Skip to content
Snippets Groups Projects

Linking objects from GFM references

Merged gitlab-qa-bot requested to merge github/fork/smashwilson/linking-issues into master

Created by: smashwilson

Hello!

This is an implementation in progress of linking issues from comments, also referenced in #2676 (closed), #2714 (closed), #1170 (closed), #302 (closed), and probably others. I'm pushing it out now to get early feedback on what I've already done and to hear any thoughts about edge cases I might not be accounting for.

Any mention of issues, merge requests, or commits via GitLab-flavored markdown references in descriptions, titles or follow-on notes to any of the above attaches a system note to the referenced object containing a backtracking link. Furthermore, pushing commits that match a (configurable) regexp to a project's default branch will close any issues mentioned in the closing phrase.

Left to do:

  • Only close issues from commits pushed to the project's default branch -- meaning, actually make it do that thing I said it does :wink:
  • If any Mentionable is edited, create additional cross-references for any references that weren't there to begin with.
  • Merge request and issue creation don't create cross-references correctly, only notes and edits. D'oh.
  • Accepting a merge request that brings in a commit that "fixes" an issue works fine, but if the "fixes" text is in a line other than the first, it happens without warning. Detect and add an "accepting this request will close issues X, Y, and Z" banner on the merge request page.
  • Look into using ActiveRecord::Dirty to implement Mentionable#watch_references, which would let us push reference updating into the observers where it belongs.
  • Duplicate "mentioned in commit..." notes when a commit that mentions a thing is merged into a new branch.
  • Link an issue closed notification back to the commit that closed it, when closing from a commit. "Status changed to closed by..."
  • Catch references in the commits in the first push to a new branch.
  • Do a final rebase and squash on upstream master.

Merge request reports

Loading
Loading

Activity

Filter activity
  • Approvals
  • Assignees & reviewers
  • Comments (from bots)
  • Comments (from users)
  • Commits & branches
  • Edits
  • Labels
  • Lock status
  • Mentions
  • Merge request status
  • Tracking
  • Created by: coveralls

    Coverage Status

    Coverage decreased (-0%) when pulling 4d70f35db1a775d3a05a160525c543a2d5d0d4a7 on smashwilson:linking-issues into 5315fd5a on gitlabhq:master.

    By Administrator on 2013-07-07T01:54:16 (imported from GitLab project)

    By Administrator on 2013-07-07T01:54:16 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage remained the same when pulling e44e50bd2da250c493abe7353efaa65982556de6 on smashwilson:linking-issues into 5315fd5a on gitlabhq:master.

    By Administrator on 2013-07-07T14:48:04 (imported from GitLab project)

    By Administrator on 2013-07-07T14:48:04 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage remained the same when pulling 100fca5bb0b8a76f445448c5114c25456a115113 on smashwilson:linking-issues into 5315fd5a on gitlabhq:master.

    By Administrator on 2013-07-07T18:20:12 (imported from GitLab project)

    By Administrator on 2013-07-07T18:20:12 (imported from GitLab)

  • Created by: dzaporozhets

    @smashwilson cool. Looking forward

    By Administrator on 2013-07-12T13:08:25 (imported from GitLab project)

    By Administrator on 2013-07-12T13:08:25 (imported from GitLab)

  • Created by: smashwilson

    Some screenshots of an issue cross-reference note: xref

    ...and the merge request "will close issues" notification. (The issue ids are proper GFM links.) issue-close-notification

    By Administrator on 2013-07-27T18:43:29 (imported from GitLab project)

    By Administrator on 2013-07-27T18:43:29 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage remained the same when pulling f07b2fdf935983be17bf737f2d9e35f656f8ec0b on smashwilson:linking-issues into 79bea312 on gitlabhq:master.

    By Administrator on 2013-07-29T21:32:55 (imported from GitLab project)

    By Administrator on 2013-07-29T21:32:55 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage decreased (-0%) when pulling ff2c52daafd7d530ffddd1a2de687ac182f321d0 on smashwilson:linking-issues into 79bea312 on gitlabhq:master.

