Re: Using rbt post to review topmost patch of mq series
Christian, No, my patch (also posted at https://reviews.reviewboard.org/r/5965/) only appears to fix the problem with parent diffs: I've had users complaining today both that valid patches won't upload or that the diff viewer doesn't work for existing reviews ("FileNotFoundError: The file 'code.py' (rNone) could not be found in the repository") - since the patches don't contain revision info. -- Bruce On Tue, Jun 10, 2014 at 5:54 PM, Christian Hammond wrote: > Hi Bruce, > > You may have seen a couple other threads today on the list about Mercurial > and issues with revisions. Based on those, and the testing you’ve done, do > you think this patch would fix those issues? > -- Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/ --- Sign up for Review Board hosting at RBCommons: https://rbcommons.com/ --- Happy user? Let us know at http://www.reviewboard.org/users/ --- You received this message because you are subscribed to the Google Groups "reviewboard" group. To unsubscribe from this group and stop receiving emails from it, send an email to reviewboard+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: Using rbt post to review topmost patch of mq series
Hi Bruce, You may have seen a couple other threads today on the list about Mercurial and issues with revisions. Based on those, and the testing you’ve done, do you think this patch would fix those issues? Christian -- Christian Hammond - christ...@beanbaginc.com Review Board - http://www.reviewboard.org Beanbag, Inc. - http://www.beanbaginc.com On June 9, 2014 at 3:38:55 PM, Bruce Cran (bruce.c...@gmail.com) wrote: I think the following should fix the problem: if we can't fetch a file using the revision recorded in the diff, try using the base commit id instead: diff --git a/reviewboard/scmtools/models.py b/reviewboard/scmtools/models.py index 7f1dd43..820814b 100644 --- a/reviewboard/scmtools/models.py +++ b/reviewboard/scmtools/models.py @@ -17,6 +17,7 @@ from reviewboard.scmtools.managers import RepositoryManager, ToolManager from reviewboard.scmtools.signals import (checked_file_exists, checking_file_exists, fetched_file, fetching_file) +from reviewboard.scmtools.core import FileNotFoundError from reviewboard.site.models import LocalSite @@ -367,7 +368,10 @@ class Repository(models.Model): revision, base_commit_id=base_commit_id) else: - data = self.get_scmtool().get_file(path, revision) + try: + data = self.get_scmtool().get_file(path, revision) + except FileNotFoundError: + data = self.get_scmtool().get_file(path, base_commit_id) log_timer.done() On Thu, Jun 5, 2014 at 2:50 PM, Bruce Cran wrote: I don't know if it used to work with RBTools, but people have said it used to work with 'hg postreview -m qparent'. Running 'rbt post tip' or 'rbt post --parent=qparent tip' results in Review Board saying it can't display the diff where there are changes in a file which also has changes in other patches (which, since they haven't been pushed yet, are unknown to RB). -- Bruce On Thu, Jun 5, 2014 at 2:37 PM, Christian Hammond wrote: I don’t know of anything on the RB side that would have changed here. It should just accept what’s given by the client. Were you able to do it before with RBTools? If you’re just trying to post the top-most commit, then (at least with Git) you’d do: $ rbt post HEAD Replace HEAD with whatever you need to identify the particular commit you care about. RBTools should take care of calculating the parent for you. -- Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/ --- Sign up for Review Board hosting at RBCommons: https://rbcommons.com/ --- Happy user? Let us know at http://www.reviewboard.org/users/ --- You received this message because you are subscribed to the Google Groups "reviewboard" group. To unsubscribe from this group and stop receiving emails from it, send an email to reviewboard+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout. -- Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/ --- Sign up for Review Board hosting at RBCommons: https://rbcommons.com/ --- Happy user? Let us know at http://www.reviewboard.org/users/ --- You received this message because you are subscribed to the Google Groups "reviewboard" group. To unsubscribe from this group and stop receiving emails from it, send an email to reviewboard+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: Using rbt post to review topmost patch of mq series
