> On Fri, 20 May 2005, Madan US wrote:
>
>> > this is not a complete review, just clafirying some of philip's
>> > points.
>> >
>> > On Tue, 10 May 2005, Philip Martin wrote:
>> >
>> >> "Madan US" <madan@collab.net> writes:
>> >>
>> [...]
>> >>
>> >> Hmm, a quick experiment shows that the existing hook output appears
>> >> to get escaped over ra_dav, but not encoded. The hook needs to
>> >> output UTF-8 if it is to be displayed by the client, otherwise I
>> >> just see "svn: General svn error from server".
>> >>
>> > It definitely needs to be XML-escaped. Whether it should be recoded
>> > from the locale encoding is another topic - and should be a separate
>> > change IMO. Oh, and it needs base4 encoding as well, or filtering so
>> > that invalid XML chars don't come through.
>>
>> Is the pre-commit hook stderr xml-escaped???? I wonder why no one
>> pointed that out.... :?
>>
> that's marshallade as a normal error and errors are XML escaped.
^^^^^^^^^^^^^^^^^^^^^^
okay... I didnt know that.
> (Have
> you tested and found out that it isn't the case?
I have not tested it. I was saying that because the way the stderr is
captured is the same for both pre-commit and post-commit.
> In that case, feel
> free to correct me!) A test for this is a hook script that outputs some
> characters that need to be entities, such as < and &. Might be good to
> include that in the test case for this feature.
Great Idea.
I should also include command like import, cp, mv that do an implicit commit
in a new test case.
I will do both of these.
Thanks, Peter.
Regards,
Madan.
>
> //Peter
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri May 20 20:29:31 2005