From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.2)
Description of problem:
If writer offers to complete a word that contains a non-ASCII
character and you keep typing it, when you enter the non-ASCII
character, it will complete the word all the way to the end, without
the non-ASCII character. This makes it extremely difficult to use for
languages that use non-ASCII characters.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1.Start a new text
2.Disable auto spell check (otherwise you won't get offers to
auto-complete words, and the bug won't occur)
3.Enter the following words: camÃ©lia caÃda Ã³rgÃ£o dominÃ³ (if you have
no clue on how to enter the accents, choose Preferences->Keyboard,
Layouts, U.S. English w/ deadkeys, then type 'e for Ã©)
4.In a separate line, attempt to enter the same words. It will
attempt to auto-complete them, but disregard that and just keep typing
Actual Results: The first line will be correct.
The second line will read: camlialia cadada Ã³rgoo dominÃ³
Expected Results: The second line should read the same as the first
Note that, when you enter a non-ASCII character during
auto-completion, it will remove the non-ASCII character and add the
rest of the word, such that, when you keep typing the word, whatever
follows the first non-ASCII character will be duplicated. The problem
does not affect non-ASCII characters that are typed before
auto-completion figures out the word you're tying, nor does it affect
non-ASCII characters at the end of a word.
LANG=en_US.UTF-8, in case it matters.