D1621: transaction: split changes['phases'] into sets for src and target phase

[joerg.sonnenberger (Joerg Sonnenberger)]

joerg.sonnenberger created this revision.
Herald added a subscriber: mercurial-devel.
Herald added a reviewer: hg-reviewers.

REVISION SUMMARY
  changes['phases'] used to be a dictionary mapping revisions to
  (old, new) tuples. The encoding is highly redundant and eats ~40MB
  for the test case in issue5691. Recognize that for new, only the three
  phases are valid and record the state using three sets. For old, keep
  three sets for the phases as well and encode unknown as lack of
  membership in the three sets. Provide a helper function to enumerate the
  changes.

REPOSITORY
  rHG Mercurial

REVISION DETAIL
  https://phab.mercurial-scm.org/D1621

AFFECTED FILES
  mercurial/localrepo.py
  mercurial/phases.py
  tests/testlib/ext-phase-report.py

CHANGE DETAILS

diff --git a/tests/testlib/ext-phase-report.py b/tests/testlib/ext-phase-report.py
--- a/tests/testlib/ext-phase-report.py
+++ b/tests/testlib/ext-phase-report.py
@@ -1,11 +1,12 @@
 # tiny extension to report phase changes during transaction
 
 from __future__ import absolute_import
+from mercurial import phases
 
 def reposetup(ui, repo):
 
     def reportphasemove(tr):
-        for rev, move in sorted(tr.changes['phases'].iteritems()):
+        for rev, move in sorted(phases.phasechanges(tr.changes['phases'])):
             if move[0] is None:
                 ui.write(('test-debug-phase: new rev %d:  x -> %d\n'
                           % (rev, move[1])))
diff --git a/mercurial/phases.py b/mercurial/phases.py
--- a/mercurial/phases.py
+++ b/mercurial/phases.py
@@ -186,17 +186,38 @@
         headsbyphase[phase].append(node)
     return headsbyphase
 
+def phasechanges(data):
+    offset = len(allphases)
+    for tgtphase in allphases:
+        for rev in data[tgtphase + offset]:
+            srcphase = None
+            for phase in allphases:
+                if rev in data[phase]:
+                    srcphase = phase
+                    break
+            yield (rev, (srcphase, tgtphase))
+
 def _trackphasechange(data, rev, old, new):
-    """add a phase move the <data> dictionnary
+    """add a phase move the <data> sets
 
     If data is None, nothing happens.
     """
     if data is None:
         return
-    existing = data.get(rev)
-    if existing is not None:
-        old = existing[0]
-    data[rev] = (old, new)
+    assert old in allphases or old is None
+    assert new in allphases
+    offset = len(allphases)
+
+    if rev not in data[new + offset]:
+        found = False
+        for phase in allphases:
+            if phase == new:
+                data[phase + offset].add(rev)
+            elif rev in data[phase + offset]:
+                data[phase + offset].discard(rev)
+                found = True
+        if not found and old is not None:
+            data[old].add(rev)
 
 class phasecache(object):
     def __init__(self, repo, phasedefaults, _load=True):
diff --git a/mercurial/localrepo.py b/mercurial/localrepo.py
--- a/mercurial/localrepo.py
+++ b/mercurial/localrepo.py
@@ -1242,7 +1242,8 @@
                               **pycompat.strkwargs(args))
             if hook.hashook(repo.ui, 'pretxnclose-phase'):
                 cl = repo.unfiltered().changelog
-                for rev, (old, new) in tr.changes['phases'].items():
+                phasechanges = phases.phasechanges(tr.changes['phases'])
+                for rev, (old, new) in phasechanges:
                     args = tr.hookargs.copy()
                     node = hex(cl.node(rev))
                     args.update(phases.preparehookargs(node, old, new))
@@ -1277,7 +1278,7 @@
                                      checkambigfiles=_cachedfiles)
         tr.changes['revs'] = xrange(0, 0)
         tr.changes['obsmarkers'] = set()
-        tr.changes['phases'] = {}
+        tr.changes['phases'] = [set() for i in range(7)]
         tr.changes['bookmarks'] = {}
 
         tr.hookargs['txnid'] = txnid
@@ -1306,7 +1307,7 @@
 
                 if hook.hashook(repo.ui, 'txnclose-phase'):
                     cl = repo.unfiltered().changelog
-                    phasemv = sorted(tr.changes['phases'].items())
+                    phasemv = sorted(phases.phasechanges(tr.changes['phases']))
                     for rev, (old, new) in phasemv:
                         args = tr.hookargs.copy()
                         node = hex(cl.node(rev))



To: joerg.sonnenberger, #hg-reviewers
Cc: mercurial-devel