    By Administrator on 2013-07-29T23:12:50 (imported from GitLab project)

    By Administrator on 2013-07-29T23:12:50 (imported from GitLab)

  • Created by: Razer6

    :thumbsup:

    By Administrator on 2013-08-02T18:09:47 (imported from GitLab project)

    By Administrator on 2013-08-02T18:09:47 (imported from GitLab)

  • Created by: fprochazka

    :thumbsup:

    By Administrator on 2013-08-02T18:14:31 (imported from GitLab project)

    By Administrator on 2013-08-02T18:14:31 (imported from GitLab)

  • Created by: archonik

    Definitely :thumbsup:

    By Administrator on 2013-08-02T19:32:57 (imported from GitLab project)

    By Administrator on 2013-08-02T19:32:57 (imported from GitLab)

  • Created by: Morgul

    This looks great. I'm really excited about it!

    By Administrator on 2013-08-02T20:06:11 (imported from GitLab project)

    By Administrator on 2013-08-02T20:06:11 (imported from GitLab)

  • Created by: jtreml

    :thumbsup:

    By Administrator on 2013-08-02T20:15:52 (imported from GitLab project)

    By Administrator on 2013-08-02T20:15:52 (imported from GitLab)

  • Created by: smashwilson

    Almost ready -- I've got one more bug that keeps poking through the specs. Apparently, faker-generated issue text will very occasionally include a word that the ReferenceExtractor recognizes as a commit when the repository isn't stubbed out to handle it. Also I have problems running more than a few of the Spinach features at a time in my local Vagrant VM so there might be other bugs lurking there.

    I'm hoping to have some time to put the finishing touches on this this weekend. Thanks for the encouragement :grinning:

    By Administrator on 2013-08-02T20:49:55 (imported from GitLab project)

    By Administrator on 2013-08-02T20:49:55 (imported from GitLab)

  • Created by: asyd

    :thumbsup:

    By Administrator on 2013-08-03T15:55:23 (imported from GitLab project)

    By Administrator on 2013-08-03T15:55:23 (imported from GitLab)

  • Created by: smashwilson

    Currently flushing out all of the problem spots by temporarily hardcoding the sentence factory to always contain a bad commit reference, and fixing them by stubbing with wild abandon. The specs are clean now, working on the features next.

    By Administrator on 2013-08-03T20:02:34 (imported from GitLab project)

    By Administrator on 2013-08-03T20:02:34 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage increased (+0%) when pulling ab16034063557bc8debdf2e0c37fea1211b37b5c on smashwilson:linking-issues into 8886b900 on gitlabhq:master.

    By Administrator on 2013-08-03T21:04:38 (imported from GitLab project)

    By Administrator on 2013-08-03T21:04:38 (imported from GitLab)

  • Created by: smashwilson

    Hmm -- that failure looks unrelated to me, and I can't replicate locally.

    @randx and others: ready for any feedback you've got to offer on this, whenever you get the chance.

    By Administrator on 2013-08-03T21:20:44 (imported from GitLab project)

    By Administrator on 2013-08-03T21:20:44 (imported from GitLab)

  • Created by: MrKeiKun

    Current Status Please?

    By Administrator on 2013-08-08T08:46:37 (imported from GitLab project)

    By Administrator on 2013-08-08T08:46:37 (imported from GitLab)

  • Created by: smashwilson

    @MrKeiKun: Functionally complete, awaiting initial feedback. It's a big PR, too, so I'd be surprised if there weren't a few problems to sort out before it's ready to merge. (I'd also be happy to re-rebase if it drifts too far from master.)

