feat: smart (semantic) diff for subtitle regression outputs - #1142

Open

gaurav02081 wants to merge 10 commits into

CCExtractor:masterfrom

gaurav02081:feat/smart-diff

Commits on Aug 18, 2026

feat(smartdiff): semantic subtitle diff core (SRT parser + classifier)
gaurav-dev02
committed
72fdafb
feat(smartdiff): add WebVTT support and a format dispatcher
gaurav-dev02
committed
5d502d0
feat(smartdiff): CCExtractor-grounded normalization + cosmetic-diff kinds
gaurav-dev02
committed
14211c3
feat(smartdiff): timing drift, split/merged cues, real golden fixtures
gaurav-dev02
committed
f6a0753
feat(smartdiff): detect encoding-only differences (non-ASCII/accents)
gaurav-dev02
committed
b5d570f
test(smartdiff): real DVB Spanish golden fixture + strict & robustness tests
gaurav-dev02
committed
2ba7abd
feat(smartdiff): expose smart diff via endpoint + a "Smart" option in the UI
gaurav-dev02
committed
d19bedc
feat(smartdiff): per-cue change detail (which cues, expected/actual, offset)
gaurav-dev02
committed
6a05d92
style(smartdiff): stack Fail / Smart vertically with a divider
gaurav-dev02
committed
5ec68fd
fix(smartdiff): address code-review findings
gaurav-dev02
committed
fe76db1