feat(ebook): add protected phrase extraction library with config-driven tuning

Browse Source

Refactor protected phrase handling from a single module into a
python/ebook_search/protected_phrases package covering extraction,
storage, and runtime matching. Phrase filtering is now data-driven via
bundled TOML files: ignored_phrases, bad_starts, bad_ends, and
most_common_words.

Add phrase-tuning settings to EbookSearchConfig so candidate generation,
scoring, LLM judging, and matching are configurable rather than hardcoded:
token bounds, entity token limit, raw n-gram min count, frequency and
chapter-spread score thresholds, candidate/LLM/target caps, confidence
threshold, nesting defaults, and the phrase hit boost.

...

This commit is contained in:

Richie

2026-07-09 11:04:59 -04:00

parent 3e164831b5

commit 1eecf7181d

7 changed files with 2296 additions and 0 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

python/ebook_search/protected_phrases/lib.py

+2010

File diff suppressed because it is too large Load Diff