    By Administrator on 2013-08-09T20:27:21 (imported from GitLab project)

    By Administrator on 2013-08-09T20:27:21 (imported from GitLab)

  • Created by: rtripault

    :thumbsup:

    By Administrator on 2013-08-15T11:52:14 (imported from GitLab project)

    By Administrator on 2013-08-15T11:52:14 (imported from GitLab)

  • Created by: MrKeiKun

    :thumbsup:

    By Administrator on 2013-08-19T13:59:31 (imported from GitLab project)

    By Administrator on 2013-08-19T13:59:31 (imported from GitLab)

  • Created by: dzaporozhets

    scheduled for 6.1. After 6.0 being released I'll get back to this PR

    By Administrator on 2013-08-19T16:28:23 (imported from GitLab project)

    By Administrator on 2013-08-19T16:28:23 (imported from GitLab)

  • Created by: dzaporozhets

    I'm highly interested in this PR :)

    By Administrator on 2013-08-19T16:30:04 (imported from GitLab project)

    By Administrator on 2013-08-19T16:30:04 (imported from GitLab)

  • Created by: CiTroNaK

    :thumbsup:

    By Administrator on 2013-08-20T06:28:10 (imported from GitLab project)

    By Administrator on 2013-08-20T06:28:10 (imported from GitLab)

  • gitlab-qa-bot
65 68 end
66 69 end
67 70
71 # Regular expression that identifies commit message clauses that trigger issue closing.
72 def issue_closing_regex
73 @issue_closing_regex ||= Regexp.new(Gitlab.config.gitlab.issue_closing_pattern)
  • Created by: dzaporozhets

    Please dont use @@ variables

    By Administrator on 2013-08-25T22:59:08 (imported from GitLab project)

    By Administrator on 2013-08-25T22:59:08 (imported from GitLab)

  • Created by: dzaporozhets

    @smashwilson Looks good. Can you please make it mergeable. After ping me - I'll test it and merge if works.

    By Administrator on 2013-08-20T19:25:30 (imported from GitLab project)

    By Administrator on 2013-08-20T19:25:30 (imported from GitLab)

  • Created by: smashwilson

    Hmmm, I re-rebased, but it looks like the MergeRequest model changes broke some things. I'll be working on it today.

    By Administrator on 2013-08-23T15:30:33 (imported from GitLab project)

    By Administrator on 2013-08-23T15:30:33 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage decreased (-0.18%) when pulling 5ff44274f6b60413f6def79f28a8dd8c8e7c6b87 on smashwilson:linking-issues into 21e3d848 on gitlabhq:master.

    By Administrator on 2013-08-23T16:53:48 (imported from GitLab project)

    By Administrator on 2013-08-23T16:53:48 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage decreased (-12.74%) when pulling bdeb7d67d4222814f39fb2d1f3f46843088e45cf on smashwilson:linking-issues into 21e3d848 on gitlabhq:master.

    By Administrator on 2013-08-23T17:52:13 (imported from GitLab project)

    By Administrator on 2013-08-23T17:52:13 (imported from GitLab)

  • Created by: MrKeiKun

    Ready to test if on gitlabhq/gitlabhq:master

    By Administrator on 2013-08-23T17:52:10 (imported from GitLab project)

    By Administrator on 2013-08-23T17:52:10 (imported from GitLab)

  • Created by: smashwilson

    I'm getting some weird issues in the specs and features. It looks like ActivityObserver is expecting Thread.current[:current_user] to be set in a few places where it isn't being populated. What's weirder is that the features occasionally pass anyway.

    I'm also not sure yet what's going on with the Event count tests in activity_observer_spec that are creating Events for system notes (but only sometimes). Maybe there are factories being triggered during the block?

    By Administrator on 2013-08-23T18:04:04 (imported from GitLab project)

    By Administrator on 2013-08-23T18:04:04 (imported from GitLab)

  • Created by: smashwilson

    Ah! Yes - I bet the :issue factories are being invoked within that block and creating Events that shouldn't be there. I'll re-push in a minute.

    By Administrator on 2013-08-23T18:06:49 (imported from GitLab project)

    By Administrator on 2013-08-23T18:06:49 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage decreased (-12.99%) when pulling e57c56c1d448b896de7760fa329ddab8c4a0c808 on smashwilson:linking-issues into 21e3d848 on gitlabhq:master.

