Correctly handle unicode linebreaks (closes bgo#602933, bgo#627940)
As it stands, Meld relies on Python's universal newline support to translate various traditional newline sequences into '\n', which we then split on. This approach fails for unicode linebreaks, and when Python's universal newline support isn't available. This patch moves to using a combination of Python's unicode splitlines() and asking our gtk.TextBuffer where it thinks linebreaks are in order to handle unicode linebreaks.
Showing with 39 additions and 23 deletions