Commit Graph

5 Commits

Author SHA1 Message Date
prosolis
28ec11ba4e Consolidate word filtering into shared isJunkWord helper
Replace copy-pasted hasLetter/containsDigit/containsSpace/containsNonLatin
checks across 9 loaders with a single isJunkWord() call. SCOWL and CEDICT
keep their own filtering for valid reasons (SCOWL rejects hyphens/apostrophes,
CEDICT handles non-Latin Chinese characters).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-03 20:01:02 -07:00
prosolis
aff583dc28 Prune ghost words from affix expansion and stop frequency loader creating words
- Add PruneOrphanWords post-import step that deletes words with no
  definitions, synonyms, translations, pronunciations, etymology,
  synsets, or frequency data
- Remove INSERT path from CETEMPublico loader — frequency data should
  annotate existing words, not create new ones (was inflating pt-PT
  from 136K real words to 718K)
- Add hasLetter filter to hunspell, affix, and dicionario loaders to
  block punctuation entries like ".", ",", "(" from entering the DB
- Move hasLetter helper to scowl.go alongside other shared filters

Impact: en 655K→129K, pt-PT 1.5M→136K, fr 80K→56K (on reimport)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-03 19:55:37 -07:00
prosolis
207086da08 Add Phase 2+3 features: antonyms, backing, pronunciation, etymology, difficulty, affix expansion
New endpoints: /antonyms, /backing, /pronunciation, /etymology with difficulty
scoring on /random. Cross-language synset backing links French/Portuguese words
to English equivalents via WordNet 3.0 synset IDs (matching WOLF and OMW offsets).

New loaders: Hunspell affix expansion (fr, pt-PT), English programmatic inflector,
CMU Pronouncing Dictionary, SUBTLEX-US frequency, CETEMPúblico frequency,
Open Multilingual Wordnet (Portuguese), and SQL-based difficulty scoring.

Schema v2 adds tables: antonyms, synsets, word_synsets, pronunciations, etymology
with migration support for existing databases.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-02 01:01:06 -07:00
prosolis
eb471b6f6d Add word frequency endpoint and SCOWL frequency data
SCOWL loader now sets frequency based on tier (10=1000 most common,
70=50 rare). New GET /frequency?word=X&lang=Y endpoint returns frequency
score for a word. Enables downstream consumers to identify rare/
sophisticated vocabulary from dictionary data rather than LLM guessing.

Also fix .gitignore matching cmd/server/ directory instead of just the
server binary.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-30 21:27:45 -07:00
prosolis
eeea8d4d34 Initial commit: DreamDict HTTP dictionary service
Go-based dictionary backend providing word validation, definitions, synonyms,
and cross-language translation for English, French, and Portuguese. Designed
as a self-hosted replacement for Wordnik, backing GogoBee.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-28 16:57:40 -07:00