    By Administrator on 2013-08-23T18:33:05 (imported from GitLab project)

    By Administrator on 2013-08-23T18:33:05 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage decreased (-12.79%) when pulling 73d45c2f3cf81f5b35c776b0e21cc7de2e309a49 on smashwilson:linking-issues into 21e3d848 on gitlabhq:master.

    By Administrator on 2013-08-23T19:24:17 (imported from GitLab project)

    By Administrator on 2013-08-23T19:24:17 (imported from GitLab)

  • Created by: smashwilson

    There we go! At least, I think that last failure in the 2.0 build is unrelated. @randx, it should be ready to go, let me know if you find any other problems.

    By Administrator on 2013-08-23T19:26:39 (imported from GitLab project)

    By Administrator on 2013-08-23T19:26:39 (imported from GitLab)

  • Created by: dzaporozhets

    @smashwilson thank you. I'll test it and give you some feedback

    By Administrator on 2013-08-25T19:24:42 (imported from GitLab project)

    By Administrator on 2013-08-25T19:24:42 (imported from GitLab)

  • gitlab-qa-bot
  • 56 57
    57 58 # Both open and reopened issues should be listed as opened
    58 59 scope :opened, -> { with_state(:opened, :reopened) }
    60
    61 # Mentionable overrides.
    62
    63 def gfm_reference
    64 "issue ##{iid}"
  • gitlab-qa-bot
  • 255 255 target_project
    256 256 end
    257 257
    258 # Return the set of issues that will be closed if this merge request is accepted.
    259 def closes_issues
    260 if target_branch == project.default_branch
    261 unmerged_commits.map { |c| c.closes_issues(project) }.flatten.uniq.sort_by(&:id)
    262 else
    263 []
    264 end
    265 end
    266
    267 # Mentionable override.
    268 def gfm_reference
    269 "merge request !#{iid}"
    • Created by: dzaporozhets

      same here

      By Administrator on 2013-08-25T22:59:08 (imported from GitLab project)

      By Administrator on 2013-08-25T22:59:08 (imported from GitLab)

  • Created by: dzaporozhets

    @smashwilson I tested this PR briefly and I really like it. Ping me if you need help with iid and we will merge this one

    By Administrator on 2013-08-25T19:41:15 (imported from GitLab project)

    By Administrator on 2013-08-25T19:41:15 (imported from GitLab)

  • Created by: smashwilson

    @randx Thanks! This one was fun to code. :grinning:

    I changed it to use iid for Issues and MergeRequests and updated the specs accordingly. I think I got them all, but If Travis finds more problems I'll keep at it.

    By Administrator on 2013-08-25T21:46:00 (imported from GitLab project)

    By Administrator on 2013-08-25T21:46:00 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage increased (+0.22%) when pulling 26a42d16c1c28126e4df5f75e28735dff70f7e46 on smashwilson:linking-issues into 2b36dee6 on gitlabhq:master.

    By Administrator on 2013-08-25T22:13:40 (imported from GitLab project)

    By Administrator on 2013-08-25T22:13:40 (imported from GitLab)

  • Created by: smashwilson

    @randx The build is green. Let me know if you think of anything else!

    By Administrator on 2013-08-25T22:16:26 (imported from GitLab project)

    By Administrator on 2013-08-25T22:16:26 (imported from GitLab)

  • Created by: smashwilson

    Ah, I missed one iid reference: the issue GFM links in the "accepting this merge request will close..." banner.

    By Administrator on 2013-08-25T22:59:58 (imported from GitLab project)

    By Administrator on 2013-08-25T22:59:58 (imported from GitLab)

  • Created by: coveralls

    Coverage Status

    Coverage increased (+0.22%) when pulling c8a115c0 on smashwilson:linking-issues into 2b36dee6 on gitlabhq:master.

