view tests/test-import-merge.t @ 32200:4d504e541d3d

rebase: use matcher to optimize manifestmerge

The old merge code would call manifestmerge and calculate the complete diff
between the source to the destination. In many cases, like rebase, the vast
majority of differences between the source and destination are irrelevant
because they are differences between the destination and the common ancestor
only, and therefore don't affect the merge. Since most actions are 'keep', all
the effort to compute them is wasted.

Instead, let's compute the difference between the source and the common ancestor
and only perform the diff of those files against the merge destination. When
using treemanifest, this lets us avoid loading almost the entire tree when
rebasing from a very old ancestor. This speeds up rebase of an old stack of 27
commits by 20x.

In mozilla-central, without treemanifest, when rebasing a commit from
default~100000 to default, this speeds up the manifestmerge step from 2.6s to
1.2s.  However, the additional diff adds an overhead to all manifestmerge calls,
especially for flat manifests. When rebasing a commit from default~1 to default
it appears to add 100ms in mozilla-central. While we could put this optimization
behind a flag, I think the fact that it makes merge O(number of changes being
applied) instead of O(number of changes between X and Y) justifies it.

  $ echo "[extensions]" >> $HGRCPATH
  $ echo "mq=" >> $HGRCPATH

  $ tipparents() {
  > hg parents --template "{rev}:{node|short} {desc|firstline}\n" -r tip
  > }

Test import and merge diffs

  $ hg init repo
  $ cd repo
  $ echo a > a
  $ hg ci -Am adda
  adding a
  $ echo a >> a
  $ hg ci -m changea
  $ echo c > c
  $ hg ci -Am addc
  adding c
  $ hg up 0
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ echo b > b
  $ hg ci -Am addb
  adding b
  created new head
  $ hg up 1
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ hg merge 3
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  (branch merge, don't forget to commit)
  $ hg ci -m merge
  $ hg export . > ../merge.diff
  $ grep -v '^merge$' ../merge.diff > ../merge.nomsg.diff
  $ cd ..
  $ hg clone -r2 repo repo2
  adding changesets
  adding manifests
  adding file changes
  added 3 changesets with 3 changes to 2 files
  updating to branch default
  2 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ cd repo2
  $ hg pull -r3 ../repo
  pulling from ../repo
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files (+1 heads)
  (run 'hg heads' to see heads, 'hg merge' to merge)

Test without --exact and diff.p1 == workingdir.p1

  $ hg up 1
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ cat > $TESTTMP/editor.sh <<EOF
  > env | grep HGEDITFORM
  > echo merge > \$1
  > EOF
  $ HGEDITOR="sh $TESTTMP/editor.sh" hg import --edit ../merge.nomsg.diff
  applying ../merge.nomsg.diff
  HGEDITFORM=import.normal.merge
  $ tipparents
  1:540395c44225 changea
  3:102a90ea7b4a addb
  $ hg strip --no-backup tip
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved

Test without --exact and diff.p1 != workingdir.p1

  $ hg up 2
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ hg import ../merge.diff
  applying ../merge.diff
  warning: import the patch as a normal revision
  (use --exact to import the patch as a merge)
  $ tipparents
  2:890ecaa90481 addc
  $ hg strip --no-backup tip
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved

Test with --exact

  $ hg import --exact ../merge.diff
  applying ../merge.diff
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ tipparents
  1:540395c44225 changea
  3:102a90ea7b4a addb
  $ hg strip --no-backup tip
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved

Test with --bypass and diff.p1 == workingdir.p1

  $ hg up 1
  0 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ hg import --bypass ../merge.diff
  applying ../merge.diff
  $ tipparents
  1:540395c44225 changea
  3:102a90ea7b4a addb
  $ hg strip --no-backup tip

Test with --bypass and diff.p1 != workingdir.p1

  $ hg up 2
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ hg import --bypass ../merge.diff
  applying ../merge.diff
  warning: import the patch as a normal revision
  (use --exact to import the patch as a merge)
  $ tipparents
  2:890ecaa90481 addc
  $ hg strip --no-backup tip

Test with --bypass and --exact

  $ hg import --bypass --exact ../merge.diff
  applying ../merge.diff
  $ tipparents
  1:540395c44225 changea
  3:102a90ea7b4a addb
  $ hg strip --no-backup tip

  $ cd ..

Test that --exact on a bad header doesn't corrupt the repo (issue3616)

  $ hg init repo3
  $ cd repo3
  $ echo a>a
  $ hg ci -Aqm0
  $ echo a>>a
  $ hg ci -m1
  $ echo a>>a
  $ hg ci -m2
  $ echo a>a
  $ echo b>>a
  $ echo a>>a
  $ hg ci -m3
  $ hg export 2 | head -7 > ../a.patch
  $ hg export tip > out
  >>> apatch = open("../a.patch", "ab")
  >>> apatch.write("".join(open("out").readlines()[7:]))

  $ cd ..
  $ hg clone -qr0 repo3 repo3-clone
  $ cd repo3-clone
  $ hg pull -qr1 ../repo3

  $ hg import --exact ../a.patch
  applying ../a.patch
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  patching file a
  Hunk #1 succeeded at 1 with fuzz 1 (offset -1 lines).
  transaction abort!
  rollback completed
  abort: patch is damaged or loses information
  [255]
  $ hg verify
  checking changesets
  checking manifests
  crosschecking files in changesets and manifests
  checking files
  1 files, 2 changesets, 2 total revisions