view tests/test-evolve-order.t @ 3846:f9dad99a90d5

evolve: create a new commit instead of amending one of the divergents

This patch changes the behavior of evolve command while resolving
content-divergence to create a new commit instead of amending one of the
divergent ones.

In past, I have made this change, backed out this change and now today again I
am doing this change, so let's dive in some history.

Using cmdrewrite.amend() was never a good option as that requires hack to delete
the evolvestate and also gives us less control over things. We can't make the
commit on top of different parents as that of content-divergent ones. Due to all
these, I first made this change to create a new commit instead of amending one.

But, after few days, there was flakiness observed in the tests and turned out
that we need to do some dirstate dance as repo.dirstate.setparents() does not
always fix the dirstate. That flakiness was a blocker for progress at that time
and we decided to switch to amend back so that we can have things working with
some hacks and we can later fix the implementation part.

Now, yesterday while tackling resolving content-divergence of a stack which is
as follows:

C1 C2
 | |
B1 B2
 | |
A1 A2
 \/
base

where, A1-A2, B1-B2, C1-C2 are content-divergent with each other. Now we can
resolve A1-A2 very well because they have the same parent and let's say that
resolution leads to A3.
Now, we want to resolve B1-B2 and make the new resolution commit on top of A3 so
that we can end up something like:

C3
 |
B3
 |
A3
 |
base

however, amending one of the divergent changesets, it's not possible to create a
commit on a different parent like A3 here without some relocation. We should
prevent relocation as that may leads to some conflicts and should change the
parent before committing.

So, looking ahead, we can't move with using amend as still using that we will
need some relocation hacks making code ugly and prone to bad behaviors, bugs.
Let's change back to creating a new commit so that we can move forward in a good
way.

About repo.dirstate.setparents() not setting the dirstate, I have researched
yesterday night about how we can do that and found out that we can use
cmdrewrite._uncommitdirstate() here. Expect upcoming patches to improve the
documentation of that function.

There are lot of test changes because of change in hash but there is no behavior
change. The only behavior change is in test-evolve-abort-contentdiv.t which is
nice because creating a new commit helps us in stripping that while aborting.
We have a lot of testing of content-divergence and no behavior change gives
enough confidence for making this change.

I reviewed the patch carefully to make sure there is no behavior change and I
suggest reviewer to do the same.

evolve --rev reordering
-----------------------

  $ cat >> $HGRCPATH <<EOF
  > [defaults]
  > amend=-d "0 0"
  > fold=-d "0 0"
  > [web]
  > push_ssl = false
  > allow_push = *
  > [phases]
  > publish = False
  > [diff]
  > git = 1
  > unified = 0
  > [ui]
  > logtemplate = {rev}:{node|short}@{branch}({phase}) {desc|firstline}\n
  > [extensions]
  > EOF
  $ echo "evolve=$(echo $(dirname $TESTDIR))/hgext3rd/evolve/" >> $HGRCPATH
  $ mkcommit() {
  >    echo "$1" > "$1"
  >    hg add "$1"
  >    hg ci -m "add $1"
  > }

  $ mkstack() {
  >    # Creates a stack of commit based on $1 with messages from $2, $3 ..
  >    hg update "$1" -C
  >    shift
  >    mkcommits $*
  > }

  $ mkcommits() {
  >   for i in $@; do mkcommit $i ; done
  > }

Initial setup
  $ hg init testrevorder
  $ cd testrevorder
  $ mkcommits p _a _b _c
  $ hg phase --public 0
  $ hg up 'desc(_a)'
  0 files updated, 0 files merged, 2 files removed, 0 files unresolved
  $ echo "aaa" > _a
  $ hg amend
  2 new orphan changesets
  $ hg log -G
  @  4:12d519679175@default(draft) add _a
  |
  | *  3:4d156641b718@default(draft) add _c
  | |
  | *  2:4d7242ebb004@default(draft) add _b
  | |
  | x  1:2d73fcd7f07d@default(draft) add _a
  |/
  o  0:f92638be10c7@default(public) add p
  

evolve --rev reorders the rev to solve instability, trivial case 2 revs wrong order
  $ hg evolve --rev 'desc(_c) + desc(_b)'
  move:[2] add _b
  atop:[4] add _a
  move:[3] add _c
  atop:[5] add _b
  working directory is now at 52b8f9b04f83

evolve --rev reorders the rev to solve instability. Harder case, obsolescence
accross three stacks in growing rev numbers.
  $ hg up "desc(_c)"
  0 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ mkcommit _d
  $ hg up "desc(_a)"
  0 files updated, 0 files merged, 3 files removed, 0 files unresolved
  $ hg amend -m "aprime"
  3 new orphan changesets
  $ hg evolve --rev "desc(_b)"
  move:[5] add _b
  atop:[8] aprime
  working directory is now at 476c9c052aae
  $ hg up "desc(_b) - obsolete()"
  0 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ hg amend -m "bprime"
  $ hg up "desc(aprime)"
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ hg amend -m "asecond"
  1 new orphan changesets
  $ hg log -G
  @  11:9a584314f3f3@default(draft) asecond
  |
  | *  10:a59c79776f7c@default(draft) bprime
  | |
  | x  8:81a687b96d4d@default(draft) aprime
  |/
  | *  7:464731bc0ed0@default(draft) add _d
  | |
  | *  6:52b8f9b04f83@default(draft) add _c
  | |
  | x  5:59476c3836ef@default(draft) add _b
  | |
  | x  4:12d519679175@default(draft) add _a
  |/
  o  0:f92638be10c7@default(public) add p
  