    By Administrator on 2013-08-25T23:27:14 (imported from GitLab project)

    By Administrator on 2013-08-25T23:27:14 (imported from GitLab)

  • Created by: dzaporozhets

    @smashwilson thank you. I've merged this one. If you find any missed referenced to iid - please send PR

    By Administrator on 2013-08-26T09:32:34 (imported from GitLab project)

    By Administrator on 2013-08-26T09:32:34 (imported from GitLab)

  • Created by: rufinus

    @smashwilson Hi sorry to write this here, didnt find a other way to contact you. As written in http://feedback.gitlab.com/forums/176466-general/suggestions/3692754-linking-issues-from-comments-and-automatically-clo?tracking_code=9df9f2ef47771388afbbd2da3e44a130 i want to reward your effort with USD 100 $. How can get the money to you? (PayPal, Skrill, Amazon all fine)

    By Administrator on 2013-08-29T14:28:48 (imported from GitLab project)

    By Administrator on 2013-08-29T14:28:48 (imported from GitLab)

  • Created by: smashwilson

    @rufinus: I appreciate the thought, but I'm just a guy who pokes at things on nights and weekends. I'd rather see the money go to a charity or an open-source project that you use or like, who have needs for hosting and other upkeep... like GitLab :grinning:

    By Administrator on 2013-09-15T20:05:32 (imported from GitLab project)

    By Administrator on 2013-09-15T20:05:32 (imported from GitLab)

  • Created by: rufinus

    @smashwilson as you wish, I donated it to gitlab via pledgie. ( http://pledgie.com/campaigns/17027 )

    By Administrator on 2013-09-15T23:42:36 (imported from GitLab project)

    By Administrator on 2013-09-15T23:42:36 (imported from GitLab)

  • Created by: dosire

    GitLab received the money, thank you @rufinus and @smashwilson !

    By Administrator on 2013-09-16T07:48:40 (imported from GitLab project)

    By Administrator on 2013-09-16T07:48:40 (imported from GitLab)

  • Created by: rgbkrk

    Wowsers. @rufinus and @smashwilson -- you are both so awesome!

    :thumbsup:

    By Administrator on 2013-09-19T18:45:08 (imported from GitLab project)

    By Administrator on 2013-09-19T18:45:08 (imported from GitLab)

  • Created by: t-oster

    The feature does not seem to be present in 6.1. Did I do anything wrong?

    By Administrator on 2013-09-24T12:43:20 (imported from GitLab project)

    By Administrator on 2013-09-24T12:43:20 (imported from GitLab)

  • Created by: edude03

    I updated yesterday and it works for me. Note that the default regex used to test commits is set to only work if the closes / fixes comes at the beginning of the message.

    Michael Francis Sent with Sparrow (http://www.sparrowmailapp.com/?sig)

    On Tuesday, September 24, 2013 at 8:43 AM, Thomas Oster wrote:

    The feature does not seem to be present in 6.1. Did I do anything wrong?

    — Reply to this email directly or view it on GitHub (https://github.com/gitlabhq/gitlabhq/pull/4507#issuecomment-24999326).

    By Administrator on 2013-09-24T14:27:05 (imported from GitLab project)

    By Administrator on 2013-09-24T14:27:05 (imported from GitLab)

  • Created by: mikpalmen

    Im sorry if this post is in the wrong place.. but

    I love this fix/update to close issues thrue commits.

    I also altered this after update issue_closing_pattern: (?i)(([)\s_(close|fix)(s|es|d|ed)\s#\d+\s_(])|(()\s_(close|fix)(s|es|d|ed)\s#\d+\s_())) in commit.rb

    and yes it now works with inline.. -m 'this (fixes #12 (closed)) issue and thats good'

    but i also want the multiple issues in one commit fixed.

