As at HEAD(r1052844) i could still see that docstring in
subversion/include/svn_error.h
----------
Sent via Nokia Email
------Original message------
From: Daniel Shahaf <d...@daniel.shahaf.name>
To: "Kamesh Jayachandran" <kam...@collab.net>
Cc: <dev@subversion.apache.org>,<comm...@subversion.apache.org>
Date: Sunday, December 26, 2010 10:59:43 AM GMT+0200
Subject: Re: svn commit: r1052817 -
/subversion/trunk/subversion/tests/libsvn_subr/error-test.c
You're quoting an old version of the docstring.
Kamesh Jayachandran wrote on Sun, Dec 26, 2010 at 12:10:46 +0530:
> >+
> >+ err2 = svn_error_purge_tracing(err);
> >+ for (child = err2; child; child = child->child)
> >+ if (svn_error__is_tracing_link(child))
> >+ {
> >+ return svn_error_create(SVN_ERR_TEST_FAILED, err,
> >+ "Tracing link found after purging the "
> >+ "following chain:");
> >+ }
> >+
> >+ svn_error_clear(err);
>
> The doc of svn_error_purge_tracing(err) says err should *not* be cleared.
>
> <snip>
> /**
> * Purge from @a err and its child chain any links associated with
> * error tracing placeholders, and return the new top-level error
> * chain item. @a err should be considered unusable after passing
> * through this function, but should *not* be cleared (as the returned
> * error shares memory with @a err). @a err can be #SVN_NO_ERROR.
> *
> * @since New in 1.7.
> */
> svn_error_t *svn_error_purge_tracing(svn_error_t *err);
> </snip>
>
> Are you sure?
>
> With regards
> Kamesh Jayachandran
>
>
