You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This crash happens very often in the fuzzing test. It seems to be caused by Tesseract using the wrong unicharset, so tessdata_best, tessdata_fast and the initial revision of tessdata are not affected. It can occur with traineddata which includes different unicharsets for legacy and LSTM recognizers. The fuzzer uses tessdata/eng.traineddata where eng.lstm-unicharset has 112 entries while eng.unicharset has 113 entries. It tries to get entry 112 which only exists in the legacy unicharset.
Obviously PR #1954 did not fix all code locations.
…-ocr#2298)
Credit to OSS-Fuzz: This fixes an issue which was reported by OSS-Fuzz:
Add also some assertions to catch similar bugs.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
stweil
added a commit
to stweil/tesseract
that referenced
this issue
Mar 10, 2019
OSS Fuzz reports a security issue.
The text was updated successfully, but these errors were encountered: