fallback to `en`. |
+
+### Address data: real or synthetic, and correlation
+
+| Lib | Cities | Street names | Postcodes | Hierarchy / correlation |
+|---|---|---|---|---|
+| JS | `en`: real list `city_name` (~1000 US cities) is one of 5 `city_pattern`s; the other 4 are `{prefix} {firstName}{suffix}` style synthetic. `de`: 327 real cities; `sv`: no `city_name`, pattern-only synthetic. Varies per locale. | `en`: `street_pattern` = `{firstName} {street_suffix}` / `{lastName} …` (synthetic); `de`: 1800 real street names (from one NRW region). | Format masks (`#####`, `#####-####`). `en_US` has `postcode_by_state` so `zipCode({state:'CA'})` gives a state-prefixed range. | None between city↔state↔zip. `county` is a flat list. `timeZone` is per-locale. `nearbyGPSCoordinate` is the only "correlated" geo function. Issue #983: en_CA returned US cities; PR #2141 added real cities for ZA locales — real-city coverage is locale-by-locale volunteer work. |
+| PY | Per locale. `en_US`/`en_GB`: synthetic (`{city_prefix} {first_name}{city_suffix}`). `de_DE` 400+ real, `sv_SE` 45 real, `pl_PL` real, `ja_JP` real prefectures/cities/towns, `es_ES` real provinces + regions. | `en_US`: `{first_name} {street_suffix}`; `sv_SE`: prefix+suffix ("Björkgatan"); `de_DE`: synthetic name+Straße; `pl_PL`: real street-name pools. | Masks. `en_US`: `postcode_in_state` with real per-state numeric ranges (still random within range, "not correlated with cities"). `fr_FR`: postcode built from a real department number (correlated to department, not to city). `en_GB`: real outward-area letters but random assembly. `sv_SE`: `%####`. | No country→state→city hierarchy anywhere; `ja_JP` has prefecture–city pairs in data but methods pick independently. `geo.local_latlng(country)` returns real geonames places (name, lat, lon, country, tz) — the only real, correlated location record in the library. `address_formats` are weighted (`25.0` standard vs `1.0` military). |
+| GO | Real list of major US cities | Real-looking suffix lists; StreetName is a pool | `#####` mask | None: `Address()` picks city, state, zip independently; lat/lon are uniform random on the globe, not near the city. Open issue #196 "Better address" unanswered. |
+| DF | `en`: `city` formats are all synthetic (`#{city_prefix} #{Name.first_name}#{city_suffix}`); `cityName` draws from a separate real list. `sv-SE`: real `city_name` list; `de`: synthetic. | `en`: `#{Name.first_name} #{street_suffix}`; `de`: real `street_root` list combined randomly. | Masks; `en-US` `postcode_by_state` (`350##`), `zipCodeByState`, `countyByZipCode` (en-US only), `eircode` (IE). | None by default. Issue #1551 (Locale.CHINA yields "福建省南京市") open as "proposal". Issue #1477 unresolved directive for city names in some locales. |
+| BG | Whatever faker.js v5 had per locale (`en`: synthetic patterns only — real `city_name` list postdates the copy) | synthetic | Masks; `en_US` `postcode_by_state` | None. Issue #481 "Invalid Zip Codes (US)" open. Issue #342 "City returns people's names" (side effect of `{firstName}` patterns). |
+| MM | Real lists per locale (`en`: real US cities; `sv`: real municipalities incl. small towns) | `en`: real San Francisco street names; `sv`: real Swedish street names | `postal_code_fmt` mask per locale; `sv` has none | `state` (name+ISO 3166-2 abbr) real, but no link to city or postcode; `default_country` tied to locale. Open question #1584 on choosing default city per locale. |
+| CH | Synthetic (3-syllable word) | Synthetic word + real suffix | Real-format masks | None |
+| RB | `en`: synthetic patterns (`city_with_state` exists but is `city, state` random) ; some locales have real lists | `{first_name} {street_suffix}` | `#####`; `zip_code(state_abbreviation:)` uses per-state pattern (`900##`); doc says "may not be an actual state zip" | `full_address_as_hash` returns fields but does not correlate them. Issue #2881 (state vs state_abbr mismatch, country vs country_code mismatch) open; #1958 asked for geocodable city+zip; #2899 fake zips closed "not planned". |
+
+**Which libraries correlate any address fields at all:** Python Faker (`fr_FR` postcode↔department, `en_US` zip↔state range, `geo.local_latlng` real records), faker-js/Datafaker/Bogus/Ruby (`zip↔state` range only, en_US), Mimesis (locale↔country only). None ships a country→region→city→postcode hierarchy or city↔lat/lon.
+
+### Other localized fields (what varies per locale)
+
+- JS: `sv` overrides color, commerce, company, date, internet, location, person, phone_number, team; `en` defines 24 modules; `base` 9. So finance/vehicle/animal/food/science are effectively English-only.
+- PY: per-locale providers exist for address, automotive (plate formats), bank (IBAN/BBAN/SWIFT by country), company (VAT IDs), currency, date_time (month/day names), internet (TLDs, free email domains, user-name formats), job, lorem (word lists), person, phone_number, ssn (national IDs with check digits), passport, color, barcode (EAN prefixes), misc.
+- DF `de.yml` overrides: address, company, compass, creature, lorem, hipster, name, color, commerce, book, university, chuck_norris, space, music, games, food, simpsons, dr_who, vehicle. Country files (`_SE.yml`) mainly phone_number and address formats.
+- MM: only address, datetime, finance, food, person, text are localized.
+
+### Name weighting
+
+| Lib | Weighted by frequency? |
+|---|---|
+| PY | Yes, default `use_weighting=True`. `en_US`: ~500 female + ~500 male first names weighted from SSA decade tables, ~1000 surnames weighted by US Census; `sv_SE`: ~1000/1000/1000 weighted from Skatteverket/ISOF. Plus weighted `address_formats`, `random_element(OrderedDict)`. |
+| JS | No. Plain lists (`en`: ~650 female, ~550 male, ~400 generic first names). `helpers.weightedArrayElement` exists for user data only. |
+| GO | No. ~1500 first, ~500 last, plain slices. `Weighted(options, weights)` for user data. |
+| DF | No. `en`: ~1200 male, ~4400 female first names, ~1000 last, unweighted. |
+| BG | No (faker.js data); `Randomizer.WeightedRandom` for user data. |
+| MM | No. `en`: ~4000 female, ~3000 male, ~2500 surnames, unweighted. |
+| CH | No; `weighted()` helper only. |
+| RB | No. |
+
+## 3. Common complaints (issues and posts)
+
+- **Uncorrelated address parts** — the single most repeated request across every library:
+  - faker-ruby #1958 "Need an address object that has a street, city, zip that correlate" (geocoding breaks) https://github.com/faker-ruby/faker/issues/1958; #2881 "Consistent addresses" (state ≠ state_abbr, country ≠ country_code) https://github.com/faker-ruby/faker/issues/2881; #2899 "Fake Zip Codes" closed not-planned https://github.com/faker-ruby/faker/issues/2899; #275 zips like 90099 https://github.com/stympy/faker/issues/275
+  - fzaninotto/Faker #625 "Faking Legitimate Addresses" https://github.com/fzaninotto/Faker/issues/625
+  - java-faker #378 "Italy, Lima" / "USA, Moscow" https://github.com/DiUS/java-faker/issues/378; Datafaker #1551 wrong province–city for China https://github.com/datafaker-net/datafaker/issues/1551
+  - gofakeit #196 "city may not be in the same state as the zip" https://github.com/brianvoe/gofakeit/issues/196
+  - Bogus #481 invalid US zips https://github.com/bchavez/Bogus/issues/481
+  - Python Faker #697 "city in country" https://github.com/joke2k/faker/issues/697; #1812 `states_postcode` fails for territories
+- **Synthetic city names** ("East Jarretmouth", "Nord Müller-stadt"): faker-js #2021 merged `city`/`cityName` and documented that only some patterns are real https://github.com/faker-js/faker/issues/2021; Bogus #342 `Address.City` returns people's names https://github.com/bchavez/Bogus/issues/342; faker-js PR #2141 / #2127 / #3792 adding "real" or "more realistic" cities per locale.
+- **Localized data falling through to English**: faker-js #983 en_CA cities were US cities https://github.com/faker-js/faker/issues/983; Datafaker #1477 unresolved directive for city names https://github.com/datafaker-net/datafaker/issues/1477; #1358 ru-MD city names; Python #2432 vi_VN outdated administrative units.
+- **Fields within one record don't match**: Python #1185 profile name vs username/email https://github.com/joke2k/faker/issues/1185; #1420 email consistent with name https://github.com/joke2k/faker/issues/1420 (both stale, no maintainer answer); gofakeit `Person()` picks gender and first name independently and email from a fresh name.
+- **No frequency weighting / flat distributions**: blood types come out uniform when O+ and A+ are >70% of population (https://www.statology.org/how-to-validate-enhance-faker-profile-data-generation/); Python's `use_weighting` is the only built-in answer and only some locales supply weights; #1815 weighting silently dropped after adding a provider https://github.com/joke2k/faker/issues/1815.
+- **Formats that produce impossible values**: Python #1849 US phone numbers that can't exist; #1868 fr_FR postcodes with <5 digits; faker-js #1159 unrealistic BIC; #3429 routing numbers fixed to use a real Federal Reserve lookup table; Ruby #1123 wrong pt-BR zip ranges.
+- **No locales at all** in gofakeit (#352, open) and Chance.
+- **"Column-by-column" generation, no relationships between columns/tables** as a general critique of Faker-style tools (https://securityboulevard.com/2026/04/best-synthetic-data-generation-tools-and-platforms-compared-for-2026/).
+- **Uniqueness**: faker-js removed `faker.unique` and tells users to roll their own (https://fakerjs.dev/guide/unique.html); Python `fake.unique` raises `UniquenessException` after retries.
+
+## 4. Notable and unusual features
+
+- **Datafaker** — expression language `#{Name.first_name}`, `#{regexify '[a-z]{4,10}'}`, `#{numerify '##'}`, `#{bothify}`, `#{letterify}`, `#{templatify}`, `#{examplify}`, `#{options.option 'A','B'}`, `#{csv …}`, `#{json …}`, method calls with args (`#{date.birthday 'yy DDD'}`); nested resolution (https://www.datafaker.net/documentation/expressions/). `Schema.of(field(...), compositeField(...))` with transformers to CSV, JSON, YAML, XML, TOML, SQL (batch inserts, Postgres/Oracle/Spark dialects, ARRAY/MULTISET/ROW), `@FakeForSchema` POJO population (https://www.datafaker.net/documentation/schemas/). `faker.collection(...).len(3,5)`, infinite `stream()`, `faker.unique()`, sequences, custom yml via `addPath`. 263 providers, the largest pop-culture catalogue (≈110 franchises/games).
+- **gofakeit** — Go `text/template` based `Template()`, `Markdown()`, `EmailText()`, `FixedWidth()`; `Generate("{firstname} {regex:[a-z]{5}} {randomstring:[a,b]}")` mini-language; `Struct(&v)` fill via `fake:"{city}"`, `fakesize`, `format` tags; `Slice`, `Map`, `Regex`; `CSV/JSON/XML/SQL` output; `AddFuncLookup` registers custom functions with parameter metadata (used by its CLI and HTTP server); `Weighted()`; word families split by grammatical role (46 Noun/Verb/Adjective/Pronoun/Connective sub-kinds); `Error*` generators; Minecraft.
+- **Mimesis** — `Field`/`Fieldset`/`Schema` builder (`Field(locale)("person.full_name")`, keyed lookups with `key=` post-processors like `maybe`, `romanize`), `Schema(schema=lambda: {...}, iterations=n)` exporting to JSON/CSV/pickle, relational data with `SchemaRef` foreign keys, custom field handlers, factory-boy integration, fastest Python generator; locale = six JSON files; ISO country codes in A2/A3/numeric; DMS coordinates.
+- **Bogus** — `Faker<T>().RuleFor(x => x.Prop, f => …)` fluent rules with `StrictMode`, `CustomInstantiator`, `FinishWith`, `Rules()`, `Ignore`, rule sets; local vs global seeding, `UseDateTimeReference`; `Bogus.Distributions.Gaussian`; country ID extension packages; premium Locations/Healthcare/Hollywood/Text packages; Roslyn analyzer; `AutoBogus` auto-fills any class.
+- **faker-js** — locale fallback stack with explicit `null` = "not applicable"; `mergeLocales`; `helpers.fake('{{person.firstName}}')` template; `helpers.fromRegExp`, `weightedArrayElement`, `multiple`, `uniqueArray`, `mustache`; pluggable `Randomizer` (32/53-bit Mersenne), `Distributors` (uniform, exponential); `date.birthdate({mode:'age'|'year'})`; `location.nearbyGPSCoordinate`; `git` module; `person.bio`; typed per-locale definition files generated by script.
+- **Python Faker** — `use_weighting` (frequency-weighted names/formats from census data); `OrderedDict` weighted `random_element`; `fake.unique`, `fake.optional`; multi-locale instance with weights; `geo.local_latlng` real geonames records; `misc` file/archive/CSV/JSON/image generators; `pystr_format`; pytest fixture; CLI; 24 community providers (healthcare, market data, observability, airtravel, education, security, pyspark …).
+- **Chance** — `weighted`, `normal` (Gaussian), `unique`, `n`, `mixin`, `set` (override data), `pick/pickone/pickset`; mobile IDs (android_id, apple_token, bb_pin, wp7/8 anid); geohash/altitude/depth.
+- **Ruby faker** — `full_address_as_hash`, `Faker::Config.locale`, `Faker::UniqueGenerator`, `Faker::Config.random`; `Types` generator for random Ruby types; company registration numbers for ~12 countries; deep pop-culture catalogue (~85 themed generators).

docs/research/research-sources-se.md

@@ -0,0 +1,145 @@
+# Swedish (sv_SE) fake-data sources — research 2026-09-17
+
+Scope: names, identifiers, phone, company, plates/cars, words, dates/misc, other. Geography excluded.
+"Verified" = fetched and read this session. Local copies of inspected files: `/home/lilleman/.claude/jobs/e377c8d4/tmp/`.
+
+## Licence summary (embed in MIT repo?)
+
+| Source | Licence (exact) | Embed in MIT? | Attribution |
+|---|---|---|---|
+| SCB Statistikdatabasen + geodata | CC0 1.0 Universal | Yes | None required; SCB asks "Källa: SCB" for statistics taken from scb.se |
+| SCB files on scb.se (xlsx: names, SNI, SSYK, SUN) | Not stated per file; scb.se terms: cite "Källa: SCB" | Yes (flag: CC0 statement names only statistikdatabasen/geodata) | "Källa: SCB" |
+| Skatteverket testpersonnummer / testsamordningsnummer | CC0 1.0 (in DCAT) | Yes | None |
+| Skatteverket namn på nyfödda, efternamn lists | No licence in DCAT/page; "fri att använda, inga avtal eller avgifter" (PSI law 2022:818) | Yes, flag | Cite Skatteverket |
+| Bolagsverket statistics CSVs | CC BY 2.5 SE | Yes | Attribution required |
+| Bolagsverket/SCB bulk company files | Not stated ("värdefulla datamängder", free, no agreement) | Derive stats only; do not embed raw | Cite |
+| Bankinfrastruktur clearing CSV | Not stated; accuracy disclaimer | Facts only; flag | Cite BSAB |
+| Bankgirot PDFs | "© Bankgirocentralen BGC AB. All rights reserved." (Informationsklass: Öppen) | No verbatim copy; facts only | — |
+| PTS numbering plan PDF | Not stated (no licence on pts.se or in PDF) | Facts only; flag | Cite PTS |
+| Transportstyrelsen blocked combinations | Not stated on page | Facts (a list of 99 codes); flag | Cite |
+| SALDO morphology (Språkbanken) | CC BY 4.0 | Yes | Cite: Språkbanken (2017) SALDO's morphology, DOI 10.23695/agcm-ny22 |
+| Språkbanken word statistics | CC BY 4.0 | Yes | Cite Språkbanken |
+| Folkets lexikon | CC BY-SA 2.5 Generic | No (share-alike) | — |
+| SAOL (svenska.se) | Copyright, written permission required | No | — |
+| Hunspell sv_SE (DSSO) | MPL 1.1 / GPL 2 / LGPL 2.1 (tri); yeager/hunspell-sv LGPL-3.0 | Avoid | — |
+| Mobility Sweden registrations | "ange källa" (no open licence) | Top-N facts only | Cite Mobility Sweden |
+| Wikipedia (sv/en) | CC BY-SA 4.0 | No copying; facts only | — |
+| CLDR (via ICU/Intl) | Unicode licence | Yes | Unicode notice |
+
+## 1. Person names
+
+### SCB (historic, frozen)
+- SCB stopped producing name statistics from 2024 and refers to Skatteverket (Namnsök page, verified). Statistikdatabasen folder BE0001 now holds only "Äldre tabeller som inte längre uppdateras" (BE0001D newborns, BE0001G whole population).
+- Whole-population xlsx (verified, downloaded): https://www.scb.se/contentassets/9fe7dbb460994c72b835163dbc491ef9/namn-med-minst-tva-barare-31-december-2022.xlsx — 13.1 MB, last-modified 2024-11-21, reference date 2022-12-31, names with >= 2 bearers, names in UPPERCASE.
+
+| Sheet | Rows (incl. 4 header rows) | Columns |
+|---|---|---|
+| Efternamn | 411,802 | Efternamn, Antal bärare |
+| Förnamn kvinnor | 91,247 | Förnamn, Antal bärare |
+| Förnamn män | 79,128 | Förnamn, Antal bärare |
+| Tilltalsnamn kvinnor | 57,785 | Tilltalsnamn, Antal bärare, Medelålder (only where >= 10 bearers, else "-") |
+| Tilltalsnamn män | 49,404 | same |
+
+- Gender split: a name appearing in both kvinnor/män sheets gives P(female) = count_k / (count_k + count_m).
+- Middle names: "Förnamn" = every given name a person carries; "Tilltalsnamn" = the one used. Förnamn-minus-tilltalsnamn frequency approximates middle-name usage. Legal "mellannamn" (a second surname) is a different concept, no longer grantable under namnlagen 2016:1013 (not verified this session).
+- Junk rows to filter: single letters ("A"), initials ("A-C", "K C", "JR"), "A:SON".
+- Legacy PxWeb tables (folder BE0001G, verified via web UI): BE0001T06AR, BE0001TNamn10 (tilltalsnamn >= 10 bearers 1999–2020), BE0001T100, BE0001T08AR, BE0001T07AR, BE0001FNamn10 (förnamn >= 10 bearers), BE0001F100, BE0001T03Ar (efternamn top), BE0001ENamn10 (efternamn >= 10 bearers 1999–2020). PxWebApi v1 (`api.scb.se/OV0104/v1/doris/sv/ssd/START/BE/BE0001/BE0001G/...`) returns HTTP 400 and PxWebApi 2 (`/v2beta/api/v2/tables/BE0001T06AR`) returns "Non-existent table" — these tables are web-UI only now (verified). Use the xlsx.
