phases: update the phase set as we go during retract boundary
Apparently iterating over the `changed_revs` dictionary is very expensive.
On mozilla-try-2019-02-18, a perf::unbundle call with a 10 000 changesets
bundle gives give use the following timing.
e57d4b868a3e: 4.6 seconds
ac1c75188440: 102.5 seconds
prev-changeset: 30.0 seconds
this-changeset: 4.6 seconds
So, the performance regression is gone.
Once again: thanks to marvelous Python!
Force revlog max inline value to be smaller than default
$ mkdir $TESTTMP/ext
$ cat << EOF > $TESTTMP/ext/small_inline.py
> from mercurial import revlog
> revlog._maxinline = 8
> EOF
$ cat << EOF >> $HGRCPATH
> [extensions]
> small_inline=$TESTTMP/ext/small_inline.py
> EOF
$ hg init repo
$ cd repo
Try on an empty repository
$ hg debug-revlog-stats
rev-count data-size inl type target
$ mkdir folder
$ touch a b folder/c folder/d
$ hg commit -Aqm 0
$ echo "text" > a
$ hg rm b
$ echo "longer string" > folder/d
$ hg commit -Aqm 1
Differences in data size observed with pure is due to different compression
algorithms
$ hg debug-revlog-stats
rev-count data-size inl type target
2 138 no changelog (no-pure !)
2 137 no changelog (pure !)
2 177 no manifest (no-pure !)
2 168 no manifest (pure !)
2 6 yes file a
1 0 yes file b
1 0 yes file folder/c
2 15 no file folder/d
Test 'changelog' command argument
$ hg debug-revlog-stats -c
rev-count data-size inl type target
2 138 no changelog (no-pure !)
2 137 no changelog (pure !)
Test 'manifest' command argument
$ hg debug-revlog-stats -m
rev-count data-size inl type target
2 177 no manifest (no-pure !)
2 168 no manifest (pure !)
Test 'file' command argument
$ hg debug-revlog-stats -f
rev-count data-size inl type target
2 6 yes file a
1 0 yes file b
1 0 yes file folder/c
2 15 no file folder/d
Test multiple command arguments
$ hg debug-revlog-stats -cm
rev-count data-size inl type target
2 138 no changelog (no-pure !)
2 137 no changelog (pure !)
2 177 no manifest (no-pure !)
2 168 no manifest (pure !)