Arabic: default the unvoweled lemma to Form I, not an arbitrary derived form - #155
Merged
Conversation
…ed form
The Arabic citation is the unvoweled consonantal skeleton, which is a homograph
across the derived forms — كتب is both Form I كَتَبَ ("write") and Form II كَتَّبَ
("make write"). 775 skeletons carry more than one form class (762 include Form
I), and the kaikki adapter's arbitrary last-wins dedup was resolving many common
verbs to the derived reading: a user typing كتب, ذهب, درس, فتح got kattaba,
dhahhaba, darrasa, fattaha instead of the everyday verb.
The adapter now keeps the LOWEST form class per cell (Form I over II over III…),
using Wiktionary's own form-i/form-ii tags, so the default conjugation is the
basic verb. parts.tsv and the mined overrides are regenerated against the
Form-I-preferred gold; the golden gate still passes 100% (47456/47456 forms,
632/632 lemmas — the gold shifts slightly as some homograph agreements move to
Form I). The rarer derived reading is no longer reachable from the bare
skeleton; voweled input to disambiguate is a possible follow-up.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
UX fix for the Arabic homograph problem. The citation is the unvoweled skeleton, which is ambiguous across the derived forms — كتب is both Form I كَتَبَ ("write") and Form II كَتَّبَ ("make write"). 775 skeletons carry more than one form class (762 include Form I), and the kaikki adapter's arbitrary last-wins dedup resolved many everyday verbs to the derived reading:
The adapter now keeps the lowest form class per cell (Form I over II over III…) using Wiktionary's own
form-i/form-iitags.parts.tsv+ the mined overrides are regenerated against the Form-I-preferred gold.Golden gate still passes 100% (47456/47456 forms, 632/632 lemmas — the gold shifts slightly as some homograph agreements move to Form I). 24 ara tests, clippy, fmt all green.
Trade-off: the rarer derived reading is no longer reachable from the bare skeleton. Voweled input to disambiguate is a possible follow-up. A web PR rebuilds the vendored wasm and refreshes the Arabic sample verbs (كتب/درس/ذهب are now correct).
🤖 Generated with Claude Code