Add the fast-enough goal and record the allocation-gate decision
Tests / vet + fmt + tests (pull_request) Successful in 1m3s

This commit is contained in:
2026-09-03 21:46:22 +02:00
parent 296058fb80
commit 0d2a080b85
+9
View File
@@ -360,6 +360,9 @@ tokens add cost in proportion to the output.
7. **Zero dependencies** — standard library only. 7. **Zero dependencies** — standard library only.
8. **Docs index the grammar** — every syntax feature is a heading; every example 8. **Docs index the grammar** — every syntax feature is a heading; every example
runs under test and shows its output; a rule is stated once. 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`
parses and validates the whole set once upfront, so generating fixtures stays
noise against a test's own runtime.
## Decisions ## Decisions
@@ -451,6 +454,12 @@ tokens add cost in proportion to the output.
- **Samples say what they emit, transforms what they do.** `{upper(2)}` is two - **Samples say what they emit, transforms what they do.** `{upper(2)}` is two
letters, `{uppercase(x)}` is `x` upper-cased; one name for both would turn on letters, `{uppercase(x)}` is `x` upper-cased; one name for both would turn on
whether the argument looks like a number. whether the argument looks like a number.
- **The performance gate asserts allocations, not wall-clock time.** `AllocsPerRun`
is deterministic across machines, so a ±10% ceiling does not flake under CI load,
while time varies with the machine and its neighbours. A rendering slowdown
almost always costs an allocation too (a lost pre-size, a per-item map, an extra
copy). The benchmark suite (see Development) reports time for a human, not as a
pass/fail gate.
## Development ## Development