+- Licence: SCB terms page (verified): "Statistik och geodata som SCB tillgängliggör som öppna data i statistikdatabasen och i vår geodataplattform har licensen Creative commons 0 1.0 Universal, CC0" and "Om du använder eller sprider statistik från scb.se ska dock som källa alltid SCB anges". The xlsx sits on scb.se, not in statistikdatabasen — flag; treat as CC0 + "Källa: SCB".
+
+### Skatteverket (current)
+- Namn på nyfödda (2021+), annual: DCAT https://skatteverket.entryscape.net/store/26/metadata/416; CSV `fbf_namn_nyfodda.csv` https://skatteverket.entryscape.net/store/26/resource/417; JSON API https://skatteverket.entryscape.net/rowstore/dataset/da2556d0-c717-45e8-a1d8-3320161d3a7d/json?_limit=100&_offset=0 (47,529 rows, verified). Columns: rangordning, kön (Kvinna/Man), fodelsear, uppdateringsdatum, namn, gruppering (Kommun/…), grupperingsvärde (e.g. ALE), antal. Names in Title case. No licence field in DCAT — flag.
+- Most common surnames (> 2,000 bearers, ~500 names), 2026 text file (verified, 517 lines, format `Andersson 211808`): https://www.skatteverket.se/download/18.70685bee19c85dd5dd03b59/1775049117448/Fria_efternamn_namn_antal_textfil_2026.txt (also two PDFs). Page: https://www.skatteverket.se/privat/folkbokforing/namn/bytaefternamn/sokblanddevanligasteefternamnen.4.515a6be615c637b9aa48e09.html. Licence not stated.
+- Skatteverket general open-data terms (verified): "Vår öppna data är fri att använda och kräver varken några avtal eller innebär några avgifter" under lag 2022:818; no CC licence named except on the test-number datasets (CC0).
+- Skatteverket publishes no whole-population first-name list as a file; its statistikportalen "Så många har ett visst namn" is an interactive search (page 404 on fetch; unverified).
+
+Recommendation: embed SCB 2022 xlsx-derived lists (top-N per sheet with counts) + Skatteverket 2026 surname counts for freshness.
+
+## 2. Identifiers
+
+### Personnummer (Folkbokföringslagen 1991:481 18 §, verified via lagen.nu)
+- 10 digits: YYMMDD + födelsenummer (3 digits) + kontrollsiffra. Födelsenummer odd = male, even = female. Separator "-" ; "+" once the person turns 100. 12-digit form YYYYMMDDNNNC (no separator).
+- Check digit: Luhn/mod 10 over the 9 digits YYMMDDNNN (weights 2,1,2,1,… from the left; sum digits of two-digit products; check = (10 − sum mod 10) mod 10).
+- Before 1990 the first two födelsenummer digits encoded county; now random (Wikipedia, unverified officially).
+
+### Samordningsnummer
+- Same layout; day + 60 (3rd → 63); individual number random 001–999, odd male / even female; same Luhn (Skatteverket rättslig vägledning, via search excerpt; page itself blocked). Now regulated by Lag (2022:1697) om samordningsnummer (18 a § FBL repealed) — flag.
+- Example from Skatteverket: man born 1970-10-03, no. 239 → 701063-2391.
+
+### Official test numbers (Skatteverket, CC0 1.0)
+- Testpersonnummer dataset: https://www.dataportal.se/sv/datasets/6_67959/testpersonnummer (DCAT: https://admin.dataportal.se/store/6/metadata/67959?recursive=dcat). "Testpersonnummer med sekelsiffror för åren 1890-2025. Nya testnummer för kommande år läggs ut i december varje år." Formats: Excel (one file), CSV (split into many period files), JSON API https://skatteverket.entryscape.net/rowstore/dataset/b4de7df7-63c0-4e7e-bb59-1f156a591763/json?_limit=100 — resultCount 43,895 (verified), one column `testpersonnummer`, 12 digits.
+- Observed pattern (verified on 1890s CSV file fully and API samples at offsets 0/10000/30000/43880): every day gets two numbers, födelsenummer 238 (female) and 239 (male), through 2026-12-31; 1890s use 980/981 (plus a few 982/983/936). All Luhn-valid. Skatteverket blocks these from ever being assigned.
+- Testsamordningsnummer: DCAT https://skatteverket.entryscape.net/store/9/metadata/153; years 1914–2025; CSV per year; sample file 2022 (1,498 rows): days 61–91, individual numbers 238/239, Luhn-valid.
+- Should a generator restrict to these? Skatteverket does not mandate it, but only these are guaranteed never to collide with a real person; any other Luhn-valid number may be real. Recommended: default to the official series (date + 238/239, or 980/981 for 1890s) and embed nothing but the rule (~530 kB if the full list were embedded). Flag: the 238/239 rule was sampled, not exhaustively verified for every year.
+- Common practice elsewhere (Faker etc.): random date + random 3 digits + Luhn — valid but collision-prone.
+
+### Organisationsnummer
+- 10 digits NNNNNN-NNNN, last = Luhn over first 9 (same algorithm as personnummer). Digits 3–4 ("month") are always >= 20, distinguishing from personnummer (Wikipedia; not found on an official page — flag).
+- First digit (gruppnummer), Bolagsverket page (verified): 5 = aktiebolag, filialer, banker, försäkringsbolag, europabolag; 9 = handelsbolag, kommanditbolag; 7 or 8 = bostadsrättsföreningar, ekonomiska föreningar, näringsdrivande ideella föreningar etc.; 2 or 8 = trossamfund; 20… = state agencies (Bolagsverket itself 202100-5489); 3 = foreign companies. Wikipedia adds: 1 = dödsbon, 2 = stat/region/kommun, 6 = samfällighetsföreningar, 8 = ideella föreningar/stiftelser. Enskild firma = owner's personnummer.
+- VAT number (Skatteverket, verified): "SE" + 12 digits = organisationsnummer (10) + "01" ("De två sista siffrorna är alltid 01"). Written SE556047352101.
+
+### Bankgiro / Plusgiro (Bankgirot "10-modul" PDF 2016-12-01, verified)
+- Bankgironummer: 7 or 8 digits, last digit mod-10 (Luhn) check; printed 991-2346 (7) / 5555-5551 (8), i.e. hyphen before the last four digits. 90-konton (charity) 900-000x…904-999x, always 7 digits (Wikipedia).
+- Plusgironummer: 2–8 digits, last digit Luhn (Wikipedia/samlogic; not from an official Nordea page — flag). Printed with hyphen before check digit, e.g. "12 34 56-7" (convention, unverified).
+
+### Clearing numbers / bank accounts
+- Bankinfrastruktur i Sverige AB CSV (verified, 57 rows, 40 actors): https://www.bankinfrastruktur.se/media/1melztro/tabell-over-clearingnummer-250305.csv (also `/media/sqfjsvkp/clearingnummertabell-for-nedladdning.csv`, 63 rows, newer, includes Zimpler 2130-2139). Columns: `Clearingnummer;Aktör;BIC;IBAN ID;Konto-typ;Metod IBAN konvertering`. Semicolon, UTF-8 BOM. Page: https://www.bankinfrastruktur.se/framtidens-betalningsinfrastruktur/iban-och-svenskt-nationellt-kontonummer. No licence; disclaimer "garanterar inte att publicerade uppgifter är korrekta".
+- Key ranges: Nordea 1100–1199, 1400–2099, 3000–3399, 3410–3999, 4000–4999 (3300 and 3782 = personkonto); Danske 1200–1399, 2400–2499; SEB 5000–5999, 9120–9124, 9130–9149; Handelsbanken 6000–6999; Swedbank 7000–7999 (4-digit) and 8000–8999 (5-digit clearing with own check digit); Länsförsäkringar 3400–3409, 9020–9029, 9060–9069; Skandiabanken 9150–9169; ICA 9270–9279; Avanza 9550–9569; Nordnet 9100–9109; Klarna 9780–9789.
+- Bankgirot "Bankernas kontonummer" 2024-02-22 (verified text): Typ 1 = 4-digit clearing + 7-digit account (11 digits) with mod-11 check, weights 1,10,9,…,1; comment 1 = weigh clearing minus first digit + 7 digits; comment 2 = whole clearing + 7 digits. Typ 2 = clearing not part of the account: Handelsbanken 9-digit account mod-11 (comment 2); Swedbank 8000–8999 up to 10 digits mod-10 (comment 3); Danske 9180–9189, Nordea personkonto 3300/3782, Sparbanken Syd 9570–9579 10 digits mod-10 (comment 1).
+- IBAN SE (Bankinfrastruktur page, verified): 24 chars = "SE" + 2 check digits + 3-digit IBAN ID (from CSV column, e.g. Nordea 300, SEB 500, Handelsbanken 600, Swedbank 800, Danske 120, LF 902) + 17-digit account (zero-padded; clearing included/excluded per "Metod IBAN konvertering" 1–3).
+
+## 3. Phone
+
+- Source: PTS "The Swedish numbering plan for telephony according to ITU-T E.164", 2024-01-08 (verified, text extracted): https://pts.se/globalassets/globala-block/nummertillstand/the-swedish-numbering-plan-for-telephony-according-to-itu---2024-01-08.pdf. Swedish "nrplansammanstallning-2026-02-04.pdf" and https://nummer.pts.se/NbrPlanSearch are behind a Radware bot check (both curl and Playwright blocked). No licence stated by PTS anywhere found; pts.se says contact pts@pts.se for data.
+- 264 geographic area codes (PDF text yields 263 "Area code for" rows; one likely split across a page break — flag). Area code = trunk "0" + NDC of 1–3 digits (08 Stockholm, 031 Göteborg, 040 Malmö, 0480 Kalmar). Geographic N(S)N (NDC + subscriber) max 9 digits, min 7 (2-digit NDC) or 8 (3-digit NDC); Stockholm min 7. Each row in the PDF gives NDC, max/min length, "Area code for <name>", so a `{ndc, name}` list can be built from it.
+- Mobile: NDC 70, 72, 73, 76, 79 — N(S)N exactly 9 digits, i.e. 07X-XXX XX XX (10 digits with trunk 0). Others: 71 mobile broadband/M2M (13 digits), 74 paging, 75 personal numbering, 77 shared cost, 20 freephone, 10 location-independent (10 AXX XX XX, A=1–8), 378 M2M fixed (10 digits).
+- Formatting (Språkrådet/Isof frågelådan, via search excerpts): hyphen after area code, subscriber digits grouped 2–3 from the left: `08-668 01 50`, `031-123 45 67`, `0480-123 45`, `070-123 45 67` (also `0701-23 45 67` accepted). International: `+46 70 123 45 67` (drop trunk 0).
+- Wikipedia "Lista över riktnummer i Sverige" cites the same PTS document (CC BY-SA — do not copy).
+
+## 4. Company
+
+- Legal forms + suffixes (Bolagsverket, lag 2018:1653 om företagsnamn): aktiebolag must contain "aktiebolag" or "AB" (verified); public companies add "(publ)"; handelsbolag "HB" / "handelsbolag", kommanditbolag "KB" / "kommanditbolag", ekonomisk förening "ekonomisk förening" / "ek. för.", enskild firma no suffix — statutory wording for HB/KB/ek.för. not verified on the fetched page (flag).
+- Legal-form codes (SCB "juridisk form", from Skatteverket organisationsnummer page and SCB variabelbeskrivning): 10 fysiska personer, 21 enkla bolag, 22 partrederier, 23 värdepappersfonder, 31 handelsbolag/kommanditbolag, 32 gruvbolag, 41 bankaktiebolag, 42 försäkringsaktiebolag, 43 europabolag, 49 övriga aktiebolag, 51 ekonomiska föreningar, 53 bostadsrättsföreningar, 54 kooperativ hyresrättsförening, 61 ideella föreningar, 62 samfälligheter, 63 registrerade trossamfund, 71 familjestiftelser, 72 övriga stiftelser, 81 statliga enheter, 82 kommuner, 84 regioner, 87 offentliga korporationer, 88 hypoteksföreningar, 91 oskiftade dödsbon, 92 ömsesidiga försäkringsbolag, 93 sparbanker, 94 understödsföreningar, 95 arbetslöshetskassor, 96 övriga svenska juridiska former, 98 utländska juridiska personer — codes beyond 49 quoted from memory of the SCB list, verify against https://www.scb.se/contentassets/8a8eb5c3d45f461ea93482f8e8d4de4f/variabelbeskrivning-api.pdf (flag).
+- Distribution (Bolagsverket ftgstat_oppna.csv, CC BY 2.5 SE, 404,779 rows, latin-1, verified): https://www.bolagsverket.se/statistik/ftgstat_oppna.csv. Columns: ar, manad, handelse (1 = nyregistrerade, 2 = alla registrerade, 3 = avslutade), regfam, län/kommun, then counts per form: AB, BAB (bankaktiebolag), BF (bostadsförening), BRF, EK (ekonomisk förening), E (enskild näringsidkare registered at Bolagsverket only), SE (europabolag), FL (filial), FAB (försäkrings-AB), HB, I (näringsdrivande ideell förening), KB, KHF, MB, SF, SB (sparbank), TSF (trossamfund), BFL, OFB, SCE, S, EGTS, FOF, TPAB, OTPB, TPF. New registrations 2025 (computed): AB 50,859; E 8,468; HB 1,557; BRF 398; EK 319; FL 269; KB 176; I 59. Column description xlsx: https://www.bolagsverket.se/download/18.46f4138717c599ee403aac05/1638951733112/beskrivning-av-csv-fil.xlsx.
+- Company-name register: Bolagsverket "värdefulla datamängder" bulk file https://vardefulla-datamangder.bolagsverket.se/bolagsverket/bolagsverket_bulkfil.zip (250 MB, updated 2026-09-14; org.nr, names, legal form, business description, address) and SCB bulk https://vardefulla-datamangder.bolagsverket.se/scb/scb_bulkfil.zip (71 MB; adds SNI up to 5 codes, status). Free, no agreement; licence not stated on the page (statistics CSVs are CC BY 2.5 SE) — flag. Use offline to derive name-token frequencies; do not embed rows. API (org.nr lookup only; name search "planned, no date"): https://bolagsverket.se/apierochoppnadata/hamtaforetagsinformation/apiforatthamtaforetagsinformation.3988.html.
+- Naming rules (Bolagsverket "Välja företagsnamn", verified): must be distinctive — not only a generic activity ("Bilverkstad AB") or a bare surname; no confusion with existing names/trademarks; no professional titles without credentials; no apparent internet address; "aktiebolag/AB/handelsbolag" ignored when comparing names. Typical patterns (unverified heuristic, derive from bulk file): `<Surname> <Trade> AB`, `<Surname>s <Trade>`, `<Place> <Trade> AB`, `<Fantasy word> AB`, `<Initials> <Trade> HB`, suffix words Bygg, Konsult, Holding, Invest, Fastigheter, Förvaltning, Entreprenad, Teknik, Design.
+- SNI 2007 (SCB, verified xlsx, 94 kB): https://www.scb.se/contentassets/d43b798da37140999abf883e206d0545/sni2007.xlsx — sheets Detaljgrupp (5-digit, 821 codes), Undergrupp (615), Grupp (272), Huvudgrupp (88), Avdelning (21 letters). Code format `01.110` + Swedish label. SNI 2025 replaces it from 2025-12-08 (835/651/287/87/22): https://www.scb.se/globalassets/sni-2025.xlsx (93 kB). Skatteverket/Bolagsverket transition status unverified. Licence: no statement on the classification pages; SCB terms apply ("Källa: SCB") — flag.
+
+## 5. Licence plates and cars
+
+- Format: `ABC 123` (since 1973) and `ABC 12A` (since 2019-01-16; last character a letter, never O). Letters never used: I, Q, V, Å, Ä, Ö. Digits 001–999 (000 not issued). Personal plates 2–7 characters, may use any letters, must not mimic the standard pattern. Source: sv.wikipedia Registreringsskyltar i Sverige (facts; not confirmed on a Transportstyrelsen page — flag).
+- Blocked letter combinations, Transportstyrelsen (verified, page updated 2026-03-31, 99 codes): https://www.transportstyrelsen.se/sv/vagtrafik/fordon/aga-kopa-eller-salja-fordon/registreringsskyltar/byte-av-registreringsnummer/Sparrade-bokstavskombinationer/
+  APA ARG ASS BAJ BSS CUC CUK CUM DUM ETA ETT FAG FAN FEG FEL FEM FES FET FNL FUC FUK FUL GAM GAY GEJ GEY GHB GUD GYN HAT HBT HKH HOR HOT KGB KKK KUC KUF KUG KUK KYK LAM LAT LEM LOJ LSD LUS MAD MAO MEN MES MLB MUS NAZ NRP NSF NYP OND OOO ORM PAJ PKK PLO PMS PUB RAP RAS ROM RPS RUS SEG SEX SJU SOS SPY SUG SUP SUR TBC TOA TOK TRE TYP UFO USA WAM WAR WWW XTC XTZ XUK XXL XXX ZEX ZOG ZPY ZUG ZUP ZOO
+- Car makes, Mobility Sweden 2025 full-year new passenger cars (272,987 total; press release 2026-01-02 "Svagt fordonsår 2025", via search excerpt; "ange källa" required): Volvo 48,961; Volkswagen 38,677; Toyota 22,189; Kia 19,922; Mercedes-Benz 17,372; Skoda 16,664; BMW 15,001; Audi 14,887; Peugeot 9,059; Polestar 7,601. Top models 2025: Volvo XC60 17,933, Volvo EX40, VW ID.7, Tesla Model Y 5,822 (press excerpts). Monthly xlsx: https://mobilitysweden.se/statistik/Nyregistreringar_per_manad_1. Not an open licence — embed as a curated list with attribution.
+- Transportstyrelsen fordonsstatistik (open data, by vehicle type only, not make): https://www.transportstyrelsen.se/sv/om-oss/statistik-och-analys/statistik-inom-vagtrafik/fordonsstatistik/. Trafa (https://www.trafa.se/vagtrafik/fordon/) publishes by owner/technical/region, no make/model list found; licence not stated.
+
+## 6. Words
+
+- SALDO morphology (recommended): https://sprakbanken.se/en/resources/saldom — CC BY 4.0, 128,036 entries, LMF XML `saldom.xml` 254,107,385 bytes (https://svn.spraakbanken.gu.se/sb-arkiv/pub/lmf/saldom/saldom.xml, 2017-09-19). Fields: lemma, POS, inflection paradigm, full word forms with MSD. Derive a small `{lemma, pos}` list (nouns/adjectives/verbs) offline; embed with the citation "Språkbanken (2017). SALDO's morphology. https://doi.org/10.23695/agcm-ny22".
+- Frequencies: Språkbanken word statistics https://sprakbanken.se/resurser/ordstatistik — `stats_all.txt.zip` 763.87 MB, CC BY 4.0, 2025-04-22, aggregated over modern corpora (Korp). Column layout not confirmed (FAQ entry not found; per-corpus files are tab-separated, 6 columns incl. word form, POS, lemgram, SALDO sense, compound flag, frequency) — flag; inspect a per-corpus stats file first (smaller).
