characterminator 0.1.2

find and eliminate characters outside an allowed set, per file type and per file, so text costs fewer tokens
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
# SPEC ARCHIVE

Task rows moved out of `SPEC.md` by `mth archive`. An id is never reused
(`V12`), so a citation to an archived row still resolves -- here.

This is a SINK, not a spec. The citations inside these rows point into
`SPEC.md`, so `mth check` on this file reports every one of them as dangling,
correctly and uselessly. The verb that reads it is `mth tasks`.

## §T TASKS

T10|x|transliteration map + `fix` & `fix --check`; property tests: idempotency, untouched bytes|V4,V5,V6
T26|x|builtin map: typography defaults|V26,`src/charset:V22`
T29|x|family lines: parse, tree validation, fallback path|V27
T30|x|class lines: parse, family-labelled members, resolution along fallback path|V28,V27
T33|x|2A: vendor emoji 18.0 sequence data; generator → `emoji-seq.ctrm-map`; curated ZWJ table; fixtures ∀ target kind & a non-RGI ZWJ that stays reported; lint exemption per `src/lint/hazard:V63`|V62,V31,`src/charset:V22`,`src/lint/hazard:V34`
T53|x|`words` map: `use` & `word` lines, word-boundary space, measured entries, property tests|V51,V5,V6,`.:R11`
T59|x|builtin map: VS15, VS16 & skin-tone deletes + test|V60,V31