Re: show all merge conflicts
- From
Philip Oakley <philipoakley@iee.org>
- Date
- Jan 28, 2017, 13:48 UTC
- Message-ID
- <C0FB9362D83842EB97B4C795C91F1775@PhilipOakley>
- In-Reply-To
- <CAAj3zPzO4+9t9_L2OXFmkihw-HwFvzybb7GXs4tTeFRyZHOaNQ@mail.gmail.com>
From: "G. Sylvie Davies" <sylvie@bit-booster.com>
Show 36 quoted lines
> On Fri, Jan 27, 2017 at 9:51 AM, Jeff King <peff@peff.net> wrote: >> On Fri, Jan 27, 2017 at 11:56:08AM -0500, Michael Spiegel wrote: >> >>> I'm trying to determine whether a merge required a conflict to resolve >>> after the merge has occurred. The git book has some advice >>> (https://git-scm.com/book/en/v2/Git-Tools-Advanced-Merging) to use >>> `git show` on the merge commit or use `git log --cc -p -1`. These >>> strategies work when the merge conflict was resolved with a change >>> that is different from either parent. When the conflict is resolved >>> with a change that is the same as one of the parents, then these >>> commands are indistinguishable from a merge that did not conflict. Is >>> it possible to distinguish between a conflict-free merge and a merge >>> conflict that is resolved by with the changes from one the parents? >> >> No. You'd have to replay the merge to know if it would have had >> conflicts. >> > > > Aside from the usual "git log -cc", I think this should work (replace > HEAD with whichever commit you are analyzing): > > git diff --name-only HEAD^2...HEAD^1 > m1 > git diff --name-only HEAD^1...HEAD^2 > b1 > git diff --name-only HEAD^1..HEAD > m2 > git diff --name-only HEAD^2..HEAD > b2 > > If files listed between m1 and b2 differ, then the merge is dirty. > Similarly for m2 and b1. > > More information here: > > http://stackoverflow.com/questions/27683077/how-do-you-detect-an-evil-merge-in-git/41356308#41356308 > > > - Sylvie
This feels as though there ought to be some sort of --left-right option to get an indication of which side various changes came from
Show 25 quoted lines
> >> There was a patch series a few years ago that added a new diff-mode to >> do exactly that, and show the diff against what was resolved. It had a >> few issues (I don't remember exactly what) and never got merged. >> >> Certainly one complication is that you don't know exactly _how_ the >> merge was done in the first place (e.g., which merge strategy, which >> custom merge drivers were in effect, etc). But in general, replaying >> with a standard merge-recursive would get you most of what you want to >> know. >> >> I've done this manually sometimes when digging into erroneous merges >> (e.g., somebody accidentally runs "git reset -- <paths>" in the middle >> of a merge and throws away some changes. >> >> You should be able to do: >> >> git checkout $merge^1 >> git merge $merge^2 >> git diff -R $merge >> >> to see what the original resolution did. >> >> -Peff >