+- Folkets lexikon (KTH): CC BY-SA 2.5 — share-alike, not suitable for an MIT-embedded derivative. SAOL: svenska.se "får inte användas … utan skriftligt tillstånd" — unusable. Hunspell sv_SE/DSSO: MPL/GPL/LGPL — avoid.
+
+## 7. Dates, money, email, domains
+
+- CLDR sv-SE (verified with Node 24.18.1 ICU): short date `2026-09-17`; long `17 september 2026`; full `torsdag 17 september 2026 kl. 16:30`; months januari…december (lowercase), short `jan. feb. mars apr. maj juni juli aug. sep. okt. nov. dec.`; weekdays måndag…söndag, short `mån tis ons tors fre lör sön`; currency `1 234,56 kr`; number `1 234 567,891`. ICU uses U+00A0 (NBSP) as group separator; typographic guidance says hard space, decimal comma.
+- Språkrådet (search excerpts): ISO 8601 `ÅÅÅÅ-MM-DD` in formal/technical text, `6 juli 2022` in running text; clock `14.30` (period; colon accepted); amounts `1 234,50 kr`.
+- Public holidays, Lag (1989:253) om allmänna helgdagar (verified): nyårsdagen 1/1, trettondedag jul 6/1, långfredagen, påskdagen, annandag påsk, Kristi himmelsfärdsdag (6th Thursday after Easter), pingstdagen, nationaldagen 6/6, midsommardagen (Saturday 20–26 June), alla helgons dag (Saturday 31 Oct–6 Nov), juldagen 25/12, annandag jul 26/12, plus Sundays. Semesterlagen 3 a § (verified): "Med söndag jämställs allmän helgdag samt midsommarafton, julafton och nyårsafton". Easter needs computing (Gregorian).
+- Email domains: no official statistic found (Internetstiftelsen "Svenskarna och internet" does not report providers). Common providers by general knowledge only (unverified): gmail.com, hotmail.com, outlook.com, icloud.com, yahoo.com/yahoo.se, live.se, telia.com (Telia is phasing out its mail service), bredband.net, comhem.se, spray.se, me.com. Flag as curated.
+- TLDs (Internetstiftelsen registrar stats JSON, verified): .se active domains 1,475,495 (2026-07); .nu 203,940 (2026-07). https://internetstiftelsen.se/registrar-stats/se/activedomainsmonthend.json, `/nu/…`. Licence not stated.
+
+## 8. Other
+
+- Occupations, SSYK 2012 (SCB, verified xlsx): https://www.scb.se/contentassets/0c0089cc085a45d49c1dc83923ad933a/ssyk-2012-koder.xlsx (50 kB; sheets 1/2/3/4-siffer; 429 four-digit codes) and job-title index https://www.scb.se/contentassets/0c0089cc085a45d49c1dc83923ad933a/ssyk_index_webb_2026.xlsx (213 kB, 2026-03-18, thousands of yrkesbenämningar → code). Licence: SCB terms ("Källa: SCB") — flag as with SNI.
+- Education, SUN 2020 (SCB): https://www.scb.se/contentassets/aeeedec0e28c465aa524429407dcd5ba/sun-2020_niva_inriktning2.xlsx (304 kB; levels + orientations). Population 25–64 (SCB "Befolkningens utbildning 2024", via search excerpt): ~10 % förgymnasial, ~40 % gymnasial, ~31 % eftergymnasial >= 3 years (remainder shorter post-secondary). Approximate — flag.
+- Blood groups (geblod.nu 2007 via sv.wikipedia Blodgruppsfördelning; geblod.nu today only says RhD+ ≈ 85 %): A+ 37, O+ 32, B+ 10, AB+ 5, A− 7, O− 6, B− 2, AB− 1 (%).
+- Regional/kommun data: out of scope (geography).
+
+## Unverified / flagged items
+1. CC0 applies by SCB's statement to statistikdatabasen/geodata; xlsx files on scb.se (names, SNI, SSYK, SUN) carry only "Källa: SCB" terms.
+2. Skatteverket name datasets and surname file: no licence field; only "fri att använda" under PSI law.
+3. Testpersonnummer 238/239 rule: sampled (1890s file fully, offsets 0/10k/30k/43.9k), not exhaustively checked; per-period CSV list has many files.
+4. Organisationsnummer "digits 3–4 >= 20" and group digits 1/6 — Wikipedia only.
+5. Plusgiro length/format — Wikipedia/blog only.
+6. PTS: no licence; 264 vs 263 area codes counted in the PDF text; Swedish 2026 PDF and nummer.pts.se blocked (Radware).
+7. Plate format details (000 excluded, O excluded as last letter, 2019-01-16) — Wikipedia only.
+8. Legal-form codes above 49 — from memory; verify against the SCB variabelbeskrivning PDF.
+9. Word-statistics column layout; email-provider list; company-name patterns; education shares — heuristic/approximate.

docs/research/research-sources-world.md

@@ -0,0 +1,616 @@
+# Open data sources and format rules — US/locale-neutral fake-data categories
+
+Researched 2026-09-17 for embedding as data files in an MIT-licensed library. Geography/addresses and Swedish data are out of scope (covered elsewhere). Each item that was not read first-hand from the cited page/file is marked **UNVERIFIED**.
+
+## Licence key (what each licence means inside an MIT repo)
+
+| Licence | Embed? | Obligation |
+|---|---|---|
+| Public domain (US federal work, 17 USC §105), CC0 1.0, PDDL, Unlicense | Yes | None; cite the source in the data file header as courtesy |
+| MIT, BSD-2, ISC | Yes | Keep the copyright + permission notice next to the copied data |
+| Unicode License v3 (OSI-approved 2023-11-17, verified) | Yes | Notice in LICENSE/NOTICE or docs |
+| WordNet License (verified at wordnet.princeton.edu) | Yes | Notice + disclaimer on all copies; no Princeton name in advertising |
+| W3C Document License 2023 | Yes (data tables) | Notice "includes material copied from or derived from <title> <URI>" |
+| CC BY 3.0 / 4.0 | Yes | Creator, licence link, indicate changes (verified deed) |
+| Apache-2.0 (code), LGPL-2.1/3.0 (data) | Awkward | LGPL data files stay LGPL; keep them out of an MIT repo, use as reference for facts only |
+| CC BY-SA, ODbL, CC BY-NC-SA, GFDL | No | Share-alike / non-commercial — incompatible |
+| No licence stated, "all rights reserved", proprietary directories (SWIFT BIC, Fed RTN, CUSIP, ICAO 8643, GS1 page, ISO OBP, SAE WMI, whatismybrowser, useragents.me, MySQL manual, Pantone) | No | Generate synthetically from the format rules instead |
+
+## Embed verdict at a glance
+
+| Category | Use (clean) | Use with notice | Avoid |
+|---|---|---|---|
+| 1 Names/IDs | SSA baby names (CC0), Census 2010 surnames (PD), IRS EIN prefixes / ITIN ranges, SSA SSN rules | — | — (prefix/suffix lists: hand-curate) |
+| 2 Finance | ISO 4217 via datasets/currency-codes (PDDL) or SIX list-one.xml; algorithms (ABA, Luhn, ISIN, CUSIP, IBAN mod 97, EIP-55, Bech32) | npm currency-codes (MIT), GLEIF BIC-LEI (custom notice), CLDR currency symbols (Unicode-3.0) | Fed routing directory, SwiftRef BIC directory, real CUSIP lists, php-iban (LGPL) verbatim |
+| 3 Company | NAICS 2022 (PD), SEC company_tickers.json (free reuse), GLEIF ELF list (CC0) | — | — |
+| 4 Internet | All IANA registries (CC0): media types, HTTP status, ports, TLDs, IPv4/IPv6 special-purpose; RFC 5737/3849/9637 ranges; locally-administered MAC | mime-db (MIT), Linguist languages.yml (MIT), top-user-agents (MIT) / user-agents (BSD-2), IEEE OUI (IEEE "does not assert copyright" — second-hand) | useragents.me, whatismybrowser, MySQL manual tables |
+| 5 Vehicle | NHTSA vPIC makes/models/WMIs/value lists (PD); VIN check-digit rule | — | SAE WMI DB, openalpr plate patterns (AGPL), Wikipedia lists verbatim |
+| 6 Airline | OurAirports (PD/Unlicense), Wikidata airlines + aircraft types (CC0) | OPTD airlines (CC BY 4.0) | OpenFlights + npm mirrors (ODbL), ICAO Doc 8643 + scraped mirrors, OpenSky aircraft DB |
+| 7 Science | PubChem periodic table (PD), NASA planetary fact sheet (PD), Wikidata animals/plants/moons (CC0), Stanford/AABB blood types (facts) | BIPM SI brochure (CC BY 4.0), GBIF backbone (CC BY 4.0 + citation), faker-js lists (MIT) | Bowserinator periodic table (CC BY-SA 3.0) |
+| 8 Text | Moby POS (PD), 12dicts PD lists, XKCD colours (CC0), Lorem ipsum/Cicero, Bartlett via Gutenberg (PD) | WordNet 3.x (WordNet License), Open English WordNet (CC BY 4.0), SCOWL (MIT-like), CSS named colours (W3C notice), Google Books ngrams (CC BY 3.0) | wordfreq data (CC BY-SA), hermitdave (CC BY-SA), Norvig count_1w (LDC-derived), Wikiquote |
+| 9 Dates | IANA tzdb (PD) | CLDR cldr-dates-full (Unicode-3.0) | — |
+| 10 ISO | datasets/country-codes (PDDL), LoC ISO 639-2 (PD, UNVERIFIED layout), Census state.txt (PD) | annexare/Countries, i18n-iso-countries, langs, iso-639-1 (all MIT) | mledoze (ODbL), lukes + stefangabos (CC BY-SA), Debian iso-codes (LGPL), SIL 639-3 table (no-redistribution clause), olahol iso-3166-2 (unclear) |
+| 11 Commerce | Check-digit algorithms (EAN/UPC/GTIN, ISBN-10/13, ISSN, IMEI Luhn); GS1 demo prefix 952 | faker-js en commerce word lists (MIT) | GS1 prefix page + ISBN range file verbatim, Wikipedia tables verbatim |
+| 12 Books/music/food | Gutenberg pg_catalog.csv (CC0), MusicBrainz genre names (CC0), ID3v1 genres, Wikidata instruments/dishes/cuisines/cocktails (CC0), USDA FoodData Central (CC0) | Open Brewery DB (MIT) | MusicBrainz tag associations (CC BY-NC-SA), IBA site text |
+
+## Corrections to the brief's assumptions
+
+- Bowserinator/Periodic-Table-JSON is CC BY-SA 3.0 (not MIT) and has 119 rows; use PubChem CSV (118, PD).
+- wordfreq code is Apache-2.0 and its data CC BY-SA 4.0 — avoid the data.
+- CSS Color 4 has 148 named colours (rebeccapurple included).
+- Gutenberg #27889 Bartlett is the 9th edition (1903); the 1919 10th is not on Gutenberg (UNVERIFIED elsewhere).
+- CLDR file path is `cldr-dates-full/main/<locale>/ca-gregorian.json`; current package 48.2.0.
+- Diners Club US/Canada is a Mastercard co-brand on 55; replace the loose "Maestro 50/56–58/6xxx" rule with the explicit IINs in part 2.
+- Maestro/Discover/Diners/JCB/UnionPay lengths vary 12–19 — generate per the table, not a fixed 16.
+- The 987-65-4320..4329 SSN advertising range is repeated by third parties but not found on any ssa.gov/POMS page (UNVERIFIED).
+- DUNS dropped its mod-10 check digit in 2006 — no validation rule today.
+- IEEE's OUI "does not assert copyright" statement is only recorded second-hand (Debian/FSF); no current IEEE page restates it.
+- Sites blocking automation (403/Cloudflare): ssa.gov, www2.census.gov, swift.com, redcrossblood.org, loc.gov, gs1.org (WebFetch), standards-oui.ieee.org (needs browser UA). Download those manually with a browser UA.
+
+## Recommended default picks for an en_US library
+
+- Given names: SSA `names.zip` aggregated to top-N per sex (CC0). Surnames: Census `Names_2010Census.csv` top-N (PD).
+- Countries: `datasets/country-codes` CSV (PDDL); flag emoji derived from alpha-2. Currencies: same file's ISO 4217 columns or `datasets/currency-codes`; symbols from CLDR `en` (Unicode-3.0).
+- Internet: IANA registries + mime-db; UAs from templates (no data) or `top-user-agents` (MIT).
+- Vehicles: vPIC `GetMakesForVehicleType/car` (195) + models per make; VIN generated with check digit.
+- Airports: OurAirports rows with `iata_code` and type large/medium (4,569). Airlines: Wikidata (CC0).
+- Words: WordNet 3.0 index files (with LICENSE) or Moby POS (PD) for POS-tagged words; XKCD + CSS for colours.
+- Timezones: `zone1970.tab` names only; offsets via `Intl` at runtime.
+- Check digits: implement from the rules in parts 1, 2 and 4 — no data needed.
+
+---
+
+# Part 1 — Names, finance, company: sources and rules
+
+Researched 2026-09-17. "Embed" = ship as a data file in an MIT repo. Everything not marked UNVERIFIED was read from the cited page or file this session.
+
+## 1. US person names and identifiers
+
+### SSA baby names (given names by year and sex)
+- Page: https://www.ssa.gov/oact/babynames/limits.html — national zip https://www.ssa.gov/oact/babynames/names.zip (7 MB), state zip https://www.ssa.gov/oact/babynames/state/namesbystate.zip (23 MB), territories 228 kB.
+- Format (per SSA page + readme as mirrored by https://github.com/hackerb9/ssa-baby-names): one `yobYYYY.txt` per year 1880–2025, CSV `name,sex,count`, name 2–15 chars, sex `M`/`F`, sorted by sex then count desc. Names with < 5 occurrences in a year are excluded.
+- Counts: ~100,364 distinct names, ~2.02 M rows across all years (hackerb9 snapshot; year of snapshot UNVERIFIED). Top 1000 names cover 71.5 % of 2025 births (SSA page).
+- Licence: data.gov catalog entry https://catalog.data.gov/dataset/baby-names-from-social-security-card-applications-national-data lists licence **CC0 1.0** (`https://creativecommons.org/publicdomain/zero/1.0/`), publisher SSA. US federal work, no attribution required; embeddable.
+- Note: ssa.gov returns 403 to non-browser user agents (curl/WebFetch); download with a browser UA or via the data.gov link.
+
+### Census 2010 surnames
+- Page: https://www.census.gov/topics/population/genealogy/data/2010_surnames.html
+- Files: full set https://www2.census.gov/topics/genealogy/2010surnames/names.zip (CSV + XLSX, `Names_2010Census.csv`), top 1000 `Names_2010Census_Top1000.xlsx`, docs https://www2.census.gov/topics/genealogy/2010surnames/surnames.pdf.
+- Fields (from surnames.pdf): `name, rank, count, prop100k, cum_prop100k, pctwhite, pctblack, pctapi, pctaian, pct2prace, pcthispanic`. `(S)` = suppressed percentage. Names are UPPERCASE.
+- Count: 162,253 surnames occurring ≥ 100 times (covers 90.1 % of people with a recorded surname); the CSV also has a trailing `ALL OTHER NAMES` row (UNVERIFIED — www2.census.gov returned 403 to curl this session; download with a browser).
+- Licence: US Census Bureau work → US federal government work, public domain in the US (17 U.S.C. §105). Census only asks for citation (https://www.census.gov/about/policies/citation.html: "U.S. Census Bureau, [Table], [Product], [Vintage], [URL], accessed on [date]"). Embeddable; cite source in the data file header.
+
+### Name prefixes / suffixes
+- No authoritative enumerated list exists. USPS Publication 28 §33 (https://pe.usps.com/text/pub28/28c3_011.htm) defines the fields: Name Prefix, First Name, Middle Name or Initial, Surname, **Suffix Title** = "maturity (e.g., JR, SR) and professional (e.g., PHD, DDS) suffixes"; example "MR. WALTER W. WITHERSPOON JR.".
+- Recommendation: hand-curate. Prefixes `Mr., Mrs., Ms., Mx., Dr., Rev.`; suffixes `Jr., Sr., II, III, IV, PhD, MD, DDS, Esq.`. Rule of thumb: generational suffix (Jr/Sr/II–IV) only with a male name in traditional US usage; never combine Jr. with Sr.; at most one generational suffix.
+
+### SSN
+- Structure: `AAA-GG-SSSS` (area 3, group 2, serial 4). No check digit (Wikipedia).
+- Randomization since **2011-06-25** (https://www.ssa.gov/employer/randomization.html): "Previously unassigned area numbers were introduced for assignment excluding area numbers 000, 666 and 900-999." Area no longer geographic.
+- Invalid per SSA POMS RM 10201.035 (https://secure.ssa.gov/apps10/poms.nsf/lnx/0110201035): area `000`, `666`, `900–999`; group `00`; serial `0000`.
+- Generator rule: area ∈ [001,899] \ {666}; group ∈ [01,99]; serial ∈ [0001,9999].
+- Advertising SSNs (https://www.ssa.gov/history/ssn/misused.html): `078-05-1120` (Woolworth wallet card, 1938; > 40,000 people claimed it) and `219-09-9999` (1940 Social Security Board pamphlet). Both are structurally valid — exclude from output.
+- `987-65-4320`–`987-65-4329` "reserved for advertising": repeated by searchbug.com, ssn-check.org, a Missouri state standard (https://oa.mo.gov/sites/default/files/CC-SocialSecurityNumberNamingStandardV050305.pdf, which even misprints it as 987-54-4329). **UNVERIFIED** — not found on any ssa.gov page or POMS this session; Wikipedia no longer carries it. Safe to use these ten as the library's "obviously fake" examples, but do not cite SSA for it.
+
+### ITIN (IRS)
+- IRS Publication 4757 (Rev. 6-2026), p. 5 (https://www.irs.gov/pub/irs-pdf/p4757.pdf): "All valid ITINs are nine-digit numbers in the same format as the SSN (9XX-8X-XXXX), beginning with a "9" and the 4th and 5th digits ranging from 50 to 65, 70 to 88, 90 to 92, and 94 to 99."
+- Group `93` = ATIN (adoption TIN); `89` reserved (Intuit/TaxSlayer support pages, not IRS — UNVERIFIED).
+- Generator rule: `9` + 2 digits + group ∈ {50–65, 70–88, 90–92, 94–99} + 4 digits (serial `0000` not documented as invalid for ITIN; avoid it anyway).
+
+### EIN (IRS)
+- Format `NN-NNNNNNN`. Page: https://www.irs.gov/businesses/small-businesses-self-employed/how-eins-are-assigned-and-valid-ein-prefixes ("The first two digits of your EIN show the IRS campus that assigned it or if you applied online").
+- Valid prefixes (union of the table, 2026-09-17):
+  `01 02 03 04 05 06 10 11 12 13 14 15 16 20 21 22 23 24 25 26 27 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 71 72 73 74 75 76 77 80 81 82 83 84 85 86 87 88 90 91 92 93 94 95 98 99`
+  By campus: Andover 10,12 · Atlanta 60,67 · Austin 50,53 · Brookhaven 01–06,11,13,14,16,21–23,25,34,51,52,54–59,65 · Cincinnati 30,32,35–38,61 · Fresno 15,24 · Kansas City 40,44 · Memphis 94,95 · Ogden 80,90 · Philadelphia 33,39,41–43,46,48,62–64,66,68,71–77,85–88,91–93,98,99 · Internet 20,26,27,33,39,41,42,45–47,81–88,92,93,99 · SBA 31.
