Re: [PATCH] callers to svn_time_to_human_cstring should expect UTF8 (issue #1534)

Josh Pieper <jpieper@andrew.cmu.edu> writes:

> Here is a fix to part of issue #1534. What I haven't handled is the
> case where the date is substituted into a working copy file via the
> svn:keywords mechanism. As far as I can tell, currently other
> substitutions are not translated from UTF8 before getting stuck in
> working copy files. Also, the encoding of those WC files is not
> tracked, so this problem may be too difficult to handle right away.

Hm... I think that for right now we should just document that inserted
keywords are in 7-bit ASCII (or the ASCII subset of UTF8 if it makes
everyone feel better). Not sure I want to go down the road (just yet)
of recognizing the character set and encoding of each versioned file.
My gut instinct is that behind those gates lies pain and suffering,
more so than we want to bear pre-1.0.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Oct 6 15:48:47 2003