view tests/test-namespaces-report.t @ 6935:954d7ea5cd67 stable tip

stack: when stack base is obsolete, pick any successor, even if at random

There are situations when s0 is obsolete and we also cannot pick just one
successor for it to use in stack. In such a case, let's pick the "latest"
successor from the first set.

We're assuming that obsutil.successorssets() returns data in the same order (it
should, since it makes sure to sort data internally). Keeping that in mind,
while the successor picked for s0 by this code is not based on any sort of
sophisticated logic, it should nonetheless be the same every time.

This patch is probably not going to completely break anything that was
previously working fine, because the previous behavior was to just abort with
an exception.

============================================================
Test detection of topic name space affected by a transaction
============================================================

Reporting affected topic namespaces in transactions

  $ . "$TESTDIR/testlib/common.sh"

  $ cat >> $HGRCPATH << EOF
  > [extensions]
  > evolve =
  > topic =
  > [phases]
  > publish = no
  > [devel]
  > tns-report-transactions = push
  > EOF

  $ hg init orig

case 1: new changeset (draft with topic namespace)
==================================================

topic namespace of that changeset is reported

  $ hg clone orig case-1 -q
  $ cd orig

  $ echo apple > a
  $ hg ci -qAm apple

  $ hg push ../case-1
  pushing to ../case-1
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files

  $ echo banana > b
  $ hg debug-topic-namespace bob
  marked working directory as topic namespace: bob
  $ hg topic b
  marked working directory as topic: b
  $ hg ci -qAm 'banana'

  $ hg push ../case-1
  pushing to ../case-1
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  topic namespaces affected: bob
  added 1 changesets with 1 changes to 1 files

  $ cd ..

case 2: obsmarker affecting known changeset
===========================================

topic namespaces of both the precursor and the successor are affected

  $ hg clone orig case-2 -q
  $ cd orig

  $ echo broccoli > b
  $ hg debug-topic-namespace bruce
  $ hg ci --amend -m 'broccoli'

  $ hg push ../case-2
  pushing to ../case-2
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  topic namespaces affected: bob bruce
  added 1 changesets with 1 changes to 1 files (+1 heads)
  1 new obsolescence markers
  obsoleted 1 changesets

  $ cd ..

case 3: phase divergence
========================

3 phase divergence resolution can point to a thing but not affect it (probably not affected)

In this case, the pushed changeset comes with an obsmarker whose predecessors
has the `charlie` topic-namespace and the successors has the `carol`
topic-namespace. However, that obsolescence is part of a phase-divergence
fixup, so we should now mark `coconut` as affected since it is already public.

  $ hg clone orig case-3 -q
  $ cd orig

  $ hg debug-topic-namespace charlie
  $ hg topic c
  $ echo coconut > c
  $ hg ci -qAm 'coconut'

  $ hg debug-topic-namespace carol
  $ echo cloudberry > c
  $ hg ci --amend -m 'cloudberry'

  $ hg phase --hidden -r 'desc("coconut")' --public
  1 new phase-divergent changesets

  $ hg evolve --phase-divergent
  recreate:[s1] cloudberry
  atop:[3] coconut
  committed as 9f1abc6f4a6f
  working directory is now at 9f1abc6f4a6f

  $ hg push ../case-3
  pushing to ../case-3
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  topic namespaces affected: bruce carol
  added 2 changesets with 2 changes to 1 files
  2 new obsolescence markers

  $ cd ..

case 4: phase movement: publishing drafts
=========================================

topic namespaces of published changesets are affected

  $ hg clone orig case-4 -q
  $ cd orig

  $ hg push ../case-4 --publish
  pushing to ../case-4
  searching for changes
  no changes found
  topic namespaces affected: carol
  active topic 'c' is now empty
  (use 'hg topic --clear' to clear it if needed)
  [1]

  $ cd ..

case 5: bookmark movement
=========================

Bookmark movement that affect tns (like putting a bookmark on obsolete
changesets) their topic namespaces reappear and are therefore reported

  $ hg clone orig case-5 -q
  $ cd orig

  $ hg debug-topic-namespace dana
  $ hg topic d
  $ echo durian > d
  $ hg ci -qAm 'durian'

  $ hg push ../case-5
  pushing to ../case-5
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  topic namespaces affected: dana
  added 1 changesets with 1 changes to 1 files

  $ hg debug-topic-namespace dave
  $ echo dragonfruit > d
  $ hg ci --amend -m 'dragonfruit'

  $ hg push ../case-5
  pushing to ../case-5
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  topic namespaces affected: dana dave
  added 1 changesets with 1 changes to 1 files (+1 heads)
  1 new obsolescence markers
  obsoleted 1 changesets

  $ hg bookmark --hidden -r 'desc("durian")' @
  bookmarking hidden changeset c56d89b2348b
  (hidden revision 'c56d89b2348b' was rewritten as: 7fc662c4767d)

  $ hg push ../case-5 -B @
  pushing to ../case-5
  searching for changes
  no changes found
  topic namespaces affected: dana
  exporting bookmark @
  [1]

  $ cd ..

case 6: phase movement: publishing secret changesets
====================================================

(that are known on the server)

topic namespaces of published changesets are affected

  $ hg clone orig case-6 -q
  $ cd orig

XXX: we see "active topic is now empty" twice because stack doesn't handle topic namespaces yet

  $ hg push ../case-6 -r . --publish
  pushing to ../case-6
  searching for changes
  no changes found
  topic namespaces affected: dave
  active topic 'd' is now empty
  active topic 'd' is now empty
  (use 'hg topic --clear' to clear it if needed)
  [1]

previous topic namespace is resurrected...

  $ hg phase --secret --force -r . --config 'devel.tns-report-transactions=phase'
  topic namespaces affected: dave
  active topic 'd' grew its first changeset
  (see 'hg help topics' for more information)

...just to disappear again

  $ hg push ../case-6 -r . --config 'devel.tns-report-transactions=*'
  pushing to ../case-6
  searching for changes
  no changes found
  topic namespaces affected: dave
  active topic 'd' is now empty
  (use 'hg topic --clear' to clear it if needed)
  [1]

  $ cd ..

case 7: phase movement: secret->draft on the server
===================================================

changeset becomes visible to peers, so its topic namespace is affected

  $ hg clone orig case-7 -q
  $ cd orig

  $ hg phase --draft --force -r tip
  active topic 'd' grew its first changeset
  (see 'hg help topics' for more information)
  $ hg phase --secret --force -r tip -R ../case-7
  active topic 'd' grew its first changeset
  (see 'hg help topics' for more information)

  $ hg push ../case-7 -r . --config 'devel.tns-report-transactions=*'
  pushing to ../case-7
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  topic namespaces affected: dave
  added 0 changesets with 0 changes to 1 files
  active topic 'd' grew its first changeset
  (see 'hg help topics' for more information)

  $ cd ..

case: 99 pushing obsmarker for an unknown changeset
===================================================
doesn't affect any topic namespace, we report nothing

  $ hg clone orig case-99 -q
  $ cd orig

  $ hg debugobsolete aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa `getid "desc('dragonfruit')"`
  1 new obsolescence markers

  $ hg push ../case-99
  pushing to ../case-99
  searching for changes
  no changes found
  1 new obsolescence markers
  [1]

  $ cd ..