+- Not valid: 00, 07, 08, 09, 17, 18, 19, 28, 29, 49, 69, 70, 78, 79, 89, 96, 97. No check digit.
+
+## 2. Finance
+
+### ABA routing number (RTN)
+- Wikipedia https://en.wikipedia.org/wiki/ABA_routing_transit_number: 9 digits; check: `(3(d1+d4+d7) + 7(d2+d5+d8) + (d3+d6+d9)) mod 10 = 0`.
+- First two digits: `00` US Government; `01–12` Federal Reserve districts (normal banks); `21–32` thrifts (historic, still valid); `61–72` non-bank payment processors/clearinghouses; `80` traveler's checks. Other prefixes unused (Wikipedia does not say so explicitly — inferred).
+- Generator rule: prefix ∈ {01–12, 21–32, 61–72}, 6 random digits, compute 9th as `(10 − (3d1+7d2+d3+3d4+7d5+d6+3d7+7d8) mod 10) mod 10`.
+- Open list of real RTNs: the Federal Reserve E-Payments Routing Directory (https://www.frbservices.org/resources/routing-number-directory/index.html) says "may not be sold, re-licensed, or otherwise used for commercial gain" → **not embeddable**. Generate synthetically.
+
+### Payment card numbers (Wikipedia "Payment card number", https://en.wikipedia.org/wiki/Payment_card_number; all Luhn-checked)
+| Network | IIN ranges | Length | CVV |
+|---|---|---|---|
+| Visa | 4 | 13, 16, 19 | 3 |
+| Mastercard | 51–55, 2221–2720 | 16 | 3 |
+| American Express | 34, 37 | 15 | 4 (front) |
+| Discover | 6011, 644–649, 65, 622126–622925 (UnionPay co-brand) | 16–19 | 3 |
+| JCB | 3528–3589 | 16–19 | 3 |
+| Diners Club International | 30 (covers 300–305, 3095), 36, 38, 39 | 14–19 | 3 |
+| Diners Club US & Canada | 55 (Mastercard co-brand) | 16 | 3 |
+| UnionPay | 62 | 16–19 | 3 |
+| Maestro | 5018, 5020, 5038, 5893, 6304, 6759, 6761, 6762, 6763 | 12–19 | 3 |
+| Maestro UK | 6759, 676770, 676774 | 12–19 | 3 |
+| Dankort 5019 · Mir 2200–2204 · RuPay 60,65,81,82,508 · Troy 65,9792 · UATP 1 (15) · Verve 506099–506198, 507865–507964, 650002–650027 (16/18/19) | | | |
+- CVV lengths: https://en.wikipedia.org/wiki/Card_security_code ("three-digit ... Visa, Mastercard, and Discover"; "American Express is a four-digit code on the front").
+- The requested "Maestro 50/56–58/6xxx" is a looser legacy rule; use Wikipedia's explicit IINs.
+- Wikipedia text is CC BY-SA 4.0 — copy the *facts* (not prose) into your table; facts are not copyrightable.
+
+### Published test cards
+- Stripe: https://docs.stripe.com/testing — Visa 4242424242424242, Mastercard 5555555555554444, Amex 378282246310005, Discover 6011111111111117, Diners 3056930009020004 (14), JCB 3566002020360505, UnionPay 6200000000000005; any future expiry, any CVC (4 for Amex).
+- PayPal: https://developer.paypal.com/tools/sandbox/card-testing/ — Visa 4005519200000004, 4012000033330026, 4012000077777777, 4012888888881881, 4217651111111119, 4500600000000061, 4772129056533503, 4915805038587737; Mastercard 2223000048400011; Amex 371449635398431, 376680816376961; Diners 36461510000039, 36461510000013; Maestro 6304000000000000, 5063516945005047; JCB 3636500000000260, 3636500000000989; CUP 6200680000000004, 6200680000000038.
+- Reuse: neither page carries an open licence (Stripe docs are under Stripe's site terms). The numbers themselves are Luhn-valid facts, widely reproduced; embedding the numbers with a "from Stripe/PayPal test docs" note is low-risk but **UNVERIFIED** as an explicit grant. Alternative: generate Luhn-valid numbers from the IIN table above — no dependency on either vendor.
+
+### BIC / SWIFT (ISO 9362)
+- https://en.wikipedia.org/wiki/ISO_9362: 4 letters bank code + 2 letters ISO 3166-1 country + 2 alphanumeric location + optional 3 alphanumeric branch (8 or 11 chars; `XXX` = primary office). Location second char `0` = test BIC, `1` = passive participant, `2` = reverse billing.
+- Generator rule: `[A-Z]{4}[A-Z]{2}[A-Z2-9][A-NP-Z0-9]([A-Z0-9]{3})?` — avoid `0`/`1` as the 8th char for a "live" BIC.
+- SWIFT's BIC Directory is proprietary (SwiftRef licence, redistribution needs a separate "SwiftRef Redistribution License": https://www.swift.com/myswift/ordering/order-products-services/swiftref-redistribution-license) → **not embeddable**.
+- Open alternative: GLEIF BIC-to-LEI mapping, https://www.gleif.org/en/lei-data/lei-mapping/download-bic-to-lei-relationship-files, files at https://mapping.gleif.org/api/v2/bic-lei/ (monthly zip; `LEI-BIC-20260828.zip` → `lei-bic-20260828T000000.csv`, columns `LEI,BIC`, 39,347 rows). Licence: "BIC/LEI Mapping Table License Agreement" (https://www.gleif.org/lei-data/lei-mapping/download-bic-to-lei-relationship-files/2017-12-21_annex-2_bic-to-lei-mapping-table-license-agreement_final.pdf) — a CC0-style grant "for any purpose whatsoever, including ... commercial" but **conditional on reproducing the notice** "SWIFT © and database rights [month year]. All rights reserved. This Mapping Table has been developed by SWIFT. Any use of the Mapping Table ... is subject to the BIC/LEI Mapping Table License Agreement ... For the latest BIC information and updates, always refer to www.swift.com/bic." Embeddable in an MIT repo with that notice in the data file; it is a list of real live BICs (bank names not included — join to GLEIF LEI data if names are wanted).
+
+### ISIN (https://en.wikipedia.org/wiki/International_Securities_Identification_Number)
+- 12 chars: 2-letter country + 9 alphanumeric NSIN + 1 check digit.
+- Check: convert letters A=10…Z=35 (ASCII − 55) to produce a digit string, then Luhn over that string (double every second digit from the right, sum digits, check makes total ≡ 0 mod 10). Pitfall: work on the *expanded* digit string; a transposed letter pair can pass.
+
+### CUSIP (https://en.wikipedia.org/wiki/CUSIP)
+- 9 chars: 6 issuer + 2 issue + 1 check. Values: digits as-is, A=10…Z=35, `*`=36, `@`=37, `#`=38.
+- Check: for i in 1..8, v = value; if i even, v ×= 2; sum += v div 10 + v mod 10; check = (10 − sum mod 10) mod 10.
+- Real CUSIPs are proprietary (CGS/FactSet); generate synthetic ones only.
+
+### IBAN (ISO 13616)
+- Validation (https://en.wikipedia.org/wiki/International_Bank_Account_Number): move first 4 chars to end, letters → 10–35, integer mod 97 must equal 1; max 34 chars; check digits 02–98.
+- SWIFT registry: https://www.swift.com/standards/data-standards/iban-international-bank-account-number, PDF release 101 https://www.swift.com/sites/default/files/files/iban-registry-v101.pdf (release 100 was Oct 2025); TXT also published (URL **UNVERIFIED** — swift.com returns 403/errors to automation this session). Terms: free of charge; no explicit licence text found (**UNVERIFIED**). Country count: 89 countries per Wikipedia (Dec 2024).
+- Embeddable per-country length tables already extracted from the registry:
+  - php-iban `registry.txt` (https://github.com/globalcitizen/php-iban, **LGPL-3.0**): pipe-separated, 121 rows (116 official + unofficial), columns `country_code|country_name|domestic_example|bban_example|bban_format_swift|bban_format_regex|bban_length|iban_example|iban_format_swift|iban_format_regex|iban_length|bban_bankid_start_offset|...|country_sepa|swift_official|...|currency_iso4217|central_bank_url|central_bank_name|membership`. LGPL data in an MIT repo is awkward; use it as a *reference* to build your own table (formats are facts) rather than copying the file.
+  - ibankit-js (https://github.com/koblas/ibankit-js, **Apache-2.0**, registry v95) — same caveat.
+  - Best route: own table `{country, length, bban_format}` derived from the SWIFT PDF (facts), cite SWIFT.
+
+### Currencies (ISO 4217)
+- ISO says use is free: https://www.iso.org/iso-4217-currency-codes.html — "ISO allows free-of-charge use of its country, currency and language codes from ISO 3166, ISO 4217 and ISO 639". Official list from SIX: https://www.six-group.com/dam/download/financial-information/data-center/iso-currrency/lists/list-one.xml (+ `.xls`, `list-three` historic). Fields `CtryNm, CcyNm, Ccy, CcyNbr, CcyMnrUnts`; published 2026-01-01; 280 entity rows, 178 distinct codes. No symbols. No licence text inside the XML.
+- Alternatives:
+  - datasets/currency-codes (https://github.com/datasets/currency-codes): **PDDL** (public domain); `data/codes-all.csv` fields `Entity, Currency, AlphabeticCode, NumericCode, MinorUnit, WithdrawalDate`; built from SIX list one + three; no symbols. Embeddable.
+  - npm `currency-codes` v2.2.0 (https://github.com/freeall/currency-codes): **MIT**; fields `code, number, digits, currency, countries[]`; generated from the SIX XML; no symbols. Embeddable.
+  - umpirsky/currency-list (https://github.com/umpirsky/currency-list): **MIT**; code → localised name only (311 entries in `data/en_US/currency.json`, includes historic); no symbols, no numeric codes, no decimals.
+  - Debian iso-codes (https://salsa.debian.org/iso-codes-team/iso-codes): **LGPL-2.1+**; `data/iso_4217.json` has `alpha_3, name, numeric` only (179 entries); no symbols/minor units. LGPL → avoid embedding.
+  - Symbols: none of the above carry them. Unicode CLDR (`common/main/en.xml` currency symbols, Unicode License, MIT-compatible with notice) is the usual source — **UNVERIFIED this session**.
+
+### Crypto addresses
+- Bitcoin (https://en.bitcoin.it/wiki/List_of_address_prefixes): P2PKH version 0x00 → leading `1`; P2SH 0x05 → leading `3`; Base58Check 25–34 chars (alphabet excludes `0OIl`; last 4 bytes = first 4 of double-SHA256 of version+payload). Testnet `m`/`n`, `2`, `tb1`.
+- Bech32 (BIP-173, https://github.com/bitcoin/bips/blob/master/bip-0173.mediawiki, BSD-2-Clause): HRP `bc`, separator `1`, charset `qpzry9x8gf2tvdw0s3jn54khce6mua7l`, all-lowercase (or all-uppercase), max 90 chars; P2WPKH `bc1q…` = 42 chars, P2WSH `bc1q…` = 62 chars.
+- Bech32m (BIP-350, https://github.com/bitcoin/bips/blob/master/bip-0350.mediawiki): witness v1+ (P2TR `bc1p…`, 62 chars), checksum constant `0x2bc830a3` instead of 1.
+- Ethereum EIP-55 (https://eips.ethereum.org/EIPS/eip-55, CC0): `0x` + 40 hex; keccak256 of the lowercase hex (no 0x, as ASCII); for each hex letter, uppercase iff the corresponding hash nibble ≥ 8 (bit 4·i set). Digits unchanged.
+
+## 3. Company
+
+### NAICS 2022 (US Census)
+- Page https://www.census.gov/naics/ ; files: 6-digit list https://www.census.gov/naics/2022NAICS/6-digit_2022_Codes.xlsx (82 kB; **1,012** six-digit codes, 111110…928120, verified by parsing), 2–6 digit https://www.census.gov/naics/2022NAICS/2-6%20digit_2022_Codes.xlsx, structure https://www.census.gov/naics/2022NAICS/2022_NAICS_Structure.xlsx, descriptions https://www.census.gov/naics/2022NAICS/2022_NAICS_Descriptions.xlsx, manual PDF https://www.census.gov/naics/reference_files_tools/2022_NAICS_Manual.pdf.
+- Format: xlsx, column A code (numeric), column B title (trailing spaces in some titles — trim).
+- Licence: no statement on the page or in the manual front matter; US federal government work → public domain in the US (17 U.S.C. §105). Embeddable; cite Census/OMB.
+
+### SEC EDGAR company names
+- `https://www.sec.gov/files/company_tickers.json`: object keyed `"0".."N"` of `{cik_str, ticker, title}`; **10,422** entries (2026-09-17). `company_tickers_exchange.json`: `{"fields":["cik","name","ticker","exchange"],"data":[...]}`.
+- `https://www.sec.gov/Archives/edgar/cik-lookup-data.txt`: 40 MB, **1,059,372** lines, `NAME:CIK:` (`COMPANY NAME:0001234567:`), all filers incl. individuals — filter to the ticker list for company names.
+- Access: declare `User-Agent: Company Name email@domain`, ≤ 10 req/s (https://www.sec.gov/os/webmaster-faq). Licence: "All Government-created content on sec.gov and EDGAR public filing content are free to access and reuse" (same FAQ). Embeddable; the ticker file (~800 kB) is the practical one.
+
+### DUNS
+- https://en.wikipedia.org/wiki/Data_Universal_Numbering_System: 9 digits, no significance; shown `NN-NNN-NNNN` or plain. Had a mod-10 check digit until ~Dec 2006; **dropped** since (expanded the pool by 800 M). So: any 9 digits are structurally valid today; no validation rule to enforce. Optional DUNS+4 suffix (4 alphanumerics) is user-assigned.
+
+### Legal entity suffixes per country
+- GLEIF ISO 20275 Entity Legal Forms code list v1.6 (2026-02-19): page https://www.gleif.org/en/lei-data/code-lists/iso-20275-entity-legal-forms-code-list ; CSV https://www.gleif.org/lei-data/code-lists/iso-20275-entity-legal-forms-code-list/2026-02-19-elf-code-list-v1.6.csv (758 kB; xlsx alongside).
+- Columns: `ELF Code, Country of formation, Country Code (ISO 3166-1), Jurisdiction of formation, Country sub-division code (ISO 3166-2), Entity Legal Form name Local name, Language, Language Code (ISO 639-1), Entity Legal Form name Transliterated name, Abbreviations Local language, Abbreviations transliterated, Date created, ELF Status ACTV/INAC, Modification, Modification date, Reason`.
+- Counts: 4,002 rows; 130 countries; 3,790 ACTV / 212 INAC; 1,518 rows have a local abbreviation (e.g. US `N.A.`, `FSA`; SE `AB`… multiple abbreviations separated by `;`). US has 737 rows (per-state forms). Abbreviations are the field you want for `Inc`, `LLC`, `GmbH`, `AB`, `SAS`, `Pty Ltd`; many rows have none, so filter on non-empty abbreviation and status ACTV.
+- Licence: GLEIF Open Data page (https://www.gleif.org/en/about/open-data): "The data on GLEIF's website is provided under a Creative Commons (CC0) license"; LEI Data Terms of Use: "provided under the CC0 licence, see CC0 1.0 Universal". The ELF page itself names no licence — that CC0 explicitly covers this code list is **UNVERIFIED**, but GLEIF treats all its published data this way. Embeddable.
+
+## Quick embed verdict
+| Item | Embed? | Licence / attribution |
+|---|---|---|
+| SSA baby names | yes | CC0 (data.gov) |
+| Census 2010 surnames | yes | US gov PD; cite Census |
+| EIN prefixes, ITIN/SSN rules | yes (rules) | facts from IRS/SSA |
+| NAICS 2022 | yes | US gov PD |
+| SEC company_tickers.json | yes | free to reuse (SEC) |
+| GLEIF ELF code list | yes | CC0 (GLEIF) |
+| GLEIF BIC-LEI (real BICs) | yes, with mandatory SWIFT notice | custom permissive licence |
+| ISO 4217 (SIX XML / datasets PDDL / npm MIT) | yes | free use (ISO), PDDL, MIT |
+| Fed routing directory, SwiftRef BIC directory, CUSIP lists | no | restricted/proprietary — generate synthetically |
+| php-iban / ibankit registries | copy facts only | LGPL-3.0 / Apache-2.0 |
+| Stripe / PayPal test cards | numbers only, note source | no explicit grant (UNVERIFIED) |
+# Part 2 — Internet, vehicle, airline
+
+Verdict key: **EMBED** = safe to vendor into an MIT repo; **EMBED+ATTR** = ok with a notice; **AVOID** = licence incompatible or unclear.
+
+## 4. Internet
+
+### IANA registries (all of them)
+- Licence: IANA/IETF statement at https://www.iana.org/help/licensing-terms — "IANA and IETF intend that the Protocol Registries may be freely used by any party for any purpose", data dedicated under **CC0 1.0**. No attribution required. **EMBED**.
+- All registries below share this licence; ship one `SOURCES`/NOTICE line pointing at that URL.
+
+| Registry | CSV URL | Rows (2026-09-16) | Notes |
+|---|---|---|---|
+| Media types | `https://www.iana.org/assignments/media-types/<type>.csv`, type ∈ application, audio, font, haptics, image, message, model, multipart, text, video (also `example`) | application 1800, audio 165, font 6, haptics 3, image 88, message 27, model 42, multipart 17, text 105, video 97 (header excluded) ≈ **2350** | Fields: `Name,Template,Reference`. Template is the full `type/subtype`; some rows are DEPRECATED/OBSOLETED in Name (filter). Format: RFC 6838 §4.2 — `type "/" subtype`, restricted chars, max 127 chars each, case-insensitive. |
+| HTTP status codes | https://www.iana.org/assignments/http-status-codes/http-status-codes-1.csv | 75 rows, **64 assigned** (306 and 418 are "(Unused)"; 104 is TEMPORARY) → 62 usable | Fields: `Value,Description,Reference`. Value is 3 digits 1xx–5xx. |
+| IPv4 special-purpose | https://www.iana.org/assignments/iana-ipv4-special-registry/iana-ipv4-special-registry-1.csv | 26 blocks | Fields: `Address Block,Name,RFC,Allocation Date,Termination Date,Source,Destination,Forwardable,Globally Reachable,Reserved-by-Protocol`. |
+| IPv6 special-purpose | https://www.iana.org/assignments/iana-ipv6-special-registry/iana-ipv6-special-registry-1.csv | 27 blocks | Same fields. Contains `2001:db8::/32` and `3fff::/20`. |
+| Service names / ports | https://www.iana.org/assignments/service-names-port-numbers/service-names-port-numbers.csv | 14,534 rows; 11,731 with both name and a single port; **6,102 distinct named ports**; 6,008 tcp-named; 707 tcp names < 1024 | Fields: `Service Name,Port Number,Transport Protocol,Description,Assignee,Contact,Registration Date,Modification Date,Reference,Service Code,Unauthorized Use Reported,Assignment Notes`. Port Number may be empty, single, or a range `a-b`. Well-known = 0–1023, registered 1024–49151, dynamic 49152–65535 (RFC 6335). ~1.1 MB; embed only the named-port subset. |
+| TLDs | https://data.iana.org/TLD/tlds-alpha-by-domain.txt | **1,401** TLDs (first line is a `# Version …` comment) | Upper-case ASCII; IDNs as `XN--…` punycode. Same IANA CC0 terms (UNVERIFIED that data.iana.org is explicitly covered by the licensing page; it is the same registry data). Lower-case on use. |
+
+### MIME extension mapping: jshttp/mime-db
+- https://github.com/jshttp/mime-db — **MIT** (GitHub API spdx MIT; npm `license: MIT`), version 1.54.0 on npm.
