a small service to automate the searching and downloading of eBooks from irchighway.net
- Python 62.1%
- JavaScript 16.9%
- CSS 11.1%
- Shell 6.2%
- HTML 3.4%
- Other 0.3%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
Leigh
7833a2aec8
Each bot formats result lines differently; auditing 635 cached rows showed hash/ID prefixes (TrainFiles/artemis "hex | ", Firebound %HEX%, Ashurbanipal base64 ids), trailing genre lists, archive extensions, and (retail)/(epub) tags surviving into clean_title — and format='?' for 34% of rows because rar-serving bots hide the real format in a parens tag. - search_parser: ordered cleanup pipeline (structure-keyed, not bot-keyed); format falls back ext -> parens tag -> archive type; htm normalized to HTML - results_store: PRAGMA user_version migration re-derives clean fields from stored raw titles on init, FTS reindexes via the existing trigger - tests: 11 real per-bot fixture lines, prefix/genre guards, migration test Live cache verified: unknown formats 214 -> 0, genre words no longer produce false FTS cache hits, zero junk rows in a full post-migration sweep. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> |
||
| backend | ||
| frontend | ||
| scripts | ||
| tests | ||
| .gitignore | ||
| agents.md | ||
| CLAUDE.md | ||
| CODE_REVIEW.md | ||
| docker-compose.yml | ||
| Dockerfile | ||
| PLAN.md | ||
| requirements.txt | ||