Alex-GilbertandClaude Opus 4.6 d2b149f1ba feat: per-line font detection with prior bias ...

Instead of detecting one font per page, detect the font for each line
independently. The previous line's font is used as a prior — if it
scores within 15% of the best candidate, it's kept for consistency.
This handles documents with mixed fonts (bold headers, regular body,
different sections).

Also: redirect / to the static UI, fix data endpoint to return
per-line font info.

Co-Authored-By: Claude Opus 4.6 <[email protected]>

2026-02-15 19:24:48 -06:00

..

__init__.py

feat: project scaffolding with dependencies

2026-02-15 18:36:10 -06:00

conftest.py

feat: project scaffolding with dependencies

2026-02-15 18:36:10 -06:00

test_alignment.py

feat: overlay calibration with word-by-word rendering

2026-02-15 19:04:40 -06:00

test_app.py

feat: per-line font detection with prior bias

2026-02-15 19:24:48 -06:00

test_e2e.py

test: end-to-end smoke test with visual output

2026-02-15 19:06:24 -06:00

test_font_detect.py

feat: font detection by comparing rendered vs OCR'd word widths

2026-02-15 18:46:07 -06:00

test_ocr.py

feat: OCR module with character-level bounding boxes

2026-02-15 18:41:05 -06:00

test_overlay.py

feat: green text overlay renderer

2026-02-15 18:49:39 -06:00

test_rasterize.py

feat: PDF rasterization module

2026-02-15 18:38:50 -06:00