+- `db.json` (CDN: https://cdn.jsdelivr.net/npm/mime-db/db.json): object keyed by lower-case type → `{source, extensions[], compressible, charset}`.
+- Counts (1.54.0): **2,522 types**, **1,015 with extensions**; source = iana 2,136, apache 275, nginx 13, custom 98.
+- Sources: IANA registry (CC0), Apache httpd `mime.types` (Apache-2.0), nginx `mime.types` (BSD-2). **EMBED+ATTR** (keep the MIT copyright notice). Data updates are not semver-breaking per README.
+
+### IPv4 / IPv6 generation rules (RFCs, no data file needed)
+- RFC 5737 documentation: `192.0.2.0/24` (TEST-NET-1), `198.51.100.0/24` (TEST-NET-2), `203.0.113.0/24` (TEST-NET-3). https://www.rfc-editor.org/rfc/rfc5737.html
+- RFC 1918 private: `10.0.0.0/8`, `172.16.0.0/12`, `192.168.0.0/16`. Also useful: `100.64.0.0/10` CGNAT (RFC 6598), `198.18.0.0/15` benchmarking (RFC 2544), `169.254.0.0/16` link-local.
+- IPv6 documentation: `2001:db8::/32` (RFC 3849) and `3fff::/20` (RFC 9637, "expands on the existing 2001:db8::/32 … with the reservation of an additional, larger prefix"). ULA `fc00::/7` (use `fd00::/8`), link-local `fe80::/10`.
+- Validation: generate inside those blocks; avoid network/broadcast (`.0`/`.255`) if the consumer expects host addresses. Format IPv6 per RFC 5952 (lower-case hex, `::` once, longest zero run).
+
+### MAC OUI (IEEE MA-L)
+- CSV: https://standards-oui.ieee.org/oui/oui.csv (3.8 MB, **40,161 rows**, fields `Registry,Assignment,Organization Name,Organization Address`; Assignment is 6 hex chars). Server returns HTTP 418 to non-browser user agents; fetch with a browser UA. Also MA-M `oui28.csv`, MA-S `oui36.csv`, CID `cid.csv`.
+- Licence: no licence on the IEEE pages themselves (site footer "© IEEE – All rights reserved" is generic). Debian's `ieee-data` copyright file (https://metadata.ftp-master.debian.org/changelogs/main/i/ieee-data/unstable_copyright) records IEEE's 2014 statement: *"IEEE does not assert any copyright in the OUI Public Listing or attempt to restrict distribution of the listing in any way. The IEEE Registration Authority does, however, strongly encourage those who use the list to obtain it directly from IEEE…"* — same text in the FSF directory. Debian ships it in `main` on that basis. **EMBED+ATTR** (cite the statement); the org-address column is not needed — keep `Assignment,Organization Name` only (~1 MB → few hundred KB). UNVERIFIED: no current IEEE page restates it (regauth FAQ and IPR pages do not mention it).
+- Alternative needing no data: locally administered unicast MAC — first octet's bit 1 (U/L) = 1, bit 0 (I/G) = 0 → second hex digit ∈ {2, 6, A, E}, e.g. `x2:xx:xx:xx:xx:xx`. Guaranteed never to collide with a vendor OUI. Notations: `01:23:45:67:89:ab`, `01-23-45-67-89-AB`, Cisco `0123.4567.89ab`.
+
+### User-agent strings
+| Source | Licence | Verdict |
+|---|---|---|
+| https://www.useragents.me/ | No licence stated anywhere on the page ("© 2022–2025 Useragents.me"); `/api` returns 404 now; JSON blobs embedded in page, weekly updates | **AVOID** (unlicensed = all rights reserved) |
+| https://explore.whatismybrowser.com/useragents/explore/ + API legal https://developers.whatismybrowser.com/api/about/legal/ | Proprietary: "You may NOT share the unique database download URL, or the downloaded file itself … EXCEPT for making it publicly available on the internet" | **AVOID** |
+| npm `user-agents` (intoli) https://github.com/intoli/user-agents | **BSD-2-Clause**, v2.1.185, daily-rebuilt dataset from Intoli traffic, weighted by frequency; fields `userAgent, platform, vendor, appName, deviceCategory, screenWidth/Height, viewportWidth/Height, connection{…}, oscpu, cpuClass, pluginsLength` | **EMBED+ATTR** (keep BSD notice). Count not stated on the page (UNVERIFIED; dataset is a few thousand rows). |
+| npm `top-user-agents` (microlinkhq) https://github.com/microlinkhq/top-user-agents | **MIT** (GitHub API). Top 100 UA strings from microlink.io traffic, weekly; JSON: https://cdn.jsdelivr.net/gh/microlinkhq/top-user-agents@master/src/index.json (+ desktop.json, mobile.json) | **EMBED+ATTR** — smallest clean option |
+- None of the above is CC0. Alternative: generate UAs from templates (Chrome/Firefox/Safari/Edge strings are fixed-form; only version numbers vary), no data needed.
+
+### Programming languages: GitHub Linguist
+- https://raw.githubusercontent.com/github-linguist/linguist/main/lib/linguist/languages.yml — **MIT** (repo LICENSE, "Copyright (c) 2017 GitHub, Inc."). **EMBED+ATTR**.
+- **835** top-level entries; **562** with `type: programming` (others: data, markup, prose). Per-entry fields: `type, color, extensions, aliases, tm_scope, ace_mode, language_id, group, interpreters, …`.
+
+### Operating systems
+- No open dataset found beyond awesome-lists. OS names/versions are facts — hand-author a small table (Windows 10/11, Windows Server, macOS versions+names, Ubuntu/Debian/Fedora/Arch/Alpine, iOS, Android, FreeBSD…). `user-agents` (BSD-2) carries `platform`/`oscpu` values if a data-driven list is wanted.
+
+### Database engines / collations
+- Engine names: hand-author (facts).
+- MySQL manual (https://dev.mysql.com/doc/refman/8.4/en/preface.html): Oracle terms — "you may not … copy, reproduce … distribute … any part" except distributing the docs with the software. **AVOID copying tables from MySQL docs.** Instead dump `SHOW COLLATION` from a `mysql:8.4.x` container (server output is factual, not the manual) — ~286 collations / 41 charsets in 8.x (UNVERIFIED count). PostgreSQL docs are under the PostgreSQL licence (permissive, keep notice) — `pg_collation` dump or docs both fine. MariaDB KB is CC BY-SA — avoid.
+
+## 5. Vehicle
+
+### NHTSA vPIC
+- API base `https://vpic.nhtsa.dot.gov/api/vehicles/`, `?format=json|xml|csv`. Endpoints verified: `GetAllMakes` (**12,363 makes**, fields `Make_ID,Make_Name`; 612 KB JSON, mostly small trailer/custom builders — filter with `GetMakesForVehicleType/car` → 195 makes, `/truck` → 207), `GetModelsForMake/honda` (362 models, `Make_ID,Make_Name,Model_ID,Model_Name`), `DecodeWMI/1HG`, `GetWMIsForManufacturer/honda` (45 WMIs, fields `WMI,Name,Country,VehicleType,Id,…`), `GetVehicleVariableList`, `GetVehicleVariableValuesList/<display name>` (must be the display name, URL-encoded, e.g. `Fuel%20Type%20-%20Primary`; the camel-case name returns 0 rows).
+- Value lists (counts): **Fuel Type - Primary 14**, **Body Class 71**, **Transmission Style 12**, **Drive Type 23**, **Vehicle Type 9**. Fields `Id,Name`.
+- Bulk: https://vpic.nhtsa.dot.gov/downloads/ — `vPICList_lite_2026_08.bak.zip` (MS SQL Server 2019+ backup, 190 MB, updated 2026-08-14), also `.plain.zip` / `.custom.zip`. Page says standalone DB "limited to VIN decoding"; makes/models/variables still via API. Contains `Wmi`, `Make`, `Model`, `WMIYearValidChars` tables (UNVERIFIED table names).
+- Licence: FAQ https://vpic.nhtsa.dot.gov/api/home/index/faq — "No, NHTSA is a government agency and the services provided on the API are free for use by the public as an offering as a part of our Open Data initiatives"; no rate-limit number stated, batch jobs asked to run at night. Work of the US federal government → public domain under 17 USC §105; data.gov entry lists licence as "unknown". **EMBED** (note source; no attribution legally required).
+
+### VIN (ISO 3779 / 49 CFR 565, North America)
+- 17 chars from `0-9 A-H J-N P R-Z` (**I, O, Q never**). Positions: 1–3 WMI, 4–8 VDS, 9 check digit, 10 model year, 11 plant, 12–17 serial (last 4 must be digits for US, positions 12–14 alphanumeric — 49 CFR 565.15).
+- Check digit: transliterate `A=1 B=2 C=3 D=4 E=5 F=6 G=7 H=8 J=1 K=2 L=3 M=4 N=5 P=7 R=9 S=2 T=3 U=4 V=5 W=6 X=7 Y=8 Z=9`, digits as-is; weights `8 7 6 5 4 3 2 10 0 9 8 7 6 5 4 3 2`; sum mod 11 → digit, remainder 10 → `X`.
+- Model year (pos 10): excludes `I O Q U Z 0`; `A=2010 … H=2017 J=2018 K=2019 L=2020 M=2021 N=2022 P=2023 R=2024 S=2025 T=2026 V=2027 W=2028 X=2029 Y=2030`, `1=2031…9=2039` (30-year cycle; `1–9` also = 2001–2009). Source: https://en.wikipedia.org/wiki/Vehicle_identification_number (rules are facts; do not copy prose).
+
+### WMI list
+- SAE J1044 / SAE WMI database is paid (https://www.sae.org/standards/j1044_202501-world-manufacturer-identifier) — **AVOID**.
+- Wikipedia "List of WMIs" — CC BY-SA — **AVOID**.
+- Use vPIC: `GetWMIsForManufacturer/<name or id>` per manufacturer, or the `Wmi` table in the standalone `.bak`. Public domain. Country prefix rule (first char): `1,4,5` USA, `2` Canada, `3` Mexico, `J` Japan, `K` Korea, `S` UK, `W` Germany, `Y` Sweden/Finland, `Z` Italy, `L` China (ISO 3780 facts).
+
+### US licence plate formats
+| Source | Licence | Verdict |
+|---|---|---|
+| openalpr `runtime_data/postprocess/us.patterns` (https://github.com/openalpr/openalpr) | **AGPL-3.0** | **AVOID** (291 lines, `@`=letter `#`=digit — useful only as a reference to check hand-authored patterns) |
+| Wikipedia "United States license plate designs and serial formats" | CC BY-SA | **AVOID copying**; formats themselves are facts |
+| jonnii/platekit | MIT, but SVG rendering, no serial patterns | not useful |
+- No MIT/CC0 per-state pattern list found. Recommendation: hand-author a per-state table (facts); current standard passenger formats as listed by Wikipedia (verify against DMV pages before use): AL `0AXXXXX`/`00AXXXX`, AK `ABC 123`, AZ `XXX 1XX`, AR `ABC 12D`, CA `1ABC123` (digit-3 letters-3 digits), CO `ABC-D12`, CT `AB·12345`, DE `123456`, DC `AB-1234`, FL `ABC D12`, GA `ABC1234`, HI `ABC 123`, ID `A 1234U`, IL `AB 12345`, IN `123ABC`, IA `ABC 123`, KS `1234ABC`, KY `ABC123`, LA `123 ABC`, ME `123·ABC`, MD `1AB2345`, MA `1ABC 23`, MI `ABC 1234`, MN `ABC-123`, MS `ABC 123`, MO `AB1 C2D`, MT `0-AB1234`, NE `ABC 123`, NV `123·A45`, NH `123 4567`, NJ `D12-ABC`, NM `123-ABC`, NY `ABC-1234`, NC `ABC-1234`, ND `123 ABC`, OH `ABC 1234`, OK `ABC-123`, OR `123 ABC`, PA `ABC1234`, RI `1AB 234`, SC `123ABC`, SD `0A1 234`, TN `ABC 1234`, TX `ABC-1234`, UT `A12 3BC`, VT `ABC 123`, VA `ABC-1234`, WA `ABC1234`, WV `X1A 2345`, WI `ABC-1234`, WY `1A-123A`. Most states omit I/O/Q from serials (state-specific, UNVERIFIED per state).
+
+## 6. Airline
+
+### OurAirports
+- https://ourairports.com/data/ — "All data is released to the Public Domain"; mirror repo https://github.com/davidmegginson/ourairports-data is **Unlicense** (GitHub API). Nightly updates. **EMBED**.
+- Raw URLs: `https://davidmegginson.github.io/ourairports-data/{airports,airport-frequencies,runways,navaids,countries,regions}.csv`.
+
+| File | Rows | Fields |
+|---|---|---|
+| airports.csv (12.7 MB) | **86,089**; by type: small_airport 42,734, heliport 23,216, closed 13,524, medium 4,106, seaplane_base 1,273, large 1,174, balloonport 62 | `id,ident,type,name,latitude_deg,longitude_deg,elevation_ft,continent,iso_country,iso_region,municipality,scheduled_service,icao_code,iata_code,gps_code,local_code,home_link,wikipedia_link,keywords` |
+| — with `iata_code` | **9,055** (large 1,171, medium 3,398, small 4,233, seaplane 153, heliport 100); `scheduled_service=yes` 4,335; US with IATA 2,036 (873 large/medium) | Suggested subset: `iata_code != '' AND type IN (large_airport, medium_airport)` → 4,569 rows |
+| countries.csv | **249** | `id,code,name,continent,wikipedia_link,keywords` |
+| regions.csv | **3,987** | `id,code,local_code,name,continent,iso_country,wikipedia_link,keywords` (code = ISO 3166-2, e.g. `US-KS`) |
+- No airline file. Derived: datasets/airport-codes on datahub (PDDL) — no need, use upstream.
+
+### OpenFlights
+- https://openflights.org/data.php — airports/airlines/routes/planes under **ODbL 1.0** (+ DbCL); airline/plane data partly from Wikipedia (GFDL/CC BY-SA). 5,888 airlines (2012). **AVOID** (share-alike). npm `airline-codes` (npow, ISC) is a straight OpenFlights sync → inherits ODbL → **AVOID**.
+
+### Airline IATA/ICAO codes — open options
+| Source | Licence | Count | Verdict |
+|---|---|---|---|
+| Wikidata SPARQL (https://query.wikidata.org/sparql) | **CC0** (https://www.wikidata.org/wiki/Wikidata:Licensing) | **2,928** items with IATA code (P229), **3,644** with ICAO code (P230) | **EMBED**. Query: `SELECT ?item ?itemLabel ?iata ?icao ?callsign WHERE { ?item wdt:P229 ?iata . OPTIONAL { ?item wdt:P230 ?icao } OPTIONAL { ?item wdt:P432 ?callsign } SERVICE wikibase:label { bd:serviceParam wikibase:language "en" } }`. Includes defunct airlines — filter `MINUS { ?item wdt:P576 ?dissolved }` and/or P31 = Q46970 (airline). |
+| OpenTravelData `optd_airlines.csv` (https://github.com/opentraveldata/opentraveldata) | **CC BY 4.0** | 1,620 rows; `^`-separated; fields `pk^env_id^validity_from^validity_to^3char_code^2char_code^num_code^name^name2^alliance_code^alliance_status^type^wiki_link^flt_freq^alt_names^bases^key^version^parent_pk_list^successor_pk_list` | **EMBED+ATTR** (attribution notice required) |
+- Format rules: IATA designator = 2 alphanumeric chars (`[A-Z0-9]{2}`, at least one letter in practice; a third optional char exists in the standard but is unused); "controlled duplicates" share a code between non-overlapping regional carriers. ICAO designator = 3 letters `[A-Z]{3}`, unique. Accounting/prefix code = 3 digits (ticket number prefix, e.g. `016` United).
+
+### Aircraft types
+| Source | Licence | Verdict |
+|---|---|---|
+| ICAO Doc 8643 (https://www.icao.int/operational-safety/doc-8643-aircraft-type-designators) | ICAO copyright notice: "None of the materials … may be used, reproduced or transmitted … without permission in writing from ICAO"; API Data Service is paid (25 free trial calls) | **AVOID** — including GitHub mirrors such as ColtJD45/icao-aircraft-designator-list (MIT-labelled, 7,388 rows, fields `manufacturer,model,type_designator,description,engine_type,engine_count,wtc`) because the underlying data is scraped ICAO content (UNVERIFIED provenance) |
+| OpenSky aircraft database (https://opensky-network.org/data/aircraft; CSV https://s3.opensky-network.org/data-samples/metadata/aircraftDatabase.csv) | "The aircraft database is unlicensed and does not fall under our terms of use … offered as is"; built from registries, openflights.org and ICAO Doc 8643; citation requested; page says it is no longer up to date | **AVOID** ("unlicensed" ≠ open; upstream includes ODbL/ICAO). Fields for reference: `icao24,registration,manufacturericao,manufacturername,model,typecode,serialnumber,linenumber,icaoaircrafttype,operator,operatorcallsign,operatoricao,operatoriata,owner,…,engines,…,categoryDescription` |
+| Wikidata | **CC0** | **550** items with ICAO aircraft type designator (P8305); also IATA aircraft code P9040 (UNVERIFIED property id), manufacturer P176 | **EMBED**. Query: `SELECT ?item ?itemLabel ?icao ?mfrLabel WHERE { ?item wdt:P8305 ?icao . OPTIONAL { ?item wdt:P176 ?mfr } SERVICE wikibase:label { bd:serviceParam wikibase:language "en" } }` |
+- Format: ICAO type designator 2–4 alphanumerics (`[A-Z0-9]{2,4}`, e.g. `B738`, `A20N`, `E190`); IATA aircraft code 3 alphanumerics (`738`, `32N`).
+
+### Airport / flight / seat format rules
+- IATA airport code `[A-Z]{3}`; ICAO location indicator `[A-Z0-9]{4}` (US `K…`, Canada `C…`, UK `EG…`, Sweden `ES…`). Use OurAirports rows so codes are real.
+- Flight number: 2-char IATA designator + 1–4 digits, no leading zeros in display (`QF9`, `AA1234`); systems cap at 0001–9999. Conventions (not rules): 1–999 mainline, 3000–5999 regional affiliates, ≥6000 codeshares, 8xxx charters, ≥9000 ferry/positioning. ICAO form: 3-letter designator + 1–4 alphanumerics (`AFR1`). Source: https://en.wikipedia.org/wiki/Flight_number.
