I removed the "lone ampersand" fixup stuff. I think that was left over from when we were using Pango markup escaping, because it could confuse things. Note that bubble.c no longer uses pango_parse_markup(), it only uses filter_text(). This fixes a lot of things "such as <>" being interpreted when it should, "x < 1" being interpreted when it shouldn't, etc. It also is important from the unit test perspective, because now the unit test actually tests -exactly- what will be filtered by notify- osd (whereas previously it only tested what would be filtered by filter_text(), and then Pango would go and screw things up after that).
This also adds new regexes for the numerics, which I had forgotten to include in the last patch. Hopefully the unit tests cover things pretty well and will make this easier for Mirco to integrate. ** Attachment added: "Patch for review" http://launchpadlibrarian.net/25115482/356659-3.diff -- Character reference handling for <, &, and > doesn't match notification-daemon https://bugs.launchpad.net/bugs/356659 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