    -m 'this (fixes #32 (closed)) and also (closes #43 (closed))'

    Solved it with this. Anyone know how i can still make use of the preset variable? this dont work. md = safe_message.scan(Gitlab.config.gitlab.issue_closing_pattern)

     def closes_issues project
        md = safe_message.scan(/(?i)((\[)\s*(close|fix)(s|es|d|ed)*\s*#\d+\s*(\])|(\()\s*(close|fix)(s|es|d|ed)*\s*#\d+\s*(\)))/)
        #md = issue_closing_regex.match(safe_message)
        if md
          extractor = Gitlab::ReferenceExtractor.new
          md.each do |n|
           extractor.analyze(n[0])
          end
          extractor.issues_for(project)
          #extractor = Gitlab::ReferenceExtractor.new
          #extractor.analyze(md[0])
          #extractor.issues_for(project)
        else
          []
        end
      end

    By Administrator on 2013-09-25T22:08:51 (imported from GitLab project)

    By Administrator on 2013-09-25T22:08:51 (imported from GitLab)

  • Created by: t-oster

    I tried it in a different install and it works like a charm. Thank you for this great feature. I have to check what was the problem... maybe I didn't make the reference in the first line....

    By Administrator on 2013-09-25T14:45:54 (imported from GitLab project)

    By Administrator on 2013-09-25T14:45:54 (imported from GitLab)

  • Created by: t-oster

    The problem was, that the redis-cli executable was on /usr/local/bin instead of /usr/bin, so gitlab-shell couldn't find it. Since this did not show up in any logs or checks, maybe it's worth adding it to the check-script?

    By Administrator on 2013-10-01T06:06:00 (imported from GitLab project)

    By Administrator on 2013-10-01T06:06:00 (imported from GitLab)

  • Created by: dosire

    @t-oster That is a great idea, feel free to make a merge request for this.

    By Administrator on 2013-10-01T07:59:30 (imported from GitLab project)

    By Administrator on 2013-10-01T07:59:30 (imported from GitLab)

  • Created by: t-oster

    Hi...i have a working solution, but i have to test it.I will submit my merge request this evening.

    By Administrator on 2013-10-01T14:52:32 (imported from GitLab project)

    By Administrator on 2013-10-01T14:52:32 (imported from GitLab)

  • Created by: dosire

    @t-oster Ok, great!

    By Administrator on 2013-10-01T14:56:10 (imported from GitLab project)

    By Administrator on 2013-10-01T14:56:10 (imported from GitLab)

  • Created by: CodeBrauer

    :thumbsup:

    By Administrator on 2013-12-10T13:38:43 (imported from GitLab project)

    By Administrator on 2013-12-10T13:38:43 (imported from GitLab)

  • Created by: ssehovic

    @smashwilson I have one question. Your implementation works on default (eg. master) branch but if I create another one let's say develop branch and trying to close an issuse on that branch closing issue does not work. Is it possible to make issue close on every branch not only master?

    By Administrator on 2014-04-16T08:56:36 (imported from GitLab project)

    By Administrator on 2014-04-16T08:56:36 (imported from GitLab)

  • Created by: smashwilson

    @ssehovic The idea is that if you're using something like GitHub flow, you'll be pushing commits on other branches out to GitLab before they've been peer-reviewed and are ready for integration, so you'll have the "fixes" comment on a commit pushed well before you've actually dealt with the issue.

    Currently you can't configure it to behave differently. Rather than toggle it to handle issue-closing commits on all branches, it might be interesting to do something like use a regex match against branch names, or handle them on all protected branches...

    In the meantime, if you're using a "develop" branch as your integration target, is there any reason you couldn't make "develop" the default branch, instead?

    By Administrator on 2014-04-16T12:29:55 (imported from GitLab project)

    By Administrator on 2014-04-16T12:29:55 (imported from GitLab)

  • Created by: ssehovic

    @smashwilson Thank you for reply. I am using git-flow "philosophy" right now and main branch is develop so I was hoping that this will work. Best regards.

    By Administrator on 2014-04-16T12:51:20 (imported from GitLab project)

    By Administrator on 2014-04-16T12:51:20 (imported from GitLab)

  • Created by: dosire

    @ssehovic Have you made develop your default branch as @smashwilson asked? (see project => settings)

    By Administrator on 2014-04-16T13:55:41 (imported from GitLab project)

    By Administrator on 2014-04-16T13:55:41 (imported from GitLab)

  • Created by: ssehovic

    @dosire Yes I have and by changing develop to default branch issue closing is working, but I was hoping that issue closing will be available in any branch :-) Btw, is issue always "related" to default branch in gitlab?

