I am doing a transplant of commits where for some of them the resolution is to get rid of the changes entirely. When this is done via a filter, the commit is skipped. When this is done manually (e.g. the ones the filter cannot decide on), I get: $ hg st ? .eggs/README.txt ? .eggs/argparse-1.2.1-py2.6.egg ? NEWS~ ? utils/transplant-filter.sh $ hg transplant -v -s ../cpython.hg --filter utils/transplant-filter.sh 0f1619e539fb: --continue searching for changes sampling from both directions warning: repository is unrelated abort: commit failed I would have expected this to have skipped over the commit rather than try to commit an empty changeset.
Created attachment 1795 [details] patch for the issue I'm not on the selenic mailing list and this is a drive-by fix, so I figured this was better than subscribing just long enough to patchbomb and unsubscribe.
No subscription is necessary to submit a patch to the list. We can't review or accept patches via the BTS, hence our pleading not to submit them this way.
A patch have been submitted here: http://www.selenic.com/pipermail/mercurial-devel/2014-October/063148.html But the submitter have no interest in maturing it. So someone else will have to.
Hi, I'm on annual leave at the moment. If you have urgent queries about Helion please contact Chris Jones (cmsj@hp.com) or Monty Taylor (monty.taylor@hp.com). For OpenStack and upstream opensource things please contact Monty Taylor.
Fixed by http://selenic.com/repo/hg/rev/49caef455912 Pierre-Yves David <pierre-yves.david@fb.com> transplant: properly skip empty changeset (issue4423) If resolving a merge conflict result in an empty changesets, we now properly skip the changeset instead of crashing. Original patch from Robert Collins <robertc@robertcollins.net>. (please test the fix)
Bulk testing -> fixed