nass 0.3.0
nass: ^0.3.0 copied to clipboard
Arabic text normalization and diacritic-insensitive search for Quran, hadith and heritage texts, with match offsets in the original vocalized text.
0.3.0 #
- Experimental
splitIsnad: new rules; 88.3% on a new held-out set of 300 hadiths (dev 98.0% on 500). - No change to normalization: keys from 0.2 stay valid.
- Implements SPEC v0.3 and passes all 1,554 shared test vectors. Full notes in the repository CHANGELOG.md.
0.2.0 #
search: U+06E7 (small high yeh) becomes «ي» instead of being deleted («ٱلنَّبِيِّـۧنَ» → «النبيين»).cleankeeps ZWNJ (U+200C);searchremoves it next to Arabic letters.- Experimental
splitIsnad: new rules; 88.0% on a new held-out set of 200 hadiths (dev 98.3% on 300). - Implements SPEC v0.2 and passes all 1,254 shared test vectors. Full notes in the repository CHANGELOG.md.
0.1.0 #
- First release:
normalize,normalizeWithMap,findAll,contains,highlight,nameKey. - Experimental
splitIsnad(91% on a held-out set of 100 hadiths). - Implements SPEC v0.1 and passes all shared test vectors.