Document the clock-only fence and the samordningsnummer day, and restore two-space continuations

CHANGELOG.md

@@ -32,7 +32,7 @@ replacement, and each removed path, column or flag.
   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`.
+  `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 quoted Go layout. `sv_SE.date`, `en_US.date`, `sv_SE.time` and
   `en_US.time` render through them, so `date.year`, `date.month`, `date.day` and

README.md

@@ -540,7 +540,8 @@ between the two days is reachable, so a layout with a clock draws the time too.
 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 quoted one; a layout naming no field, which is text;
-and for `time` a layout naming a date field, naming `date`. `{seq()}` spans `Fake`
+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.
 
@@ -617,7 +618,7 @@ its groups by name; the unnamed group spans them all.
 
 ```json
 { "format": "{payer} pays {payee}; signed {signature}",
-    "payer": { "format": "{/sv_SE.person.first} {/sv_SE.person.last}", "drawGroup": "payer" },
+  "payer": { "format": "{/sv_SE.person.first} {/sv_SE.person.last}", "drawGroup": "payer" },
   "payee": "{/sv_SE.person.first} {/sv_SE.person.last}",
   "signature": { "format": "{/sv_SE.person.last}", "drawGroup": "payer" } }
 ```
@@ -1079,11 +1080,13 @@ renamed or retyped line is a major.
   load, since nothing could then select it.
 - **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
+  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.
+  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

bench_test.go

@@ -26,12 +26,12 @@ func benchPath(b *testing.B, dir, path string) {
 	}
 }
 
-func BenchmarkPerson(b *testing.B)     { benchPath(b, "data", "sv_SE.person") }
+func BenchmarkPerson(b *testing.B)       { benchPath(b, "data", "sv_SE.person") }
-func BenchmarkAddress(b *testing.B)    { benchPath(b, "data", "sv_SE.address") }
+func BenchmarkAddress(b *testing.B)      { benchPath(b, "data", "sv_SE.address") }
-func BenchmarkWord(b *testing.B)       { benchPath(b, "data", "sv_SE.word") }
+func BenchmarkWord(b *testing.B)         { benchPath(b, "data", "sv_SE.word") }
-func BenchmarkCreditcard(b *testing.B) { benchPath(b, "data", "misc.creditcard") }
+func BenchmarkCreditcard(b *testing.B)   { benchPath(b, "data", "misc.creditcard") }
-func BenchmarkSSN(b *testing.B)        { benchPath(b, "data", "sv_SE.personnummer") }
+func BenchmarkPersonnummer(b *testing.B) { benchPath(b, "data", "sv_SE.personnummer") }
-func BenchmarkUUIDv7(b *testing.B)     { benchPath(b, "data", "misc.uuid") }
+func BenchmarkUUIDv7(b *testing.B)       { benchPath(b, "data", "misc.uuid") }
 
 func tmpData(b *testing.B, name, body string) string {
 	b.Helper()

builtins.go

@@ -472,17 +472,14 @@ func iban(r rng, cc string) string {
 
 const dayLayout = "2006-01-02"
 
-// The two instants a layout is proved against: alike in nothing, so a layout that
+// Alike in no field; layoutDay differs from layoutProbe in its date fields only.
-// formats them alike names no field, and one that tells their days apart at the same
-// clock names a date field.
 var (
 	layoutProbe  = time.Date(2001, 2, 3, 4, 5, 6, 0, time.UTC)
 	layoutProbe2 = time.Date(2010, 11, 12, 13, 14, 15, 0, time.UTC)
 	layoutDay    = time.Date(2010, 11, 12, 4, 5, 6, 0, time.UTC)
 )
 
-// layoutArg is the Go layout a quoted arg holds; unquoted, it is refused naming the
+// layoutArg is the Go layout a quoted arg holds, refused when unquoted or constant.
-// quoted spelling, since a layout may carry the comma that splits args.
 func layoutArg(a string) (string, error) {
 	if len(a) < 2 || a[0] != '\'' || a[len(a)-1] != '\'' {
 		return "", fmt.Errorf("layout %s is not quoted; write '%s'", a, strings.Trim(a, "'"))
@@ -503,8 +500,8 @@ func layoutOf(a string) string {
 	return layout
 }
 
