https://bugzilla.redhat.com/show_bug.cgi?id=984537
Bug ID: 984537 Summary: regression in gtk_text_ functions Product: Fedora Version: 19 Component: pango Keywords: Regression Severity: medium Priority: medium Assignee: tagoh@redhat.com Reporter: ksrot@redhat.com QA Contact: extras-qa@fedoraproject.org CC: fonts-bugs@lists.fedoraproject.org, i18n-bugs@lists.fedoraproject.org, tagoh@redhat.com
Description of problem:
There are several failures in LSB test suite that were not present in Fedora 18 and are related to text_iter_* functions, for example:
---------------------------------------
/tests/functions/GtkTextIter/GtkTextIter 234 failed Message from the test: test case: /opt/lsb/test/desktop/gtkvts/tet_tmp_dir/02926aa/tests/functions/GtkTextIter/GtkTextIter, TP number: 234 GtkTextIter.c, line 14345: gtk_text_starts_word () returned not 1 incorrectly, line = 'word1word2 word3', pos = 4
or
Message from the test: test case: /opt/lsb/test/desktop/gtkvts/tet_tmp_dir/02926aa/tests/functions/GtkTextIter/GtkTextIter, TP number: 340 GtkTextIter.c, line 18892: gtk_text_ends_word () returned not 1 incorrectly, line = 'word1word2 word3', pos = 4
Message from the test: test case: /opt/lsb/test/desktop/gtkvts/tet_tmp_dir/02926aa/tests/functions/GtkTextIter/GtkTextIter, TP number: 951 GtkTextIter.c, line 53871: 'gtk_text_iter_backward_word_start' moved iterator incorrectly 'Word1 Word2' iter = (0, 0), should be (0, 4)
Message from the test: test case: /opt/lsb/test/desktop/gtkvts/tet_tmp_dir/02926aa/tests/functions/GtkTextIter/GtkTextIter, TP number: 974 GtkTextIter.c, line 55239: 'gtk_text_iter_forward_word_end' moved iterator incorrectly 'Word 1. Word2' iter = (1, 5), should be (1, 4)
Message from the test: test case: /opt/lsb/test/desktop/gtkvts/tet_tmp_dir/02926aa/tests/functions/GtkTextIter/GtkTextIter, TP number: 991 GtkTextIter.c, line 56265: 'gtk_text_iter_forward_word_ends' moved iterator incorrectly 'Word 1, Word2?! Word3' count = 5, iter = (0, 21), should be (0, 20)
--------------------------------------
Actually this is a regression between pango-1.34.0-1 and pango-1.34.1-1.
https://bugzilla.redhat.com/show_bug.cgi?id=984537
--- Comment #1 from Karel Srot ksrot@redhat.com --- Created attachment 773727 --> https://bugzilla.redhat.com/attachment.cgi?id=773727&action=edit standalone reproducer for gtk_text_iter_starts_word() function
as an example I am adding a standalone reproducer for the gtk_text_iter_starts_word() function.
# ./a.out | grep FAIL FAIL: gtk_text_starts_word () returned not 1 incorrectly, line = 'word1word2 word3', pos = 4 FAIL: gtk_text_starts_word () returned not 1 incorrectly, line = 'word1word2 word3', pos = 5 FAIL: gtk_text_starts_word () returned not 1 incorrectly, line = 'word1_word2 word3', pos = 4
https://bugzilla.redhat.com/show_bug.cgi?id=984537
Akira TAGOH tagoh@redhat.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |CLOSED Resolution|--- |NOTABUG Last Closed| |2013-07-15 23:33:20
--- Comment #2 from Akira TAGOH tagoh@redhat.com --- This behavior was introduced by this change: https://git.gnome.org/browse/pango/commit/?id=1aeb5c840e25a7d8538f701659d77d...
According to the commit log, this change was needed to meet Unicode 6.2 specs and the previous behavior was wrong. so this is not a bug. please fix the test case instead.
i18n-bugs@lists.fedoraproject.org