  $ hg evolve --rev "orphan()"
  move:[10] bprime
  atop:[11] asecond
  move:[6] add _c
  atop:[12] bprime
  move:[7] add _d
  atop:[13] add _c
  working directory is now at 739f18ac1d03
  $ hg log -G
  @  14:739f18ac1d03@default(draft) add _d
  |
  o  13:e5960578d158@default(draft) add _c
  |
  o  12:4ad33fa88946@default(draft) bprime
  |
  o  11:9a584314f3f3@default(draft) asecond
  |
  o  0:f92638be10c7@default(public) add p
  

Evolve --rev more complex case: two sets of stacks one with prune an no successor, the other one
partially solvable
 
First set of stack:
  $ mkstack "desc(_d)" c1_ c2_ c3_ c4_ >/dev/null
  $ mkstack "desc(_d)" c1prime c2prime >/dev/null
  $ mkstack "desc(_d)" c1second >/dev/null
  $ hg prune "desc(c1_)" -s "desc(c1prime)"
  1 changesets pruned
  3 new orphan changesets
  $ hg prune "desc(c2_)" -s "desc(c2prime)"
  1 changesets pruned
  $ hg prune "desc(c1prime)" -s "desc(c1second)"
  1 changesets pruned
  1 new orphan changesets
  $ hg log -G -r "desc(_d)::"
  @  21:dcf786e878fd@default(draft) add c1second
  |
  | *  20:507d52d715f6@default(draft) add c2prime
  | |
  | x  19:c995cb124ddc@default(draft) add c1prime
  |/
  | *  18:d096a2437fd0@default(draft) add c4_
  | |
  | *  17:cde95c6cba7a@default(draft) add c3_
  | |
  | x  16:e0d9f7a099fe@default(draft) add c2_
  | |
  | x  15:43b7c338b1f8@default(draft) add c1_
  |/
  o  14:739f18ac1d03@default(draft) add _d
  |
  ~

Second set of stack with no successor for b2_:
  $ mkstack "desc(_d)" b1_ b2_ b3_ b4_ >/dev/null
  $ mkstack "desc(_d)" b1prime b3prime >/dev/null
  $ hg prune "desc(b1_)" -s "desc(b1prime)"
  1 changesets pruned
  3 new orphan changesets
  $ hg prune "desc(b3_)" -s "desc(b3prime)"
  1 changesets pruned
  $ hg prune "desc(b2_)"
  1 changesets pruned

  $ hg log -G -r "desc(_d)::"
  @  27:b253ff5b65d1@default(draft) add b3prime
  |
  o  26:4acf61f11dfb@default(draft) add b1prime
  |
  | *  25:594e1fbbd61f@default(draft) add b4_
  | |
  | x  24:be27500cfc76@default(draft) add b3_
  | |
  | x  23:b54f77dc5831@default(draft) add b2_
  | |
  | x  22:0e1eba27e9aa@default(draft) add b1_
  |/
  | o  21:dcf786e878fd@default(draft) add c1second
  |/
  | *  20:507d52d715f6@default(draft) add c2prime
  | |
  | x  19:c995cb124ddc@default(draft) add c1prime
  |/
  | *  18:d096a2437fd0@default(draft) add c4_
  | |
  | *  17:cde95c6cba7a@default(draft) add c3_
  | |
  | x  16:e0d9f7a099fe@default(draft) add c2_
  | |
  | x  15:43b7c338b1f8@default(draft) add c1_
  |/
  o  14:739f18ac1d03@default(draft) add _d
  |
  ~

Solve the full second stack and only part of the first one
  $ echo "(desc(_d)::) - desc(c3_)"
  (desc(_d)::) - desc(c3_)
  $ hg evolve --rev "(desc(_d)::) - desc(c3_)"
  cannot solve instability of d096a2437fd0, skipping
  move:[20] add c2prime
  atop:[21] add c1second
  move:[25] add b4_
  atop:[27] add b3prime
  working directory is now at ea93190a9cd1

Cleanup
  $ hg evolve --rev "(desc(_d)::)"
  move:[17] add c3_
  atop:[28] add c2prime
  move:[18] add c4_
  atop:[30] add c3_
  working directory is now at 35e7b797ace5
  $ hg log -G -r "desc(_d)::"
  @  31:35e7b797ace5@default(draft) add c4_
  |
  o  30:0b9488394e89@default(draft) add c3_
  |
  | o  29:ea93190a9cd1@default(draft) add b4_
  | |
  o |  28:881b9c092e53@default(draft) add c2prime
  | |
  | o  27:b253ff5b65d1@default(draft) add b3prime
  | |
  | o  26:4acf61f11dfb@default(draft) add b1prime
  | |
  o |  21:dcf786e878fd@default(draft) add c1second
  |/
  o  14:739f18ac1d03@default(draft) add _d
  |
  ~

Test multiple revision with some un-evolvable because parent is splitted
------------------------------------------------------------------------

  $ hg up 'desc(c2prime)'
  0 files updated, 0 files merged, 2 files removed, 0 files unresolved
  $ mkcommit c3part1
  created new head
  $ hg prev
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
  [28] add c2prime
  $ mkcommit c3part2
  created new head
  $ hg prune -s 'desc(c3part1)' 'desc(c3_)'
  1 changesets pruned
  1 new orphan changesets
  $ hg prune -s 'desc(c3part2)' 'desc(c3_)'
  1 changesets pruned
  2 new content-divergent changesets
  $ hg up 'desc(b3prime)'
  2 files updated, 0 files merged, 3 files removed, 0 files unresolved
  $ hg amend -m 'b3second'
  1 new orphan changesets
  $ hg evolve --rev "orphan()"
  move:[29] add b4_
  atop:[34] b3second
  skipping 0b9488394e89: divergent rewriting. can't choose destination
  working directory is now at 31809a198477