Repository navigation
v4.3.1
Bug Fixes
- Masks landing on the wrong characters (#56): the regex and phonetic drivers matched against normalized text but masked the original at the same positions. Whenever normalizing changed the length, every later mask was shifted. This affected:
- German
ä/ö/ü/ßandsch - Spanish
ll - French
œ/æ - Repeated whitespace in any language
- German
| Input | Before | After |
|---|---|---|
a fuck this shit |
a ****k thi****it |
a **** this **** |
Du blöde Scheiße, du Arschloch |
Du ******S******* du A******** |
Du ***** *******, du ********* |
llama mierda |
llama******a |
llama ****** |
Matched word text, positions and clean() now all refer to the right characters. No API changes.