+- Seat: `row + letter`. Rows typically 1–~60 (A380 up to ~90); many airlines skip 13 (and 14/17 on some carriers). Letters `A–K` **skipping I** (looks like 1/l); narrow-body 3-3 = `ABC DEF` (A/F windows, C/D aisles); wide-body 3-4-3 = `ABC DEFG HJK`, 2-4-2 = `AB DEFG JK`. Regex: `^[1-9][0-9]?[A-HJK]$`.
+# Part 3 — Science/nature, Text, Dates: open data sources for an MIT fake-data library
+
+Researched 2026-09-17. "OK to embed" = redistributable inside an MIT repo with the stated notice. Anything not fetched first-hand is marked **UNVERIFIED**.
+
+## 7. Science / nature
+
+### Periodic table
+
+| Source | Licence | Embed in MIT repo | Fields | Rows | Format |
+|---|---|---|---|---|---|
+| PubChem periodic table — CSV https://pubchem.ncbi.nlm.nih.gov/rest/pug/periodictable/CSV, JSON https://pubchem.ncbi.nlm.nih.gov/rest/pug/periodictable/JSON (page: https://pubchem.ncbi.nlm.nih.gov/periodic-table/) | US-government work, public domain. NLM policy (https://www.ncbi.nlm.nih.gov/home/about/policies/): "Information that is created by or for the US government on this site is within the public domain … may be freely distributed and copied. However, it is requested that in any subsequent use of this work, NLM be given appropriate acknowledgment." | Yes; add an acknowledgment line ("Element data: PubChem/NLM"). | `AtomicNumber, Symbol, Name, AtomicMass, CPKHexColor, ElectronConfiguration, Electronegativity, AtomicRadius, IonizationEnergy, ElectronAffinity, OxidationStates, StandardState, MeltingPoint, BoilingPoint, Density, GroupBlock, YearDiscovered` (17) | 118 (Z=1..118, verified) | CSV; JSON is `{Table:{Columns:{Column:[...]},Row:[{Cell:[...]}]}}` |
+| NIST periodic table https://www.nist.gov/pml/periodic-table-elements | Public domain (17 USC 105); NIST asks for citation (https://www.nist.gov/open/copyright-fair-use-and-licensing-statements-srd-data-software-and-technical-series-publications) | Yes, but | PDF only (no CSV/JSON) — not usable as a data source | 118 | PDF |
+| Bowserinator/Periodic-Table-JSON https://github.com/Bowserinator/Periodic-Table-JSON | **CC BY-SA 3.0** (LICENSE.md, verified; data derived from Wikipedia) | **No** — ShareAlike is incompatible with MIT redistribution | 33 keys incl. `name, symbol, number, atomic_mass, category, phase, period, group, block, shells, electron_configuration, summary, cpk-hex, …` | **119** (includes hypothetical Ununennium) | JSON, CSV |
+| faker-js `src/locales/en/science/chemical_element.ts` https://github.com/faker-js/faker (branch `next`) | MIT (LICENSE verified) | Yes, keep MIT notice | `{symbol, name, atomicNumber}` | 118 | TS |
+
+Validation rules: `AtomicNumber` 1..118 unique; `Symbol` 1–2 letters, first uppercase, unique; `Name` unique. PubChem leaves numeric fields empty for Z≥104 where unknown; `ElectronConfiguration` for Og is `"[Rn]7s2 7p6 5f14 6d10 (predicted)"` and `StandardState` is `"Expected to be a Gas"` — strip "(predicted)"/"Expected to be a" if you want an enum. `AtomicMass` for unstable elements is the mass number of the most stable isotope as a bare decimal (e.g. `295.216`), no brackets. `CPKHexColor` is 6 hex digits without `#`, empty for some.
+
+### SI units
+
+- Source: BIPM SI Brochure, 9th ed. (2019; current revision v4.01, 2026) https://www.bipm.org/en/publications/si-brochure ; PDF https://www.bipm.org/documents/20126/41483022/SI-Brochure-9-EN.pdf
+- Licence: **CC BY 4.0** — stated on the brochure page and on BIPM's copyright page (https://www.bipm.org/en/copyright): content may be adapted, copied, used in commercial products with "appropriate acknowledgement of the BIPM and its source"; BIPM name/logo must not imply endorsement. Verified via the HTML pages; the PDF's own front-matter licence line is **UNVERIFIED** (could not text-extract the PDF here).
+- Embed: Yes — the unit *names/symbols* are facts (not copyrightable); attribute "SI Brochure, BIPM, CC BY 4.0" anyway.
+- Data (cross-checked against https://en.wikipedia.org/wiki/International_System_of_Units):
+  - 7 base units: second s (time), metre m (length), kilogram kg (mass), ampere A (electric current), kelvin K (thermodynamic temperature), mole mol (amount of substance), candela cd (luminous intensity).
+  - 22 coherent derived units with special names: radian rad, steradian sr, hertz Hz, newton N, pascal Pa, joule J, watt W, coulomb C, volt V, farad F, ohm Ω, siemens S, weber Wb, tesla T, henry H, degree Celsius °C, lumen lm, lux lx, becquerel Bq, gray Gy, sievert Sv, katal kat.
+  - 24 prefixes: quetta Q 10^30, ronna R 10^27, yotta Y 10^24, zetta Z 10^21, exa E 10^18, peta P 10^15, tera T 10^12, giga G 10^9, mega M 10^6, kilo k 10^3, hecto h 10^2, deca da 10^1, deci d 10^-1, centi c 10^-2, milli m 10^-3, micro µ 10^-6, nano n 10^-9, pico p 10^-12, femto f 10^-15, atto a 10^-18, zepto z 10^-21, yocto y 10^-24, ronto r 10^-27, quecto q 10^-30.
+- Format rules for valid output: unit symbols are case-sensitive (`s` vs `S`, `m` vs `M`); prefix and unit symbol are joined without a space (`kJ`, `µs`); a space separates number and unit (`12 kg`), except the degree/minute/second of plane angle; `kg` already carries a prefix — never `µkg`; micro is Greek mu U+03BC (U+00B5 micro sign is the legacy compatibility character); ohm is U+03A9 (U+2126 OHM SIGN is deprecated); `°C` not `° C`.
+- MIT alternative: faker-js `en/science/unit.ts` — 29 `{name, symbol}` objects, MIT.
+
+### Animals by class
+
+- **Wikidata** (https://www.wikidata.org/wiki/Wikidata:Licensing): all structured data **CC0** — no attribution required. Embed: Yes.
+  - Feasibility, live SPARQL (https://query.wikidata.org/sparql), pattern `?t wdt:P31 wd:Q16521; wdt:P105 wd:Q7432; wdt:P171* wd:<class>; wdt:P1843 ?cn FILTER(LANG(?cn)='en')` (taxon, rank species, parent-taxon chain, English common name):
+    - Mammalia Q7377: **5,759** species with ≥1 English common name.
+    - Aves Q5113: **12,432**.
+    - Reptilia Q10811: **16,369** — suspiciously large; Wikidata's parent chain likely pulls birds/Sauropsida in. Use Squamata Q122422 + Testudines Q223044 + Crocodilia Q1387 (or check the chain) instead.
+    - Insecta Q1390 and Plantae Q756: **timed out (502/504) every attempt** — trees too big for the 60 s public endpoint. Options: paginate by order (`P171` one level at a time), use the weekly JSON dump, or take GBIF instead.
+  - Sample rows (mammals): `Notoryctes typhlops` → "Central Desert Marsupial Mole", "Itjaritjari", "Marsupial Mole", "Southern marsupial mole", "Southern Marsupial Mole" — i.e. **several P1843 values per taxon, inconsistent casing, and indigenous-language names tagged `en`**. Validation: pick one name per taxon (prefer the one matching the item's English `rdfs:label`), normalise case, drop names that equal the scientific name.
+- **GBIF Backbone Taxonomy** https://www.gbif.org/dataset/d7dddbf4-2cf0-4f39-9b2a-bb099caae36c — licence **CC BY 4.0** (verified via https://api.gbif.org/v1/dataset/d7dddbf4-2cf0-4f39-9b2a-bb099caae36c, `license: http://creativecommons.org/licenses/by/4.0/legalcode`); required citation text: "GBIF Secretariat (2023). GBIF Backbone Taxonomy. Checklist dataset https://doi.org/10.15468/39omei". Embed: Yes with that citation. Vernacular names come via `/v1/species/{key}/vernacularNames`, each row from a *different* source checklist with its own licence (some CC BY-NC) — record and check per row. GBIF's site terms page (https://www.gbif.org/terms) was unreachable (timeout) — **UNVERIFIED** beyond the API field.
+- **faker-js** `src/locales/en/animal/*` (MIT; counts = quoted string lines on branch `next`): bird 821, snake 533, dog 493, cow 467, horse 342, rodent 161, insect 129, fish 95, cat 55, cetacean 52, rabbit 49, type 44 (generic: "bat, bear, bee, …, zebra"), pet_name 42, crocodilia 24, bear 8, lion 7. These are breeds/species mixed with common names, no class field; fine as MIT drop-ins. (A WebFetch summariser claimed 1,036 dogs; the regex count is 493 — treat the exact count as UNVERIFIED.)
+
+### Plants
+
+- Same Wikidata approach (CC0) — but the Plantae query timed out; paginate by order or family (`?t wdt:P171 ?family . ?family wdt:P105 wd:Q35409`) or use the dump.
+- GBIF Plantae (kingdomKey 6) vernacular names, CC BY 4.0, same per-row-licence caveat.
+- faker-js has **no** plant list (en locale dirs: airline, animal, app, book, color, commerce, company, database, date, finance, food, hacker, internet, location, lorem, medical, music, person, phone_number, science, team, vehicle, word).
+- USDA PLANTS database (public domain, has common names) — **UNVERIFIED**, not fetched.
+
+### Planets / moons
+
+- NASA NSSDCA Planetary Fact Sheet https://nssdc.gsfc.nasa.gov/planetary/factsheet/ — 10 columns (Mercury, Venus, Earth, Moon, Mars, Jupiter, Saturn, Uranus, Neptune, Pluto) × 20 rows (mass 10^24 kg, diameter km, density, gravity, escape velocity, rotation period, day length, distance from Sun, perihelion, aphelion, orbital period, orbital velocity, inclination, eccentricity, obliquity, mean temp °C, surface pressure, number of moons, rings Y/N, magnetic field Y/N). HTML table only. NASA content "generally not subject to copyright in the United States" (https://www.nasa.gov/nasa-brand-center/images-and-media/); NASA asks to be acknowledged; the NASA insignia must not be used. Embed: Yes with "Source: NASA NSSDCA".
+- Wikidata (CC0): planets = `?p wdt:P31/wdt:P279* wd:Q634; wdt:P397 wd:Q525` → Mercury Q308, Venus Q313, Earth Q2, Mars Q111, Jupiter Q319, Saturn Q193, Uranus Q324, Neptune Q332 (plus hypothetical Theia Q1053432 — filter out). Moons = `?m wdt:P397 <planet>; wdt:P31/wdt:P279* wd:Q2537` (natural satellite; note **Q2199 is "dwarf planet"**, not moon): Saturn 165, Jupiter 85, Uranus 29, Neptune 16, Earth 5, Mars 2. Caveats: Wikidata lags IAU counts (Saturn has 274 recognised moons as of 2025 — UNVERIFIED figure); Earth's 5 includes quasi-satellites; provisional designations (`S/2004 S 3`) are labels too — filter `^S/\d{4}` if you want proper names only.
+
+### Blood type distribution
+
+- **US** (Stanford Blood Center https://stanfordbloodcenter.org/donate-blood/blood-donation-facts/blood-types/, verified; cites AABB Technical Manual 18th ed.): O+ 37.4 %, A+ 35.7 %, B+ 8.5 %, AB+ 3.4 %, O− 6.6 %, A− 6.3 %, B− 1.5 %, AB− 0.6 % (sums to 100.0). Facts — freely embeddable; cite Stanford/AABB.
+- American Red Cross https://www.redcrossblood.org/donate-blood/blood-types.html — **UNVERIFIED**: site returns HTTP 403 (Akamai) to both WebFetch and the headless browser. Search snippets say the Red Cross gives type O by ethnicity (Latino 57 %, African American 51 %, Caucasian 45 %) rather than an 8-way national table.
+- **Global**: no authoritative figure found. Wikipedia "Blood type distribution by country" gives a population-weighted row (O+ 38.4, A+ 27.3, B+ 8.1, AB+ 2.0, O− 13.1, A− 8.1, B− 2.0, AB− 0.01) that its own editors flag "unreliable source"; WorldAtlas gives O+ 42, A+ 31, B+ 15, AB+ 5, O− 3, A− 2.5, B− 1, AB− 0.5 with no source. Recommend shipping only the US table, or per-country tables from Wikipedia (facts; Wikipedia prose is CC BY-SA but the numbers are cited to national blood services — cite those).
+- Validation: ABO ∈ {A, B, AB, O}; Rh ∈ {+, −}; weights must sum to 100 ± rounding.
+
+## 8. Text
+
+### English word lists with part of speech
+
+| Source | Licence | Embed | Counts | Format |
+|---|---|---|---|---|
+| Princeton WordNet 3.0 / 3.1 https://wordnet.princeton.edu/license-and-commercial-use ; download https://wordnetcode.princeton.edu/wn3.1.dict.tar.gz | "WordNet License" (SPDX `WordNet`), BSD-style: permission to use/copy/modify/distribute "for any purpose and without fee or royalty … provided that … the following copyright notice and statements, including the disclaimer … appear on ALL copies … including modifications". Notice text: "WordNet 3.0 Copyright 2006 by Princeton University. All rights reserved." + AS-IS disclaimer; no use of Princeton's name in advertising. | Yes — ship the LICENSE text alongside the data | WordNet 3.0 (wnstats(7WN), verified): unique strings noun 117,798 / verb 11,529 / adj 21,479 / adv 4,481 (total 155,287); synsets 82,115 / 13,767 / 18,156 / 3,621 | WNDB: `index.noun|verb|adj|adv` (one lemma per line: `lemma pos synset_cnt p_cnt [ptr…] sense_cnt tagsense_cnt synset_offset…`), `data.<pos>` (synsets with glosses). Multi-word lemmas use `_` for spaces; lower-case; entries can contain digits/apostrophes/hyphens. |
+| Open English WordNet 2025 https://en-word.net/ ; https://github.com/globalwordnet/english-wordnet | **CC BY 4.0**; cite "Open English WordNet … derived from Princeton WordNet" | Yes, with attribution | 135,969 words, 107,519 synsets (README, verified) | `english-wordnet-2025.zip` (WNDB, 9.2 MB), `english-wordnet-2025.xml.gz` (WN-LMF, 10.8 MB), `english-wordnet-2025-json.zip` (9.5 MB), `.ttl.gz` (16.9 MB) |
+| SCOWL v2 / ESDB https://github.com/en-wl/wordlist (branch v2) ; classic http://wordlist.aspell.net/ | Custom MIT-like (Copyright 2000-2026 Kevin Atkinson: use/copy/modify/distribute/sell "provided that the above copyright notice appears in all copies and that both the above copyright notice and this notice appear in supporting documentation"). Sources 12dicts + ENABLE2K are public domain; lists above size 80 add the UKACD notice; **POS assignments partly from WordNet, so the WordNet notice "MIGHT apply"** (their words). Australian-English parts carry an extra notice. | Yes, ship the Copyright file | Sizes 35 (small), 50 (medium), 60 (spell-check default), 70 (large), 80 (incl. game words), 85 (archaic). No per-size counts published in the v2 README. Classic SCOWL sizes 10–95 — counts **UNVERIFIED** (README not reachable). | v2: `scowl.db` (SQLite), `scowl.txt`, Python `libscowl`. ESDB rows carry POS, spelling (A/B/Z/C/D) and region codes. |
+| 12dicts http://wordlist.aspell.net/12dicts-readme/ | Public domain ("I explicitly release them to the public domain, but request acknowledgment"), **except** `2of12inf` and the `2+2+3*` lists, which depend on AGID and inherit its terms | Yes (PD lists); acknowledge Alan Beale | 3esl ≈22k, 6of12 ≈32k, 2of12 ≈41k, 2of12inf ≈82k, 3of6game ≈65k, 5d+2a ≈68k, 3of6all ≈83k, 2+2+3lem ≈84k, 2of5core ≈4.7k, 6phrase ≈22k, neol2016 ≈600 | Plain text, one word per line; marker suffixes `+ ! ^ &` (not POS). **No POS tags** in any current 12dicts list. |
+| Moby Part-of-Speech II https://www.gutenberg.org/ebooks/3203 ; file https://www.gutenberg.org/files/3203/files/mobypos.txt | Public domain ("Public Domain material by grant from the author, January, 2001", in the package README) | Yes, no notice needed | **233,356** lines (verified) | One entry per line: `word\CODES`, CRLF line ends (e.g. `A-line\NA`, `a tempo\Avh`). Codes in priority order: `N` noun, `p` plural, `h` noun phrase, `V` verb (usu. participle), `t` transitive verb, `i` intransitive verb, `A` adjective, `v` adverb, `C` conjunction, `P` preposition, `!` interjection, `r` pronoun, `D` definite article, `I` indefinite article, `o` nominative. Many entries are phrases, proper nouns, or 1990s-era; non-ASCII entries use a legacy 8-bit encoding (e.g. `a bon march\v` lost its `é`) — restrict to ASCII `[a-z]+` for a clean list. |
+| faker-js `en/word/*` (MIT) | MIT | Yes | adjective 1000, noun 1000, verb 1000, adverb 325, preposition 109, conjunction 51, interjection 46 | TS string arrays |
+
+### Word frequency lists
+
+| Source | Licence | Verdict |
+|---|---|---|
+| Google Books Ngram data v2/v3 https://storage.googleapis.com/books/ngrams/books/datasetsv3.html | **CC BY 3.0** ("This compilation is licensed under a Creative Commons Attribution 3.0 Unported License") | Embeddable with attribution. Format `ngram TAB year TAB match_count TAB volume_count`; 1-gram files split by leading letter, GBs each — aggregate offline to a top-N list. |
+| wordfreq https://github.com/rspeer/wordfreq | Code **Apache 2.0** (not MIT). Data: "may be redistributed under a Creative Commons Attribution-ShareAlike 4.0 license" (mix of Google Books, Wikipedia, OpenSubtitles, SUBTLEX, Leeds, ParaCrawl, Twitter); maintainer notes the CSV export "does not follow the CC-By-SA license". Project in sunset mode (data snapshot ≈2021). | **Avoid** embedding the data (ShareAlike). |
+| Peter Norvig count_1w.txt https://norvig.com/ngrams/ | Page says "Code … under the MIT license"; the **data** comes from the Google Web 1T corpus distributed by LDC under LDC terms — no data licence stated. | **UNVERIFIED / avoid**. 333,333 words, `word TAB count`, lowercase. |
+| hermitdave/FrequencyWords https://github.com/hermitdave/FrequencyWords | "MIT License for code. CC-by-sa-4.0 for content." | **Avoid** (ShareAlike). |
+
+### Lorem ipsum
+
+- Canonical paragraph (Letraset, 1966; public domain — garbled Cicero): "Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum."
