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

RE: svn commit: r1337844 - /subversion/trunk/subversion/libsvn_wc/conflicts.c

From: Bert Huijben <bert_at_qqmail.nl>
Date: Sun, 13 May 2012 13:50:01 +0200

> -----Original Message-----
> From: stsp_at_apache.org [mailto:stsp_at_apache.org]
> Sent: zondag 13 mei 2012 13:07
> To: commits_at_subversion.apache.org
> Subject: svn commit: r1337844 -
> /subversion/trunk/subversion/libsvn_wc/conflicts.c
>
> Author: stsp
> Date: Sun May 13 11:06:29 2012
> New Revision: 1337844
>
> URL: http://svn.apache.org/viewvc?rev=1337844&view=rev
> Log:
> * subversion/libsvn_wc/conflicts.c
> (resolve_conflict_on_node): Restore pre-r1337579 behaviour of setting the
> did_resolve output parameter to FALSE if a text conflict was implicitly
> resolved by removing the conflict marker files. Apparently this is a
> "feature" we want to keep... *grumble*

+1 on the sentiment.
>
> Modified:
> subversion/trunk/subversion/libsvn_wc/conflicts.c
>
> Modified: subversion/trunk/subversion/libsvn_wc/conflicts.c
> URL:
> http://svn.apache.org/viewvc/subversion/trunk/subversion/libsvn_wc/conflicts.
> c?rev=1337844&r1=1337843&r2=1337844&view=diff
> =================================================================
> =============
> --- subversion/trunk/subversion/libsvn_wc/conflicts.c (original)
> +++ subversion/trunk/subversion/libsvn_wc/conflicts.c Sun May 13 11:06:29
> 2012
> @@ -304,7 +304,30 @@ resolve_conflict_on_node(svn_boolean_t *
> SVN_ERR(svn_wc__wq_run(db, local_abspath,
> cancel_func_t, cancel_baton,
> pool));
> - *did_resolve = TRUE;
> +
> + /* Text conflicts may be marked resolved by removing the conflict
> + * marker files. If they're already deleted, don't provide feedback. */
> + if (resolve_text && !resolve_props && !resolve_tree)
> + {
> + svn_node_kind_t node_kind = svn_node_unknown;
> +
> + if (conflict_old)
> + SVN_ERR(svn_io_check_path(conflict_old, &node_kind, pool));
> + *did_resolve = (node_kind == svn_node_file);
> +
> + if (!*did_resolve && conflict_new)
> + SVN_ERR(svn_io_check_path(conflict_new, &node_kind, pool));
> + *did_resolve = (node_kind == svn_node_file);
> +
> + if (!*did_resolve && conflict_working)
> + SVN_ERR(svn_io_check_path(conflict_working, &node_kind, pool));
> + *did_resolve = (node_kind == svn_node_file);

Did resolve should be true when either one of those is removed by removing the working queue.

After running the working queue none of those should exist, so the tests are failing.

So you should test if any one of those exists before running the wait queue. (One file found and a successful run of the working queue is enough for a TRUE at return)

Currently the buildbots fail in the basic (11) and resolve (1,3) tests.

        Bert

> + }
> + else
> + {
> + /* Always provide feedback for property and tree conflicts. */
> + *did_resolve = TRUE;
> + }
> }
>
> return SVN_NO_ERROR;
>
Received on 2012-05-13 13:50:49 CEST

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

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.