view tests/test-topic-fold.t @ 6655:81fe0a498447 stable

topic: internal config option to fix hg pick behavior (issue6406) After some consideration, hg pick was made to always use current topic (and topic namespace), even if it's not set, in which case the resulting changeset will not have any topic. Previously the intended behavior was to only update topic if there was an active topic, and not touch commit extras at all otherwise. That wasn't ideal, since pick doesn't change active topic, and amending the just-picked commit would unset its topic without clear user's intent to do so and to their surprise.
author Anton Shestakov <av6@dwimlabs.net>
date Sat, 27 Jan 2024 17:36:39 -0300
parents caf302fb8f4d
children 2fbe91d762ef
line wrap: on
line source

test of the fold command
------------------------

  $ . $TESTDIR/testlib/common.sh
  $ cat >> $HGRCPATH <<EOF
  > [ui]
  > interactive = true
  > [extensions]
  > evolve =
  > topic =
  > EOF

  $ logtopic() {
  >    hg log -G -T "{rev}:{node}\ntopics: {topics}" 
  > }

Check that fold keep the topic if all revisions have the topic
--------------------------------------------------------------

  $ hg init testfold
  $ cd testfold
  $ mkcommit ROOT
  $ hg topic myfeature
  marked working directory as topic: myfeature
  $ mkcommit feature1
  active topic 'myfeature' grew its first changeset
  (see 'hg help topics' for more information)
  $ mkcommit feature2
  $ logtopic
  @  2:6e0fb07fa151928b633485a01b6d815b27f5a26d
  |  topics: myfeature
  o  1:95b29fc10be4e6343c8d344ad4354d6ff4098df1
  |  topics: myfeature
  o  0:ea207398892eb49e06441f10dda2a731f0450f20
     topics:
  $ hg fold --exact -r "(tip~1)::" -m "folded"
  2 changesets folded
  0 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ hg stack
  ### topic: myfeature
  ### target: default (branch)
  s1@ folded (current)
  s0^ ROOT (base)
  $ logtopic
  @  3:ba602426356f35854a83b02183d999749142443c
  |  topics: myfeature
  o  0:ea207398892eb49e06441f10dda2a731f0450f20
     topics:
  $ hg summary
  parent: 3:ba602426356f tip
   folded
  branch: default//myfeature
  commit: (clean)
  update: (current)
  phases: 2 draft
  topic:  myfeature

Check that fold dismis the topic if not all revisions have the topic
--------------------------------------------------------------------

(I'm not sure this behavior make senses, but now it is tested)

  $ hg topic --clear
  $ mkcommit feature3
  created new head
  (consider using topic for lightweight branches. See 'hg help topic')
  $ hg topic myotherfeature
  marked working directory as topic: myotherfeature
  $ mkcommit feature4
  active topic 'myotherfeature' grew its first changeset
  (see 'hg help topics' for more information)
  $ logtopic
  @  5:85e76d22bde1cb5ce44b00fc91a88cb805c93b1b
  |  topics: myotherfeature
  o  4:6508e0bfb6a188bb94d77c107f4e969291010b42
  |  topics:
  o  3:ba602426356f35854a83b02183d999749142443c
  |  topics: myfeature
  o  0:ea207398892eb49e06441f10dda2a731f0450f20
     topics:
  $ hg fold --exact -r "(tip~1)::" -m "folded 2"
  active topic 'myotherfeature' is now empty
  2 changesets folded
  clearing empty topic "myotherfeature"
  0 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ logtopic
  @  6:9fa327ea84f90ba000b75b90446680c993972df0
  |  topics:
  o  3:ba602426356f35854a83b02183d999749142443c
  |  topics: myfeature
  o  0:ea207398892eb49e06441f10dda2a731f0450f20
     topics:
  $ hg summary
  parent: 6:9fa327ea84f9 tip
   folded 2
  branch: default
  commit: (clean)
  update: (current)
  phases: 3 draft