-// layoutArity checks a call that ends in a layout takes n args; an unquoted layout
+// layoutArity checks a call ending in a layout takes n args, naming the quoted
-// carrying a comma splits into more, so the error names its quoted spelling.
+// layout when an unquoted one split into more.
 func layoutArity(name string, n int, a []string) error {
 	if len(a) == n {
 		return nil
@@ -531,8 +528,14 @@ func dateArgs(_ map[string]node, a []string) error {
 	if !from.Before(to) {
 		return fmt.Errorf("date(from,to,layout): from %s is not before to %s", a[0], a[1])
 	}
-	_, err = layoutArg(a[2])
+	layout, err := layoutArg(a[2])
-	return err
+	if err != nil {
+		return err
+	}
+	if layoutProbe.Format(layout) == layoutDay.Format(layout) {
+		return fmt.Errorf("date(from,to,layout): '%s' names no date field; write time('%s')", layout, layout)
+	}
+	return nil
 }
 
 func timeArg(_ map[string]node, a []string) error {

builtins_test.go

@@ -310,6 +310,7 @@ func TestBuiltinDateArgs(t *testing.T) {
 		`"{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) {

data-import/names-se.py

@@ -19,7 +19,7 @@ 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", "de", "den", "der", "di", "du", "la", "le", "van", "von"}
+PARTICLES = {"af", "av", "da", "de", "del", "den", "der", "di", "dos", "du", "la", "le", "van", "von"}
 
 
 def cased(name):

data-import/names-us.py

@@ -23,6 +23,7 @@ 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):
@@ -41,7 +42,7 @@ def csv_or_zip_rows(data, member):
 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:
+        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
 

data/en_US/first-name.tsv

@@ -379,7 +379,6 @@ Ayla	f	23690
 Aylin	f	12970
 Azalea	f	7052
 Azariah	m	3910
-Baby	m	6059
 Bailee	f	14798
 Bailey	f	102509
 Bailey	m	20062
@@ -468,6 +467,7 @@ Bobbie	f	78660
 Bobbie	m	14570
 Bobby	f	8180
 Bobby	m	299694
+Bode	m	3348
 Bodhi	m	8375
 Bodie	m	4314
 Bonita	f	35102
@@ -1674,7 +1674,6 @@ Imran	m	3415
 Ina	f	15116
 India	f	21903
 Inez	f	22303
-Infant	m	4133
 Ingrid	f	25835
 Ira	m	33974
 Irene	f	150081
@@ -2615,6 +2614,7 @@ Lynn	f	180561
 Lynn	m	44787
 Lynne	f	61433
 Lynnette	f	15029
+Lynwood	m	3349
 Lyric	f	17861
 Lyric	m	3810
 Mabel	f	24359
@@ -3274,6 +3274,7 @@ Ricky	m	221617
 Rico	m	9313
 Ridge	m	5086
 Rigoberto	m	11018
+Rihanna	f	5970
 Rikki	f	6846
 Riley	f	117947
 Riley	m	92931
@@ -3804,6 +3805,7 @@ Tylor	m	7980
 Tyquan	m	3547
 Tyra	f	14416
 Tyree	m	13959
+Tyrel	m	3354
 Tyrell	m	15575
 Tyrese	m	8673
 Tyron	m	7456
@@ -3814,8 +3816,6 @@ Tyson	m	44397
 Ulises	m	10786
 Ulysses	m	13923
 Unique	f	7103
-Unknown	f	9225
-Unknown	m	8912
 Uriah	m	9786
 Uriel	m	17743
 Urijah	m	3365

table.go

@@ -243,7 +243,7 @@ func (t *table) proveNamesInsideParent() error {
 	for r := 0; r < t.rows(); r++ {
 		k := t.cell(r, t.parent) + "\t" + t.cell(r, t.name)
 		if first, dup := inside[k]; dup {
-			return fmt.Errorf("%s line %d: name %q repeats line %d inside %s %q; a name selects one row inside its parent", t.file, r+2, t.cell(r, t.name), first+2, t.columns[t.parent], t.cell(r, t.parent))
+			return fmt.Errorf("%s line %d: name %q repeats line %d inside %s %q; a name selects one row inside its parent; drop one, or add a key column", t.file, r+2, t.cell(r, t.name), first+2, t.columns[t.parent], t.cell(r, t.parent))
 		}
 		inside[k] = r
 	}