strings merged with '+' are being parsed by gettext wrongly
Example (src/chat_control.py:1298):
self._tune_image.set_tooltip_markup(
_('\<b>"%(title)s"\</b> by ' +
'\<i>%(artist)s\</i>\n' +
'from \<i>%(source)s\</i>') %
{'title': title, 'artist': artist,
'source': source})
When I run intltool-update -p, I get only _<b>"%(title)s"</b> by _ the latter part of the message is just lost. The plus sign in the end of the string makes gettext think it is the end and stops extracting this string. However, if we delete it (with any backslashes after it), gettext behaves appropriately (and of course, the module also interprets normally).
Well, it is PITA for translators to figure out why those messages translated in po file are still in English in the interface. Please make sure every such string is fixed before release! This simple command will hopefully find such strings:
egrep -nr "_\('[^']*'( )*\+|_\(\"[^\"]*\"( )*\+" *