[PATCH] merge_tracking merge_tests.py testcase failures fix for delete_file_and_dir

[Kamesh Jayachandran <ka...@collab.net>]

Hi All,
Find the attached patch.

With regards
Kamesh Jayachandran
[[[
Patch by: Kamesh Jayachandran <ka...@collab.net>

Even though the given merge leads to 'Skip' svn:mergeinfo is recorded.
This causes subsequent force merge to fail.

* subversion/tests/cmdline/merge_tests.py
  (delete_file_and_dir):
   'svn pd svn:mergeinfo' to make subseuent forced merges to succeed.
]]]

[merge tracking] Handling notifications resulting from a merge

[Daniel Rall <dl...@collab.net>]

On Fri, 07 Jul 2006, Kamesh Jayachandran wrote:

> Garrett Rooney wrote:
> >
> >Uhh, maybe it's just me, but this seems like you're working around the
> >actual problem here.

Kamesh and I spent some time looking into this test failure today, and
this did indeed turn out to be the case.

...
> Even though the given merge leads to 'Skip' svn:mergeinfo is recorded.
> This causes subsequent force merge to fail.
> 
> * subversion/tests/cmdline/merge_tests.py
>  (delete_file_and_dir):
>   'svn revert -R B2_path' to make subsequent forced merges to succeed.
...

This turned out to be part of the broader scoped notification handling
changes necessary on the merge-tracking branch.  My working plan for
handling these issues is written up in TODO [1].  I'd love to get some
feedback on the approach:

  * Handle multiple notifications for single WC items. Possible
    solutions include:

    * Output multiple notifications, but print divider lines
      indicating the revisions range to which a set of notifications
      applies.  Introduce a new type of "skipped" notification for WC
      items which are already in conflict.

    * Collate changes as merge ranges are applied.  Detect and handle
      collisions (multiple notifications to the same WC item), giving
      preference to later notifications (?).

  * Handle skips.  Merge test 3 fails because a merge of a revision
    range which contains a delete will not delete locally modified
    files (at least, not without --force), but is still recording
    merge info.

    * If all changes in a merge are skipped, no merge info should be
      recorded for the target.

    * If only some changes are skipped, merge info should be recorded
      for the target, and recorded as empty (or with no modifications,
      if there is pre-existing merge info) for the skipped items.

  * Handle conflicts.

    * If a conflict is encountered, invoke a conflict resolution
      callback to give a Subversion client a chance to intervene.  If
      resolution is successful, convert the notification from a 'C' to
      something else (e.g. 'M').  (Phase 2?)

    * Otherwise, stop applying merge ranges as soon as a second
      conflict is encountered in a WC item (as it might generate
      overlapping conflict makers, or apply a merge inside a conflict
      marker!), being sure to record partial application of merge
      ranges.



[1] http://svn.collab.net/repos/svn/branches/merge-tracking/TODO

Re: [PATCH] merge_tracking merge_tests.py testcase failures fix for delete_file_and_dir

[Kamesh Jayachandran <ka...@collab.net>]

Garrett Rooney wrote:
>
> Uhh, maybe it's just me, but this seems like you're working around the
> actual problem here.  Why should we have to delete the mergeinfo prop?
> We don't expect users to do that, do we?
>
> -garrett
>
Yes more proper one would be to 'svn revert'. (Was wondering how to 
revert changes to one property alone!. Anyway I can revert the workarea 
itself.
Thanks for finding that out.

Please find the attached patch.

With regards
Kamesh Jayachandran

[[[
Patch by: Kamesh Jayachandran <ka...@collab.net>

Even though the given merge leads to 'Skip' svn:mergeinfo is recorded.
This causes subsequent force merge to fail.

* subversion/tests/cmdline/merge_tests.py
  (delete_file_and_dir):
   'svn revert -R B2_path' to make subsequent forced merges to succeed.
]]]

Re: [PATCH] merge_tracking merge_tests.py testcase failures fix for delete_file_and_dir

[Garrett Rooney <ro...@electricjellyfish.net>]

On 7/7/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> Hi All,
> Find the attached patch.
>
> With regards
> Kamesh Jayachandran
> [[[
> Patch by: Kamesh Jayachandran <ka...@collab.net>
>
> Even though the given merge leads to 'Skip' svn:mergeinfo is recorded.
> This causes subsequent force merge to fail.
>
> * subversion/tests/cmdline/merge_tests.py
>   (delete_file_and_dir):
>    'svn pd svn:mergeinfo' to make subseuent forced merges to succeed.
> ]]]
>
>
>
> Index: subversion/tests/cmdline/merge_tests.py
> ===================================================================
> --- subversion/tests/cmdline/merge_tests.py     (revision 20465)
> +++ subversion/tests/cmdline/merge_tests.py     (working copy)
> @@ -626,6 +626,11 @@
>                                         expected_status,
>                                         expected_skip)
>
> +  #subversion/libsvn_client/diff.c:do_merge has no way of knowing
> +  #if at all any changes are caused by this merge.
> +  #delete the svn:mergeinfo so that subsequent force merge would work.
> +  svntest.actions.run_and_verify_svn(B2_path, None, [],
> +                                     'pd', SVN_PROP_MERGE_INFO, B2_path)
>    expected_output = wc.State(B2_path, {
>      'E'       : Item(status='D '),
>      'E/alpha' : Item(status='D '),

Uhh, maybe it's just me, but this seems like you're working around the
actual problem here.  Why should we have to delete the mergeinfo prop?
 We don't expect users to do that, do we?

-garrett

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org