+- Source: Cicero, *De finibus bonorum et malorum* 1.10.32–33 (45 BC), public domain; the "dolorem ipsum" fragment begins "Neque porro quisquam est qui dolorem ipsum quia dolor sit amet, consectetur, adipisci velit…". Reference: https://en.wikipedia.org/wiki/Lorem_ipsum
+- MIT word pool alternative: faker-js `en/lorem/word.ts` — 999 words.
+
+### Quote collections
+
+- Project Gutenberg #27889 https://www.gutenberg.org/ebooks/27889 — Bartlett, *Familiar Quotations*, **9th edition** (title page "NINTH EDITION", copyright 1875/1882/1891/1903) — not the 1919 10th. "Public domain in the USA." Plain text 3.1 MB, HTML zip 50 MB, EPUB.
+- Project Gutenberg #16732 https://www.gutenberg.org/ebooks/16732 — an early (Hurst & Co. reprint) edition, no edition number given. PD in USA.
+- 10th edition (1914/1919, ed. Nathan Haskell Dole) exists on archive.org (https://archive.org/details/bartlettsfamilia0000john_o4b4) — not found on Gutenberg; **UNVERIFIED** availability as clean text.
+- Either Bartlett text needs parsing (quote / author / source headings); no structured file. Wikiquote is CC BY-SA — avoid.
+- Validation: strip Gutenberg header/footer (`*** START OF THE PROJECT GUTENBERG EBOOK` … `*** END …`) before extraction; Gutenberg's trademark terms only bind if you keep the "Project Gutenberg" name on the output — strip it.
+
+### Colour names
+
+| Source | Licence | Embed | Count / format |
+|---|---|---|---|
+| CSS Color Module Level 4 §named colors https://www.w3.org/TR/css-color-4/#named-colors | W3C Document License (2023) https://www.w3.org/copyright/document-license-2023/ — copy/distribute allowed for any purpose; derivatives allowed only "to facilitate implementation of the technical specifications"; notice required: "Copyright © 2023 W3C®. This software or document includes material copied from or derived from [title and URI]." | Yes — the name→hex table is factual data and is copied into every browser; include the W3C notice line with the spec title + URI. | **148** names (verified: aliceblue … yellowgreen, incl. `rebeccapurple #663399`); all names lowercase ASCII, hex 6 lowercase digits. |
+| XKCD colour survey https://xkcd.com/color/rgb.txt | **CC0** (file header: `# License: https://creativecommons.org/publicdomain/zero/1.0/`) | Yes, no notice needed | **949** lines (verified). Format `name TAB #rrggbb TAB` (note the trailing tab). Max name length 26; all hex valid lowercase; no duplicate names. Contains crude names ("poo", "baby poop green", "diarrhea", "vomit") — apply a blocklist. |
+| faker-js `en/color/human.ts` | MIT | Yes | 31 names |
+| Pantone | Proprietary | **Avoid** | — |
+
+## 9. Dates
+
+### IANA tz database
+
+- Source https://www.iana.org/time-zones ; files https://data.iana.org/time-zones/tzdb/ (current version **2026d**, verified). LICENSE: "Unless specified below, all files in the tz code and data (including this LICENSE file) are in the public domain" (only `date.c`, `newstrftime.3`, `strftime.c` are BSD-3). Embed: Yes, no notice needed.
+- `zone1970.tab` — **312** data rows; tab-separated; UTF-8; `#` comments. Columns: (1) comma-separated ISO 3166-1 alpha-2 codes of countries overlapping the zone, most-populous first; (2) coordinates in ISO 6709 `±DDMM±DDDMM` or `±DDMMSS±DDDMMSS` (lat then lon, no separator); (3) TZ name (`Europe/Zurich`); (4) comment, present iff the country has several zones. One row per timezone where civil time has agreed since 1970.
+- `zone.tab` — **418** rows, older one-country-per-row table (column 1 is a single country code); kept for backward compatibility. Prefer `zone1970.tab`.
+- `backward` — **252** `Link TARGET LINK-NAME` lines mapping old/merged names (`US/Eastern`, `Asia/Calcutta`, `Europe/Kiev`) to current ones; also a few `Zone` entries. Use it to (a) accept aliases as input and (b) never emit them as output.
+- UTC offsets: do **not** embed — offsets depend on DST rules that change several times a year. Derive at runtime: `new Intl.DateTimeFormat('en-US', { timeZone, timeZoneName: 'longOffset' }).formatToParts(date)` → part `timeZoneName` = `GMT-05:00` (or `GMT` for zero) — strip `GMT`, treat bare `GMT` as `+00:00`. Validation of a generated zone name: `Intl.supportedValuesOf('timeZone')` (canonical names only; whether links like `US/Eastern` appear depends on the ICU build) or `try { new Intl.DateTimeFormat(undefined, { timeZone }) } catch { invalid }`, which accepts links too.
+
+### ISO 8601 layouts (https://en.wikipedia.org/wiki/ISO_8601)
+
+| Layout | Extended | Basic | Example |
+|---|---|---|---|
+| Calendar date | `YYYY-MM-DD` | `YYYYMMDD` | 2009-01-06 |
+| Ordinal date | `YYYY-DDD` | `YYYYDDD` | 1981-095 |
+| Week date | `YYYY-Www-D` | `YYYYWwwD` | 2009-W01-1 (weeks start Monday, week 1 contains 4 Jan) |
+| Time | `hh:mm:ss[.fff]` | `hhmmss` | 13:47:30 (24 h; `T` prefix optional standalone) |
+| UTC / offset | `Z` or `±hh:mm` | `±hhmm`, `±hh` | 2007-04-05T14:30Z, 2024-06-01T09:00-05:00 |
+| Date-time | date `T` time offset | | 2007-04-05T14:30:00+02:00 |
+| Duration | `PnYnMnDTnHnMnS` / `PnW` | | P3Y6M4DT12H30M5S |
+| Interval | start`/`end, start`/`duration, duration`/`end | | 2007-03-01T13:00:00Z/2008-05-11T15:30:00Z |
+| Recurring | `Rn/`interval | | R5/2008-03-01T13:00:00Z/P1Y2M10DT2H30M |
+
+RFC 3339 profile (what JSON/APIs usually mean): extended forms only, `T`/`Z` may be lowercase, space allowed instead of `T`, `-00:00` = unknown offset, no durations/intervals/week/ordinal dates. Generate RFC 3339 by default.
+
+### Unicode CLDR (locale month/weekday names, date patterns)
+
+- Licence: **Unicode License v3** (SPDX `Unicode-3.0`), OSI-approved 2023-11-17 (https://opensource.org/license/unicode-3-0). Text https://www.unicode.org/license.txt — MIT-style: use/copy/modify/sell "provided that either (a) this copyright and permission notice appear with all copies of the Data Files or Software, or (b) this copyright and permission notice appear in associated Documentation"; no use of the Unicode name in advertising. Embed: Yes; put the notice in LICENSE/NOTICE.
+- npm: `cldr-core`, `cldr-dates-full`, `cldr-numbers-full` (peer of dates), etc.; current **48.2.0** (CLDR 48, verified via npm registry, `license: Unicode-3.0`); repo https://github.com/unicode-org/cldr-json .
+- Path is **`cldr-dates-full/main/<locale>/ca-gregorian.json`** (not `dates/gregorian.json`), rooted at `main.<locale>.dates.calendars.gregorian`. Keys (verified for `en`): `months`, `days`, `quarters`, `dayPeriods`, `eras`, `dateFormats`, `dateSkeletons`, `timeFormats`, `timeSkeletons`, `dateTimeFormats`, `dateTimeFormats-atTime`, `dateTimeFormats-relative`.
+  - `months.{format|stand-alone}.{abbreviated|narrow|wide}` keyed `"1".."12"`; `days.{format|stand-alone}.{abbreviated|narrow|short|wide}` keyed `sun..sat`.
+  - `dateFormats`: en = full `EEEE, MMMM d, y`, long `MMMM d, y`, medium `MMM d, y`, short `M/d/yy`.
+  - `timeFormats`: en = full `h:mm:ss a zzzz`, long `h:mm:ss a z`, medium `h:mm:ss a`, short `h:mm a` — **the `a` is preceded by U+202F NARROW NO-BREAK SPACE**; `*-alt-ascii` variants use a plain space. Pick one spelling (ASCII) and ignore the other.
+  - `dateTimeFormats.{full|long|medium|short}` = glue like `{1}, {0}` ({1} date, {0} time); `availableFormats` = skeleton→pattern map (`yMMMd` → `MMM d, y`), plus `intervalFormats`, `appendItems`.
+  - `dayPeriods.format.abbreviated`: `am: AM`, `pm: PM`, plus `midnight`, `noon`, `morning1`… variants.
+- Pattern letters follow LDML (https://unicode.org/reports/tr35/tr35-dates.html#Date_Field_Symbol_Table): `y` year, `M`/`L` month (format/stand-alone), `d` day, `E`/`c` weekday, `h` 1–12, `H` 0–23, `a` AM/PM, `z`/`zzzz` zone name, `x`/`X` ISO offset; literal text in single quotes. Same patterns drive `Intl.DateTimeFormat` internally, so runtime `Intl` can replace embedding for month/weekday names if bundle size matters.
+
+## Summary of licence verdicts
+
+- Embed freely (PD/CC0): PubChem periodic table, NIST, NASA fact sheet, Wikidata, Moby POS, 12dicts (PD lists), Bartlett via Gutenberg, XKCD colours, IANA tz, Lorem ipsum/Cicero.
+- Embed with notice: WordNet 3.x (WordNet License), Open English WordNet (CC BY 4.0), SCOWL/ESDB (custom MIT-like), GBIF backbone (CC BY 4.0 + citation), BIPM SI (CC BY 4.0), CSS named colours (W3C notice), CLDR (Unicode-3.0), Google Books ngrams (CC BY 3.0), faker-js lists (MIT).
+- Avoid: Bowserinator periodic table (CC BY-SA 3.0), wordfreq data (CC BY-SA 4.0), hermitdave (CC BY-SA 4.0), Norvig count_1w (LDC-derived, unstated), Wikiquote (CC BY-SA), Pantone.
+- UNVERIFIED: American Red Cross figures (site blocks fetches), any global blood-type table, BIPM PDF front-matter licence line, classic SCOWL per-size counts, USDA PLANTS, IAU moon counts, exact faker dog count, Bartlett 10th-edition text availability.
+# Part 4 — ISO lists, commerce, books/music/food
+
+Researched 2026-09-17. "Embed?" = can the data ship inside an MIT-licensed repo. Anything marked UNVERIFIED was not confirmed against the primary source.
+
+## 10. ISO lists
+
+### ISO 3166-1 countries
+
+| Source | Licence | Embed in MIT repo? | Rows | Fields | Format |
+|---|---|---|---|---|---|
+| [datasets/country-codes](https://github.com/datasets/country-codes) | PDDL (README: "Public Domain Dedication and License"; GitHub API detects no licence file) | Yes, no attribution required. README caveat: ISO itself says its list is "for internal use and non-commercial purposes free of charge" — the repo argues no rights subsist in a list of facts | 249 | ISO3166-1-Alpha-2/-Alpha-3/-numeric, official_name_en, UNTERM names (ar/zh/en/fr/ru/es short+formal), Dial, TLD, Capital, Continent, Languages, ISO4217 currency code/name/numeric/minor_unit, Region/Sub-region/Intermediate names+codes (M49), FIPS, IOC, FIFA, MARC, ITU, WMO, GAUL, EDGAR, Geoname ID, wikidata_id, CLDR display name, is_independent, LDC/LLDC/SIDS flags | CSV `data/country-codes.csv` |
+| [annexare/Countries](https://github.com/annexare/Countries) (npm `countries-list` 3.4.1, 2026-07-14) | MIT | Yes; keep MIT notice | 252 countries, 185 languages (incl. non-ISO entries like AC, XK) | per country: name, native, phone[] (calling codes), continent, capital, currency[], languages[], optional alias[], partOf; separate ISO 4217 currency table (name, native, symbol, numeric, decimals); languages: name, native | TS source, exported JSON/CSV/SQL |
+| [Debian iso-codes](https://salsa.debian.org/iso-codes-team/iso-codes) | LGPL-2.1-or-later (Debian copyright file) | Not safely as embedded data in an MIT repo — LGPL applies to the data files; you would have to ship it as a separately licensed component with LGPL notice. Avoid unless accepting that | 3166-1: 249 | alpha_2, alpha_3, numeric, name, official_name, common_name, flag | JSON `data/iso_3166-1.json` (raw URL needs `?inline=false`) |
+| [mledoze/countries](https://github.com/mledoze/countries) | ODbL-1.0 | No — share-alike + attribution; derived databases must be ODbL. Avoid | ~250 | common/official names, cca2/cca3/ccn3/cioc, tld, currencies, idd (calling codes), capital, region/subregion, languages, translations, latlng, borders, area, demonyms, flag emoji, UN member | JSON/CSV/XML/YAML |
+| [restcountries](https://gitlab.com/restcountries/restcountries) | MPL-2.0 (source repo; hosted API v5 is proprietary/API-key) | File-level copyleft: a copied data file stays MPL and must carry the notice; OK to ship next to MIT code but not to relicense. Prefer PDDL/MIT sources | 250+ | name, tld, cca2, ccn3, cca3, cioc, fifa, independent, status, unMember, currencies, idd, capital, capitalInfo, altSpellings, region, subregion, continents, languages, translations, latlng, landlocked, borders, area, flag (emoji), demonyms, flags, coatOfArms, population, maps, gini, car, postalCode, startOfWeek, timezones | JSON `src/main/resources/countriesV3.1.json` |
+| [lukes/ISO-3166-Countries-with-Regional-Codes](https://github.com/lukes/ISO-3166-Countries-with-Regional-Codes) | CC BY-SA 4.0 (LICENSE.md) | No — share-alike. Avoid | 249 | name, alpha-2, alpha-3, country-code, iso_3166-2, region, sub-region, intermediate-region + codes | CSV/JSON/XML (all, slim-2, slim-3) |
+| [stefangabos/world_countries](https://github.com/stefangabos/world_countries) | Data: CC BY-SA 4.0 (README, GitHub licence detection); npm `world_countries_lists` package.json says LGPL-3.0-or-later — contradictory | No — share-alike either way. Avoid | 249 world / 193 UN; subdivisions 5046 | id (numeric), alpha2, alpha3, name (37 languages); subdivisions: country, code, name, name_en, type, parent | JSON/CSV/PHP/SQL/XML |
+| [i18n-iso-countries](https://www.npmjs.com/package/i18n-iso-countries) 7.14.0 | MIT | Yes | 78 language files | alpha2, alpha3, numeric, localized names | JSON per language |
+| Wikipedia ISO 3166-1 | CC BY-SA 4.0 | No (share-alike) | — | — | — |
+| [ISO OBP](https://www.iso.org/obp) | Proprietary; ISO grants free use only for "internal use and non-commercial purposes" (quoted in datasets/country-codes README) | No | — | — | — |
+
+- Flag emoji: derive, no data needed — for alpha-2 `XY`, emit U+1F1E6 + (X − 'A') and U+1F1E6 + (Y − 'A') (regional indicator symbols).
+- Recommendation: `datasets/country-codes` (PDDL) covers every requested field (name, alpha2, alpha3, numeric, calling code, TLD, capital, currency, languages) in one CSV; flag emoji derived. `annexare` (MIT) is the fallback for native names/calling codes if PDDL's ISO caveat worries you; it lacks TLD.
+- Validation rules: alpha-2 `^[A-Z]{2}$`, alpha-3 `^[A-Z]{3}$`, numeric `^\d{3}$` (zero-padded, keep as string); calling code 1–3 digits, NANP countries share `1`; TLD `^\.[a-z]{2}$` (ccTLD = lowercase alpha-2, exceptions: GB uses `.uk`).
+
+### ISO 639 languages
+
+| Source | Licence | Embed? | Rows | Fields | Format |
+|---|---|---|---|---|---|
+| [langs](https://www.npmjs.com/package/langs) 2.0.0 (2017, unmaintained) | MIT | Yes | 184 | name, local, 1 (639-1), 2, 2T, 2B, 3 | `data.js` array |
+| [iso-639-1](https://www.npmjs.com/package/iso-639-1) 3.1.6 (2026-07-02) | MIT | Yes | 183 | code, name, nativeName | `src/data.js` |
+| annexare/Countries languages | MIT | Yes | 185 | code (639-1), name, native | TS/JSON |
+| Debian iso-codes 639-2 / 639-3 / 639-5 | LGPL-2.1+ | See above (LGPL) | 639-2: 487; 639-3: 7923 | 639-2: alpha_2, alpha_3, bibliographic, name, common_name; 639-3: + inverted_name, scope, type | JSON |
+| [SIL ISO 639-3 tables](https://iso639-3.sil.org/code_tables/download_tables) | SIL terms: may incorporate into software (commercial or not) with attribution to iso639-3.sil.org, must not modify identifiers, and product "does not provide a means to redistribute the code set" | Risky — a git repo with the table IS a means to redistribute; avoid embedding the full table. Codes themselves are facts | 7927 | Id, Part2b, Part2t, Part1, Scope, Language_Type, Ref_Name, Comment | tab-delimited `iso-639-3.tab` |
+| [Library of Congress ISO 639-2](https://www.loc.gov/standards/iso639-2/ascii_8bits.html) | US federal government work → public domain in the US | Yes | ~487 (matches Debian 639-2 count) | pipe-delimited: alpha3-B, alpha3-T, alpha2, English name, French name — UNVERIFIED: loc.gov is behind a Cloudflare bot check (WebFetch, curl and Playwright all blocked); field layout from memory | `ISO-639-2_utf-8.txt` |
+
+- Recommendation: `langs` or `iso-639-1` (MIT) for 639-1 names; both are small and derived from public code tables. Anything 639-3-sized is LGPL or SIL-restricted.
+- Rules: 639-1 `^[a-z]{2}$`, 639-2/3 `^[a-z]{3}$`; 639-2 has B/T pairs (e.g. `fre`/`fra`, `ger`/`deu`) — pick one consistently (T codes = 639-3).
