| name | rime-spelling-algebra |
| description | Router and checklist for Rime spelling algebra and formatting rules. Use for xlit, xform, erase, derive, fuzz, abbrev, derive fourth-argument variants, Boost.Regex replacement syntax, speller/algebra, translator/preedit_format, translator/comment_format, and chord_composer algebra/output/prompt formats. Load references/full-reference.md for operator details and examples. |
| metadata | {"author":"RimeInn","version":"0.2.0"} |
Rime Spelling Algebra
How To Use This Skill
Use this skill when a Rime task involves spelling transformations or display formatting. Load full-reference for exact operator semantics, Boost.Regex replacement sequences, and examples.
First decide whether the rule changes input behavior or display only:
| Config Node | Effect |
|---|
speller/algebra | changes accepted spellings and compiled prism behavior |
translator/preedit_format | changes composition display only |
translator/comment_format | changes candidate comments only |
chord_composer/algebra | maps chord keys to spellings |
chord_composer/output_format | formats chord output |
chord_composer/prompt_format | formats chord prompt text |
Operator Choice
| Need | Operator |
|---|
| character-by-character transliteration | xlit |
| regex rewrite and discard original spelling | xform |
| remove matching spellings | erase |
| add aliases while keeping originals | derive |
| add fuzzy spellings for multi-syllable contexts | fuzz |
| add abbreviation spellings | abbrev |
Rules run in order. In speller/algebra, changes require redeployment.
Syntax Reminder
The first non-letter after the operator is the separator and must be used consistently in that rule:
- xform/^([nl])ue$/$1ve/
- xform|^([nl])ue$|$1ve|
- "xform ^([nl])ue$ $1ve "
Replacement strings use Boost.Regex syntax, including $1, $2, \u, \l, \U, \L, and \E. Load full-reference before relying on less common replacement behavior.
Debug Checklist
- Confirm whether the rule belongs in
speller/algebra, preedit_format, or comment_format.
- Check rule order; earlier rewrites affect later rules.
- Use
derive when both old and new spellings should work.
- Use
xform when the old spelling must stop working.
- Prefer
fuzz over broad derive when a fuzzy spelling should not make standalone single-character input too noisy.