I think the following should fix the problem: if we can't fetch a file using the revision recorded in the diff, try using the base commit id instead: diff --git a/reviewboard/scmtools/models.py b/reviewboard/scmtools/models.py index 7f1dd43..820814b 100644 --- a/reviewboard/scmtools/models.py +++ b/reviewboard/scmtools/models.py @@ -17,6 +17,7 @@ from reviewboard.scmtools.managers import RepositoryManager, ToolManager from reviewboard.scmtools.signals import (checked_file_exists, checking_file_exists, fetched_file, fetching_file) +from reviewboard.scmtools.core import FileNotFoundError from reviewboard.site.models import LocalSite @@ -367,7 +368,10 @@ class Repository(models.Model): revision, base_commit_id=base_commit_id) else: -data = self.get_scmtool().get_file(path, revision) +try: +data = self.get_scmtool().get_file(path, revision) +except FileNotFoundError: +data = self.get_scmtool().get_file(path, base_commit_id) log_timer.done() On Thu, Jun 5, 2014 at 2:50 PM, Bruce Cran wrote: > I don't know if it used to work with RBTools, but people have said it used > to work with 'hg postreview -m qparent'. > Running 'rbt post tip' or 'rbt post --parent=qparent tip' results in > Review Board saying it can't display the diff where there are changes in a > file which also has changes in other patches (which, since they haven't > been pushed yet, are unknown to RB). > > -- > Bruce > > > > On Thu, Jun 5, 2014 at 2:37 PM, Christian Hammond < > christ...@beanbaginc.com> wrote: > > I don’t know of anything on the RB side that would have changed here. It >> should just accept what’s given by the client. Were you able to do it >> before with RBTools? >> >> If you’re just trying to post the top-most commit, then (at least with >> Git) you’d do: >> >> $ rbt post HEAD >> >> Replace HEAD with whatever you need to identify the particular commit you >> care about. RBTools should take care of calculating the parent for you. >> >> >> -- Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/ --- Sign up for Review Board hosting at RBCommons: https://rbcommons.com/ --- Happy user? Let us know at http://www.reviewboard.org/users/ --- You received this message because you are subscribed to the Google Groups "reviewboard" group. To unsubscribe from this group and stop receiving emails from it, send an email to reviewboard+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: Using rbt post to review topmost patch of mq series
I don't know if it used to work with RBTools, but people have said it used to work with 'hg postreview -m qparent'. Running 'rbt post tip' or 'rbt post --parent=qparent tip' results in Review Board saying it can't display the diff where there are changes in a file which also has changes in other patches (which, since they haven't been pushed yet, are unknown to RB). -- Bruce On Thu, Jun 5, 2014 at 2:37 PM, Christian Hammond wrote: I don’t know of anything on the RB side that would have changed here. It > should just accept what’s given by the client. Were you able to do it > before with RBTools? > > If you’re just trying to post the top-most commit, then (at least with > Git) you’d do: > > $ rbt post HEAD > > Replace HEAD with whatever you need to identify the particular commit you > care about. RBTools should take care of calculating the parent for you. > > > -- Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/ --- Sign up for Review Board hosting at RBCommons: https://rbcommons.com/ --- Happy user? Let us know at http://www.reviewboard.org/users/ --- You received this message because you are subscribed to the Google Groups "reviewboard" group. To unsubscribe from this group and stop receiving emails from it, send an email to reviewboard+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: Using rbt post to review topmost patch of mq series
I don’t know of anything on the RB side that would have changed here. It should just accept what’s given by the client. Were you able to do it before with RBTools? If you’re just trying to post the top-most commit, then (at least with Git) you’d do: $ rbt post HEAD Replace HEAD with whatever you need to identify the particular commit you care about. RBTools should take care of calculating the parent for you. Christian -- Christian Hammond - christ...@beanbaginc.com Review Board - http://www.reviewboard.org Beanbag, Inc. - http://www.beanbaginc.com On June 5, 2014 at 1:30:51 PM, Bruce Cran (bruce.c...@gmail.com) wrote: In 1.7, it was possible to generate a review request of the topmost changeset patch1 - e.g.: > hg qapplied patch1 patch2 patch3 patch4 ... by creating a diff of patch2-patch4 and uploading it via the web interface as the parent diff. How do I do the same in 2.x using 'rbt post'? Uploading the top-most patch via --diff-filename doesn't work, since it doesn't know about the other patches. Or, is this workflow not supported? -- Bruce -- Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/ --- Sign up for Review Board hosting at RBCommons: https://rbcommons.com/ --- Happy user? Let us know at http://www.reviewboard.org/users/ --- You received this message because you are subscribed to the Google Groups "reviewboard" group. To unsubscribe from this group and stop receiving emails from it, send an email to reviewboard+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout. -- Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/ --- Sign up for Review Board hosting at RBCommons: https://rbcommons.com/ --- Happy user? Let us know at http://www.reviewboard.org/users/ --- You received this message because you are subscribed to the Google Groups "reviewboard" group. To unsubscribe from this group and stop receiving emails from it, send an email to reviewboard+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.