branchmap.py
284 lines
| 10.6 KiB
| text/x-python
|
PythonLexer
/ mercurial / branchmap.py
|
r18116 | # branchmap.py - logic to computes, maintain and stores branchmap for local repo | ||
# | ||||
# Copyright 2005-2007 Matt Mackall <mpm@selenic.com> | ||||
# | ||||
# This software may be used and distributed according to the terms of the | ||||
# GNU General Public License version 2 or any later version. | ||||
|
r18117 | |||
|
r18118 | from node import bin, hex, nullid, nullrev | ||
|
r18117 | import encoding | ||
|
r20032 | import util | ||
|
r21031 | import time | ||
|
r18117 | |||
|
r18185 | def _filename(repo): | ||
|
r18187 | """name of a branchcache file for a given repo or repoview""" | ||
|
r20185 | filename = "cache/branch2" | ||
|
r18187 | if repo.filtername: | ||
filename = '%s-%s' % (filename, repo.filtername) | ||||
return filename | ||||
|
r18185 | |||
|
r18118 | def read(repo): | ||
try: | ||||
|
r18185 | f = repo.opener(_filename(repo)) | ||
|
r18118 | lines = f.read().split('\n') | ||
f.close() | ||||
except (IOError, OSError): | ||||
|
r18212 | return None | ||
|
r18118 | |||
try: | ||||
|
r18184 | cachekey = lines.pop(0).split(" ", 2) | ||
last, lrev = cachekey[:2] | ||||
|
r18118 | last, lrev = bin(last), int(lrev) | ||
|
r18184 | filteredhash = None | ||
if len(cachekey) > 2: | ||||
filteredhash = bin(cachekey[2]) | ||||
partial = branchcache(tipnode=last, tiprev=lrev, | ||||
filteredhash=filteredhash) | ||||
|
r18132 | if not partial.validfor(repo): | ||
|
r18118 | # invalidate the cache | ||
|
r18166 | raise ValueError('tip differs') | ||
|
r18118 | for l in lines: | ||
if not l: | ||||
continue | ||||
|
r20185 | node, state, label = l.split(" ", 2) | ||
if state not in 'oc': | ||||
raise ValueError('invalid branch state') | ||||
|
r18118 | label = encoding.tolocal(label.strip()) | ||
if not node in repo: | ||||
|
r18166 | raise ValueError('node %s does not exist' % node) | ||
|
r20185 | node = bin(node) | ||
partial.setdefault(label, []).append(node) | ||||
if state == 'c': | ||||
partial._closednodes.add(node) | ||||
|
r18118 | except KeyboardInterrupt: | ||
raise | ||||
except Exception, inst: | ||||
if repo.ui.debugflag: | ||||
|
r18188 | msg = 'invalid branchheads cache' | ||
if repo.filtername is not None: | ||||
msg += ' (%s)' % repo.filtername | ||||
msg += ': %s\n' | ||||
|
r21789 | repo.ui.debug(msg % inst) | ||
|
r18212 | partial = None | ||
|
r18126 | return partial | ||
|
r18118 | |||
|
r20032 | ### Nearest subset relation | ||
# Nearest subset of filter X is a filter Y so that: | ||||
# * Y is included in X, | ||||
# * X - Y is as small as possible. | ||||
# This create and ordering used for branchmap purpose. | ||||
# the ordering may be partial | ||||
subsettable = {None: 'visible', | ||||
'visible': 'served', | ||||
'served': 'immutable', | ||||
'immutable': 'base'} | ||||
|
r18121 | def updatecache(repo): | ||
cl = repo.changelog | ||||
|
r18189 | filtername = repo.filtername | ||
partial = repo._branchcaches.get(filtername) | ||||
|
r18121 | |||
|
r18234 | revs = [] | ||
|
r18132 | if partial is None or not partial.validfor(repo): | ||
|
r18126 | partial = read(repo) | ||
|
r18212 | if partial is None: | ||
|
r20032 | subsetname = subsettable.get(filtername) | ||
|
r18234 | if subsetname is None: | ||
partial = branchcache() | ||||
else: | ||||
subset = repo.filtered(subsetname) | ||||
partial = subset.branchmap().copy() | ||||
extrarevs = subset.changelog.filteredrevs - cl.filteredrevs | ||||
revs.extend(r for r in extrarevs if r <= partial.tiprev) | ||||
revs.extend(cl.revs(start=partial.tiprev + 1)) | ||||
|
r18218 | if revs: | ||
|
r18305 | partial.update(repo, revs) | ||
|
r18128 | partial.write(repo) | ||
|
r18451 | assert partial.validfor(repo), filtername | ||
|
r18189 | repo._branchcaches[repo.filtername] = partial | ||
|
r18124 | |||
class branchcache(dict): | ||||
|
r20181 | """A dict like object that hold branches heads cache. | ||
This cache is used to avoid costly computations to determine all the | ||||
branch heads of a repo. | ||||
The cache is serialized on disk in the following format: | ||||
