zenity's text-info mistakes encoding from stdin

[Serhiy]

It is reproducible with the following bash script:

mkfifo test
grep "аванс" /usr/share/dict/ukrainian > test &
cat test | zenity --text-info
grep "аванс" /usr/share/dict/ukrainian > test &
zenity --text-info --filename test
rm test

On the first time it outputs about 50 first lines with wrong encoding.

ProblemType: Bug
DistroRelease: Ubuntu 11.04
Package: zenity 2.32.1-0ubuntu1
ProcVersionSignature: Ubuntu 2.6.38-10.46-generic 2.6.38.7
Uname: Linux 2.6.38-10-generic i686
Architecture: i386
Date: Mon Jul 25 15:47:17 2011
InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386 (20091028.5)
ProcEnviron:
 LANGUAGE=uk_UA:en
 PATH=(custom, user)
 LANG=uk_UA.UTF-8
 SHELL=/bin/bash
SourcePackage: zenity
UpgradeStatus: Upgraded to natty on 2011-07-21 (3 days ago)

[Serhiy (xintx-ua)]

Here is the reproducible code for those without /usr/share/dict/ukrainian :
for i in {1..342}; do echo "а" >> test2; done; cat test2 | zenity --text-info; rm test2

And when you reduce the number it detects right encoding:
for i in {1..341}; do echo "а" >> test2; done; cat test2 | zenity --text-info; rm test2

[Serhiy (xintx-ua)]

Yes, just what I thought:

$ for i in {1..341}; do echo "а" >> test2; done; wc -c test2; rm test2
1023 test2

$ for i in {1..342}; do echo "а" >> test2; done; wc -c test2; rm test2
1026 test2

It happens when the input exceeds 1024 bytes.

[Serhiy (xintx-ua)]

Workaround is to pipe it through
head -c 1023

[Pedro Villavicencio (pedro)]

This bug is an upstream one and it would be quite helpful if somebody experiencing it could send the bug the to the people writing the software. You can learn more about how to do this at https://wiki.ubuntu.com/Bugs/Upstream/GNOME . Thanks in advance!

[Serhiy (xintx-ua)]

Thanks, I've reported it and added a watch.