Re: [photran] Refactorings not working with non-ASCII characters ?

[Jeffrey Overbey <overbey2@xxxxxxxxxxxx>]

> I am bet we did not test for non-ASCII characters, which probably
> means that they don't work.

A few versions ago, when I was building a lot of the refactoring
infrastructure, Eclipse didn't even support multibyte characters -- it
would refuse to let you save the file in the editor -- so it was a
non-issue.  Obviously they're supported now, but, you're right, we
never added any tests, so I don't actually know how they're handled.
I just found out recently that Eclipse will let you change the
encoding on a per-file basis (so you could have UTF-8, ISO-8859-1, and
MacRoman all in the same project), and I *know* we pay no attention to
that.

> An internal error occurred during: "Photran indexer".
> java.lang.ArrayIndexOutOfBoundsException: 4095"

OK, this means that the indexer's database was corrupted somehow.
(We'll be able to detect and fix that automatically in Photran 6.)
Click on Refactor > (Debugging) > Clear and Rebuild Database.

If you can reproduce the errors after rebuilding the database, please
enter a bug in Bugzilla.  If you go into
/your/workspace/.metadata/.log, there will be a stack trace for the
various exceptions that were thrown.  If you could attach (the
relevant sections of) that log file to the bug, that would help me
track it down.  We might not be able to completely fix this by Photran
6, but it should never be throwing uncaught exceptions, anyway.

Jeff