+
+### ISO 3166-2 subdivisions
+
+| Source | Licence | Embed? | Rows | Fields |
+|---|---|---|---|---|
+| Debian iso-codes 3166-2 | LGPL-2.1+ | LGPL caveat | 5046 | code, name, type, parent |
+| [olahol/iso-3166-2.json](https://github.com/olahol/iso-3166-2.json) | package.json says ISC; no LICENSE file; GitHub detects none; data scraped from eQuest xls (provenance unclear); last push 2021 | No — unclear licence and provenance | 237 countries / 3807 divisions | `{alpha2: {name, divisions: {code: name}}}` |
+| [esosedi/3166](https://github.com/esosedi/3166) (npm `iso3166-2-db` 2.3.11) | MIT | Yes; data mixes GeoNames (CC BY 4.0), Wikipedia (CC BY-SA), OSM (ODbL) — the MIT label may not survive scrutiny for the Wikipedia/OSM-derived parts. UNVERIFIED how much is derived from each | "all countries and regions" (count not stated) | ISO 3166-2 + FIPS codes, names in 12 languages, admin level, GeoNames/OSM/Wikipedia/WOF ids |
+| stefangabos subdivisions | CC BY-SA 4.0 | No | 5046 | country, code, name, name_en, type, parent |
+| [US Census state.txt](https://www2.census.gov/geo/docs/reference/state.txt) | US government work → public domain | Yes | 57 (50 states + DC + 6 territories) | STATE (FIPS), STUSAB, STATE_NAME, STATENS |
+
+- Recommendation: for an en_US library, ship only the US list (Census/USPS, public domain: 50 states + DC, optionally territories) — ISO 3166-2 code = `US-` + USPS abbreviation. Skip the ~5000-row world list; every open compilation is LGPL, CC BY-SA or unclear.
+- Rule: `^[A-Z]{2}-[A-Z0-9]{1,3}$`.
+
+## 11. Commerce
+
+### Word lists (product name / adjective / material)
+
+- [@faker-js/faker](https://github.com/faker-js/faker) 10.6.0 — MIT (LICENSE covers code and locale data; copyright Faker contributors 2022-2025 and Marak Squires 2011-2020; acknowledges Ruby Faker and Perl Data::Faker, both MIT). No separate data licence.
+- Provenance policy in [CONTRIBUTING.md](https://github.com/faker-js/faker/blob/next/CONTRIBUTING.md): "Faker must not contain copyrighted materials"; facts (finite known lists) are OK; compilations must not be copied from a single source (Wikipedia, "most popular" articles, government sites) because "a compilation of facts can be copyrighted" — contributors are told to draw on multiple sources and their own judgement.
+- Verdict: reusing faker's `en` commerce lists (`product_name.adjective/material/product`, `department`) is fine under MIT — keep the MIT notice with the copied data. The lists are short curated words (no external dataset). Nothing in the repo or docs restricts locale-data reuse beyond MIT.
+
+### Garment and shoe sizes
+
+- Letter sizes XS–XXL: EN 13402-3 letter codes with chest/bust ranges (Wikipedia [EN 13402](https://en.wikipedia.org/wiki/EN_13402), CC BY-SA — but the table is a handful of facts; the standard itself is paywalled). Men chest / women bust cm: XXS 70–78 / 66–74, XS 78–86 / 74–82, S 86–94 / 82–90, M 94–102 / 90–98, L 102–110 / 98–107, XL 110–118 / 107–119, XXL 118–129 / 119–131, 3XL 129–141 / 131–143.
+- US numeric women's sizes (0–20, even) come from ASTM D5585 (paywalled); no open dataset. Just generate the even-number series.
+- Shoe sizes: no open dataset found (search hits are ML demos and retailer charts). Use the formulas from Wikipedia [Shoe size](https://en.wikipedia.org/wiki/Shoe_size) (facts, not copyrightable): EU (Paris point) ≈ 1.5 × foot cm + 2 (i.e. 3/20 × mm + 2); UK adult ≈ 3 × foot in − 23; US men ≈ 3 × foot in − 22 = UK + 1; US women (common) = US men + 1.5 (FIA scale: 3 × foot in − 21); Mondopoint = foot length mm. ISO/TS 19407:2023 has the official conversion tables (paywalled). Generate from a foot length (men 24–31 cm, women 21–27 cm) in half-size steps so all systems agree.
+
+### Check-digit rules (all verified against Wikipedia articles; algorithms are facts)
+
+| Code | Rule |
+|---|---|
+| EAN-13 / UPC-A / GTIN-8/12/13/14 | Weights 3,1,3,1… starting from the rightmost data digit (the one next to the check digit); check = (10 − sum mod 10) mod 10. UPC-A = EAN-13 with a leading 0 (GS1 prefix of a 12-digit GTIN is `0` + first two digits). GTIN-14 first digit is an indicator 1–8 (packaging level) or 9 (variable measure); 0 is not valid there. |
+| UPC-A number system (first digit) | 0,1,6,7,8,9 regular; 2 variable-weight; 3 drugs (NDC); 4 in-store/local; 5 coupons. |
+| ISBN-10 | Σ(weights 10..2 × first 9 digits) + check ≡ 0 mod 11; check = (11 − sum mod 11) mod 11, 10 → `X`. |
+| ISBN-13 | EAN-13 with prefix 978 or 979. 979 groups in use: 979-8 USA, 979-10 France, 979-11 Korea, 979-12 Italy. 978 groups 0 and 1 = English-language. For an en_US generator: `978-0…`, `978-1…`, `979-8…`. The official range file (`https://www.isbn-international.org/export_rangemessage.xml`) is XML only and the site says "You must not republish material from our site … without our written permission" — do not embed it; group/registrant length only matters for hyphenation, not validity. |
+| ISSN | Format `NNNN-NNNC`; weights 8..2 over first 7 digits, sum mod 11; check = 0 if remainder 0 else 11 − remainder; 10 → `X`. EAN-13 form: `977` + 7 ISSN digits (no check) + 2 variant digits + EAN check. |
+| IMEI | 15 digits: TAC 8 (RBI 2 + 6) + serial 6 + Luhn check (ISO/IEC 7812; from the right, double every second digit, sum digits, total ≡ 0 mod 10). SVN never enters the check. IMEISV = 14 digits + 2-digit SVN, no check. RBI `00` = test IMEI (Wikipedia [Reporting Body Identifier](https://en.wikipedia.org/wiki/Reporting_Body_Identifier)); live RBIs: 01 CTIA/PTCRB (US), 35 TÜV SÜD/BABT (UK), 86 TAF (China), 99 GHA. For fake data use TAC `00xxxxxx` so it can never match a real handset — UNVERIFIED: third-party pages claim the GSMA TS.06 test-TAC shape is `00 44` + 4 digits; the TS.06 PDF could not be text-extracted here. |
+| ASIN | 10 chars `[A-Z0-9]{10}`; non-books start `B0` + 8 alphanumerics (`^B0[A-Z0-9]{8}$`); books reuse their ISBN-10 verbatim. No check digit (Wikipedia). |
+| SKU | Not standardised (Wikipedia). Convention only: 8–12 uppercase alphanumerics, e.g. `[A-Z]{3}-\d{4}-[A-Z]{2}`; avoid leading 0 and O/I if you want scanner-safe. |
+
+### GS1 prefixes
+
+- Official table: [gs1.org/standards/id-keys/company-prefix](https://www.gs1.org/standards/id-keys/company-prefix) (rendered via Playwright; WebFetch gets 403). Copyright notice ([gs1.org/terms-use](https://www.gs1.org/terms-use)): reproduction allowed only "in unaltered form … for your personal, non-commercial use or use within your organisation"; "You are not permitted to re-transmit, distribute or commercialise the information or material without seeking prior written approval from GS1." → Do not copy the GS1 page. The prefix-to-country mapping is factual and also on Wikipedia ([List of GS1 country codes](https://en.wikipedia.org/wiki/List_of_GS1_country_codes), CC BY-SA 4.0 — avoid verbatim copy). Safe path: embed only the US/special ranges you need, written from the facts below.
+- Ranges relevant to an en_US generator (from the GS1 page): 001–019, 030–039, 060–139 GS1 US (UPC-A compatible); 020–029 restricted circulation within a region; 040–049 restricted within a company; 050–059 GS1 US reserved; 200–299 restricted circulation (region); 952 "used for demonstrations and examples of the GS1 system"; 977 ISSN; 978–979 ISBN; 980 refund receipts; 981–983 coupons (common currency); 990–999 coupons. Prefixes do not identify country of origin. Selected others: 300–379 France, 400–440 Germany, 450–459 & 490–499 Japan, 500–509 UK, 690–699 China, 730–739 Sweden, 750 Mexico, 754–755 Canada, 760–769 Switzerland, 800–839 Italy, 840–849 Spain, 870–879 Netherlands, 880–881 South Korea, 890 India, 930–939 Australia, 940–949 New Zealand.
+- Generator rule: for fake UPC/EAN use prefix `952` (GS1 demo) or a `2xx` restricted-circulation prefix — never a real company prefix; for US-looking UPC-A use number system 0–1/6–8 with a random 5-digit manufacturer code and accept that it may collide with real products.
+
+## 12. Books, music, food
+
+### Project Gutenberg catalog
+
+- URL: `https://www.gutenberg.org/cache/epub/feeds/pg_catalog.csv` (20 MB) / `pg_catalog.csv.gz` (5.3 MB), regenerated weekly; also RDF (`rdf-files.tar.bz2`, 121 MB) and MARC.
+- Fields: `Text#, Type, Issued, Title, Language, Authors, Subjects, LoCC, Bookshelves` (comma CSV, quoted; Authors as `Last, First, birth-death`, `;`-separated).
+- Count (2026-09-17): 79 381 rows; Type: Text 78 130, Sound 1 114, Dataset 89, Image 33, other 12; Language: en 62 860, fr 4 190, fi 3 684, de 2 425, it 1 110.
+- Licence: each RDF record carries `<cc:Work><cc:license rdf:resource="https://creativecommons.org/publicdomain/zero/1.0/"/>` → catalog metadata is CC0 1.0 (verified on `cache/epub/1/pg1.rdf`). The CSV itself has no header notice; treat as CC0 by the same source. Terms of use: do not hammer their servers (bulk feeds are the sanctioned path), "Project Gutenberg" is a trademark — royalties for commercial use of the *name*; do not brand the data file with it beyond a source citation.
+- Embed: yes (CC0). For a fake-data library ship a filtered subset (Type=Text, Language=en, title + first author, maybe 5–10k rows) rather than 20 MB.
+
+### Music genres and instruments
+
+| Source | Licence | Embed? | Count | Notes |
+|---|---|---|---|---|
+| [MusicBrainz genre list](https://musicbrainz.org/genres) | Genre is a core entity (schema doc lists 13 core entities incl. Genre) → core data dump `mbdump.tar.bz2` is CC0. The genre→entity *associations* come via user tags (CC BY-NC-SA 3.0) — do not use those | Yes, names only | 2 202 (`/ws/2/genre/all?fmt=json`, `genre-count`) | WS API needs a descriptive User-Agent; list is a JSON array of `{id, name, disambiguation}` |
+| [Discogs data dumps](https://data.discogs.com/) | CC0 | Yes | genre/style vocab is embedded in release XML (monthly dumps, GBs) — no standalone style list; UNVERIFIED count (~15 genres, ~600 styles from memory) | Impractical to extract; prefer MusicBrainz |
+| ID3v1 genres | Names 0–79 from the 1999 ID3v1 spec, 80–125 Winamp, 126–191 Winamp 5.6 (2010) — a de-facto spec table, treated as public domain facts | Yes | 192 (0–191) | Wikipedia [List of ID3v1 genres](https://en.wikipedia.org/wiki/List_of_ID3v1_genres) (CC BY-SA page; the list itself is a spec enumeration). id3.org returned HTTP 500 during research — UNVERIFIED against the original spec page |
+| Wikidata instruments | CC0 | Yes | 9 194 items `P31/P279* Q34379` with English labels; 2 372 with a Hornbostel–Sachs number (P1762) | Filter by P1762 for a clean orchestral/folk instrument list |
+| Wikidata music genres | CC0 | Yes | 6 619 (`Q188451`) | Noisier than MusicBrainz |
+
+### USDA FoodData Central
+
+- Downloads: [fdc.nal.usda.gov/download-datasets](https://fdc.nal.usda.gov/download-datasets/) (April 2026 release):
+  - Foundation Foods: JSON 459 KB zip / 6.5 MB; CSV 3.7 MB zip / 32 MB — 394 foods (API `totalHits`, dataType=Foundation).
+  - SR Legacy (final, April 2018): JSON 12.3 MB zip / 205 MB; CSV 6.7 MB zip / 54 MB — 7 793 foods.
+  - FNDDS 2021-2023: CSV 200 MB zip / 1.6 GB — 5 432 foods.
+  - Branded: CSV 428 MB zip / 2.9 GB — 433 403 foods.
+  - Full: CSV 460 MB zip / 3.1 GB.
+- Licence ([API guide](https://fdc.nal.usda.gov/api-guide/)): "USDA FoodData Central data are in the public domain and they are not copyrighted. They are published under CC0 1.0 Universal (CC0 1.0)". No permission needed; USDA *requests* listing FoodData Central as source and notifying them. Suggested citation: "U.S. Department of Agriculture, Agricultural Research Service. FoodData Central, 2019. fdc.nal.usda.gov."
+- Embed: yes (CC0). Ship a derived list (SR Legacy `description` + `food_category`) not the raw dump.
+- CSV structure (UNVERIFIED — the field-description PDF could not be text-extracted; from memory): `food.csv` (fdc_id, data_type, description, food_category_id, publication_date), `food_category.csv` (id, code, description), `nutrient.csv` (id, name, unit_name, nutrient_nbr, rank), `food_nutrient.csv` (id, fdc_id, nutrient_id, amount, …), `food_portion.csv`, `measure_unit.csv`, `sr_legacy_food.csv` (fdc_id, NDB_number), `foundation_food.csv`.
+
+### Dishes, cuisines, drinks
+
+| Source | Licence | Embed? | Count | Notes |
+|---|---|---|---|---|
+| Wikidata dishes (`Q746549`) | CC0 | Yes | 7 480 with English label | SPARQL `?i wdt:P31/wdt:P279* wd:Q746549` |
+| Wikidata cuisines (`Q1968435`) | CC0 | Yes | 219 | |
+| Wikidata cocktails (`Q134768`) | CC0 | Yes | 305 | |
+| Wikidata beer (`Q44` subclass tree) / wine (`Q282`) | CC0 | Yes | 420 / 3 025 | Wine tree is mostly appellations/brands; filter by P279 depth |
+| IBA official cocktails | IBA site "© IBA 2026 – All rights reserved"; Wikipedia list CC BY-SA 4.0 | Names only are facts (102 cocktails: 34 Unforgettables, 34 Contemporary Classics, 34 New Era, 2024 revision); do not copy recipes/descriptions | 102 | Safest: take the names from Wikidata (`P31 Q134768`) rather than the IBA page |
+| [Open Brewery DB](https://github.com/openbrewerydb/openbrewerydb) | MIT (LICENSE, © 2025 Open Brewery DB) | Yes | 11 931 breweries (US 8 308, DE 1 445, AU 514, BE 478, CA 283, NZ 243) | `breweries.csv`: id (UUID), name, brewery_type (micro 5 906, brewpub 3 947, closed 644, planning 635, regional 240, contract 208, large 137, proprietor, taproom, bar, nano, cidery, beergarden), address_1..3, city, state_province, postal_code, country, phone, website_url, longitude, latitude. Real businesses — fine for names, but generating "fake" data with real addresses/phones may be undesirable |
+| Wikipedia lists (cuisines, dishes, IBA) | CC BY-SA 4.0 | No verbatim copies | | |
+
+## Summary of safe picks
+
+- Countries: `datasets/country-codes` (PDDL) or `annexare/Countries` (MIT); flag emoji derived from alpha-2.
+- Languages: `langs` / `iso-639-1` (MIT, 639-1 only).
+- Subdivisions: US-only list from Census (public domain); skip world list.
+- Commerce words: faker `en` lists (MIT). Check digits: implement from the rules above; GS1 prefix table: hand-write the few ranges needed, use `952`/`2xx` for generated barcodes; ISBN range file: do not embed.
+- Books: Gutenberg `pg_catalog.csv` (CC0), filtered.
+- Music: MusicBrainz genre names (CC0, 2 202), ID3v1 list (192), Wikidata instruments (CC0).
+- Food: USDA FDC SR Legacy/Foundation descriptions (CC0, cite USDA), Wikidata dishes/cuisines (CC0), Open Brewery DB (MIT).
+- Avoid: mledoze (ODbL), lukes (CC BY-SA), stefangabos (CC BY-SA/LGPL), Debian iso-codes (LGPL), SIL 639-3 table (no-redistribution clause), GS1/ISBN-International pages (all rights reserved), Wikipedia tables verbatim (CC BY-SA), MusicBrainz tag associations (CC BY-NC-SA).

family.go

@@ -97,7 +97,7 @@ func (d *draws) rowOf(t *table) int {
 // pinRow pins row r of t where it agrees with the rows pinned before it.
 func (d *draws) pinRow(t *table, r int) error {
 	if pr, ok := d.pinned(t); ok && pr != r {
-		return fmt.Errorf("%s and %s are two rows of %s", t.selectorSpelling(pr), t.selectorSpelling(r), t.category)
+		return fmt.Errorf("%s and %s are two rows of %s", t.selectorSpelling(pr), t.selectorSpelling(r), t.path)
 	}
 	for a := t.parentT; a != nil; a = a.parentT {
 		if pa, ok := d.pinned(a); ok && !t.under(r, a, pa) {

fejkdata.go

@@ -5,8 +5,8 @@
 // then generate values by path:
 //
 //	f, _ := fejkdata.New(fejkdata.WithSeed(42))
-//	f.Fake("sv_SE.address")          // "Storgatan 12\n234 56 Göteborg"
+//	f.Fake("sv_SE.address")          // "Järvedsvägen 43\n891 77 Järved"
-//	f.Fake("sv_SE.address.locality") // "Göteborg"
+//	f.Fake("sv_SE.address.locality") // "Sundbyberg": a second call draws afresh
 //
 // Several sources merge in order, the last winning a name clash, so custom data
 // layers over the built-ins. The JSON template format is documented in the README.
@@ -29,8 +29,8 @@ import (
 //go:embed data
 var shippedFS embed.FS
 
-// MaxRepeat caps a repeat, and the renders nested repeats multiply to along any
+// MaxRepeat caps a repeat, and caps the renders nested repeats multiply to along
-// path; the CLI's --repeat shares it.
+// any one path; the CLI's --repeat shares it.
 const MaxRepeat = 1 << 20
 
 var _ [^uint(0)>>63 - 1]struct{} // 64-bit only, per the README's Decisions
@@ -40,8 +40,8 @@ var ErrNoData = errors.New("no data: WithoutShippedData needs at least one WithD
 
 // Generator generates fake data from a loaded namespace tree. Create one with [New].
 // It is safe for concurrent use; a seeded sequence is reproducible only when drawn
-// from one goroutine. The compiled tree is immutable after [New], and Fake,
+// from one goroutine. The compiled tree is immutable after [New], so List reads it
-// NewTemplate and List read it concurrently without a lock.
+// without a lock; mu guards the rest, which every other entry point takes.
 type Generator struct {
 	mu         sync.Mutex
 	rand       *session

fejkdata_test.go

@@ -81,3 +81,14 @@ func TestDifferentSeedsDiffer(t *testing.T) {
 	}
 	t.Fatal("seeds 1 and 2 produced identical sequences")
 }
+
+// fakeTemplate renders an inline template against a loaded generator, so its
+// references resolve.
+func fakeTemplate(t *testing.T, f *Generator, s string) string {
+	t.Helper()
+	got, err := f.FakeTemplate(s)
+	if err != nil {
+		t.Fatalf("FakeTemplate(%s) = %v", s, err)
+	}
+	return got
+}