Closed GoogleCodeExporter closed 9 years ago
Can you create patches as ettl.martin78 is doing see e.g.[1]?
Also make sure you use the latest svn revision.
[1] https://code.google.com/p/tesseract-ocr/issues/detail?id=1102
Original comment by zde...@gmail.com
on 24 May 2014 at 1:56
@zde...
Please take a look at the report from CppCat analyzer. It gives you the
necessary information about potential issues in tesseract-ocr.
@karpov2...
Many thanks for the report.
Original comment by ettl.mar...@gmail.com
on 9 Jun 2014 at 8:13
There are a few useful observations in the report, but a *lot* of false
positives.
Eg the first shift-of-a-signed is good (but currently harmless), but the rest
reported in intmatcher.cpp, are all shifts of an unsigned. Just plain wrong.
Some of the suspicious integer divisions are division of an int assigned to an
int, although most are assigned to a float.
The incorrect order of class member initialization is also a red-herring, as
the out-of-order elements are just taking a copy of the reference to the
not-yet-initialized object.
The typo in a condition section is also a false-positive.
Original comment by theraysm...@gmail.com
on 12 Aug 2014 at 1:48
CppCat has several mechanisms to suppress false positives. Read the
documentation. Hope this helps.
Original comment by karpov2...@gmail.com
on 12 Aug 2014 at 4:07
Fixed in revs 736d32747333 and 9c58701471d6.
The most significant finding is turned to a TODO in language_model.cpp, but
current functionality retained for now.
Original comment by theraysm...@gmail.com
on 12 Aug 2014 at 11:18
Original issue reported on code.google.com by
karpov2...@gmail.com
on 22 May 2014 at 1:28