    By Administrator on 2014-04-16T14:13:38 (imported from GitLab project)

    By Administrator on 2014-04-16T14:13:38 (imported from GitLab)

  • Created by: dosire

    @ssehovic For now we'll keep it related to the default branch. If you have issues not related to the default branch you'll have to close them manually.

    By Administrator on 2014-04-16T14:28:30 (imported from GitLab project)

    By Administrator on 2014-04-16T14:28:30 (imported from GitLab)

  • Created by: CodeBrauer

    Whats about closing multiple issues by one commit? fixes #39, fixes #36 closed only the first one - the 2nd just got a "mentioned"

    By Administrator on 2014-06-17T15:23:53 (imported from GitLab project)

    By Administrator on 2014-06-17T15:23:53 (imported from GitLab)

  • Created by: dosire

    @GabrielWanzek Feels broken to me, feel free to send an MR to fix this.

    By Administrator on 2014-06-17T15:25:01 (imported from GitLab project)

    By Administrator on 2014-06-17T15:25:01 (imported from GitLab)

  • Created by: CodeBrauer

    @dosire , Isn't it in the code of @mrPalm ?

    https://github.com/gitlabhq/gitlabhq/pull/4507#issuecomment-25083909

    BTW: I'm not a Ruby man :)

    By Administrator on 2014-06-17T15:27:13 (imported from GitLab project)

    By Administrator on 2014-06-17T15:27:13 (imported from GitLab)

  • Created by: dosire

    @GabrielWanzek Could be, but it is not a regression so there is no designated party to fix it.

    By Administrator on 2014-06-17T19:29:00 (imported from GitLab project)

    By Administrator on 2014-06-17T19:29:00 (imported from GitLab)

  • Created by: ds82

    Hey guys, I'm sorry if this is the wrong place to post this, but I did not find any better.

    While I love the auto-close ./. auto-reference-issues feature of gitlab, I really miss it when working in branches. I agree that issues should not be (auto-)closed if I push a fix to a branch, but they should be referenced.

    If I push a commit to the master (aka default branch) of a project gitlab closes the issue, if I use the magic keywords ( close[s], fixe[s|d] ) and a issue number. If I just use the issue number, it creates a reference in the issue linking to the commit.

    If I push a commit to a different branch (aka not the default branch) and using the magic keywords - nothing happens at all. If I just use the issue number, gitlab creates a reference to the commit in the branch.

    What I'am suggesting here is, that gitlab should always create at least a reference. That means, if I push to a different branch than the default branch using a commit message like "foo bar, fixes #XX", gitlab should create a reference in issue #XX linking to that commit in the branch. When this commit is merged to the master, it should auto-close the issue and link to the commit in the master branch.

    By Administrator on 2014-07-22T06:54:16 (imported from GitLab project)

    By Administrator on 2014-07-22T06:54:16 (imported from GitLab)

  • Created by: dosire

    @ds82 I like your idea, please post to feedback.gitlab.com or create a tested merge request

    By Administrator on 2014-07-22T06:56:40 (imported from GitLab project)

    By Administrator on 2014-07-22T06:56:40 (imported from GitLab)

  • Created by: rumpelsepp

    I also like @ds82 idea. +1

    By Administrator on 2014-10-28T10:47:36 (imported from GitLab project)

    By Administrator on 2014-10-28T10:47:36 (imported from GitLab)

  • Created by: aminadha

    +1 @ds82 idea

    By Administrator on 2015-02-01T06:02:38 (imported from GitLab project)

    By Administrator on 2015-02-01T06:02:38 (imported from GitLab)

  • Created by: ambroisemaupate

    +1 @ds82 idea

    By Administrator on 2015-03-30T12:22:55 (imported from GitLab project)

    By Administrator on 2015-03-30T12:22:55 (imported from GitLab)

  • Please register or sign in to reply
    Loading