[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: svn commit: r11296 - in trunk/subversion: libsvn_client tests/clients/cmdline

From: John Szakmeister <john_at_szakmeister.net>
Date: 2004-10-14 23:59:06 CEST

On Thursday 14 October 2004 14:35, kfogel@collab.net wrote:
> jszakmeister@tigris.org writes:
> > --- trunk/subversion/libsvn_client/export.c (original)
> > +++ trunk/subversion/libsvn_client/export.c Fri Oct 8 17:54:12 2004
> > @@ -124,6 +124,12 @@
> > return SVN_NO_ERROR;
> > }
> >
> > + /* If the revision is WORKING, then don't export directories that
> > + haven't been committed yet. */
> > + if (revision->kind != svn_opt_revision_working &&
> > + entry->schedule == svn_wc_schedule_add)
> > + return SVN_NO_ERROR;
> > +
>
> Does this comment really match the logic in the conditional?

No it doesn't... that should be "If the revision *isn't* WORKING". Thanks
for pointing that out.

> > /* Try to make the new directory. If this fails because the
> > directory already exists, check our FORCE flag to see if we
> > care. */
> > @@ -200,8 +206,10 @@
> > svn_error_clear (err);
> > }
> >
> > - /* don't copy it if it isn't versioned. */
> > - if (! entry)
> > + /* Don't copy it if it isn't versioned, or if the entry
> > hasn't been + committed unless revision is WORKING. */
> > + if (! entry || (revision->kind != svn_opt_revision_working
> > && + entry->schedule ==
> > svn_wc_schedule_add)) continue;
> >
> > if (revision->kind != svn_opt_revision_working)
>
> The part of the comment after the comma was hard to parse, mentally,
> so I can't tell if it matches the logic in the second part of the
> conditional. But the logic in that second half exactly the same as in
> the earlier hunk, so I would expect the comments to roughly match
> up...
>
> Help?

Right again... It should be nearly the same comment as above, but with
files instead of directories. I apologize for the confusion. I'll
update that tonight.

What's happening is that this function is a helper in crawling the working
copy. However, it didn't discriminate on things that had been 'added' to
the working copy when trying to export rBASE. What ultimately happened
is that the user would see an error saying that a text-base file didn't
exist. The goal of the code was that unless we're trying to export
rWORKING, then skip over anything that's been added but not committed.
Does that make more sense?

-John

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Oct 14 23:59:30 2004

This is an archived mail posted to the Subversion Dev mailing list.