<tip hex node> <tip rev number> [optional filtered repo hex hash] | ||||
|
r20185 | <branch head hex node> <open/closed state> <branch name> | ||
<branch head hex node> <open/closed state> <branch name> | ||||
|
r20181 | ... | ||
The first line is used to check if the cache is still valid. If the | ||||
branch cache is for a filtered repo view, an optional third hash is | ||||
included that hashes the hashes of all filtered revisions. | ||||
|
r20185 | |||
The open/closed state is represented by a single letter 'o' or 'c'. | ||||
This field can be used to avoid changelog reads when determining if a | ||||
branch head closes a branch or not. | ||||
|
r20181 | """ | ||
|
r18124 | |||
|
r18168 | def __init__(self, entries=(), tipnode=nullid, tiprev=nullrev, | ||
|
r20185 | filteredhash=None, closednodes=None): | ||
|
r18125 | super(branchcache, self).__init__(entries) | ||
self.tipnode = tipnode | ||||
|
r18126 | self.tiprev = tiprev | ||
|
r18168 | self.filteredhash = filteredhash | ||
|
r20185 | # closednodes is a set of nodes that close their branch. If the branch | ||
# cache has been updated, it may contain nodes that are no longer | ||||
# heads. | ||||
if closednodes is None: | ||||
self._closednodes = set() | ||||
else: | ||||
self._closednodes = closednodes | ||||
|
r18168 | |||
def _hashfiltered(self, repo): | ||||
"""build hash of revision filtered in the current cache | ||||
|
r18644 | Tracking tipnode and tiprev is not enough to ensure validity of the | ||
|
r18168 | cache as they do not help to distinct cache that ignored various | ||
revision bellow tiprev. | ||||
To detect such difference, we build a cache of all ignored revisions. | ||||
""" | ||||
cl = repo.changelog | ||||
if not cl.filteredrevs: | ||||
return None | ||||
key = None | ||||
revs = sorted(r for r in cl.filteredrevs if r <= self.tiprev) | ||||
if revs: | ||||
s = util.sha1() | ||||
for rev in revs: | ||||
s.update('%s;' % rev) | ||||
key = s.digest() | ||||
return key | ||||
|
r18128 | |||
|
r18132 | def validfor(self, repo): | ||
|
r18644 | """Is the cache content valid regarding a repo | ||
|
r18132 | |||
|
r18644 | - False when cached tipnode is unknown or if we detect a strip. | ||
|
r18132 | - True when cache is up to date or a subset of current repo.""" | ||
try: | ||||
|
r18168 | return ((self.tipnode == repo.changelog.node(self.tiprev)) | ||
and (self.filteredhash == self._hashfiltered(repo))) | ||||
|
r18132 | except IndexError: | ||
return False | ||||
|
r20186 | def _branchtip(self, heads): | ||
|
r20245 | '''Return tuple with last open head in heads and false, | ||
otherwise return last closed head and true.''' | ||||
|
r20186 | tip = heads[-1] | ||
closed = True | ||||
for h in reversed(heads): | ||||
if h not in self._closednodes: | ||||
tip = h | ||||
closed = False | ||||
break | ||||
return tip, closed | ||||
def branchtip(self, branch): | ||||
|
r20245 | '''Return the tipmost open head on branch head, otherwise return the | ||
tipmost closed head on branch. | ||||
Raise KeyError for unknown branch.''' | ||||
|
r20186 | return self._branchtip(self[branch])[0] | ||
|
r20188 | def branchheads(self, branch, closed=False): | ||
heads = self[branch] | ||||
if not closed: | ||||
heads = [h for h in heads if h not in self._closednodes] | ||||
return heads | ||||
|
r20190 | def iterbranches(self): | ||
for bn, heads in self.iteritems(): | ||||
yield (bn, heads) + self._branchtip(heads) | ||||
|
r18232 | def copy(self): | ||
"""return an deep copy of the branchcache object""" | ||||
|
r20185 | return branchcache(self, self.tipnode, self.tiprev, self.filteredhash, | ||
self._closednodes) | ||||
|
r18132 | |||
|
r18128 | def write(self, repo): | ||
try: | ||||
|
r18185 | f = repo.opener(_filename(repo), "w", atomictemp=True) | ||
|
r18184 | cachekey = [hex(self.tipnode), str(self.tiprev)] | ||
if self.filteredhash is not None: | ||||
cachekey.append(hex(self.filteredhash)) | ||||
f.write(" ".join(cachekey) + '\n') | ||||
|
r21031 | nodecount = 0 | ||
|
r18357 | for label, nodes in sorted(self.iteritems()): | ||
|
r18128 | for node in nodes: | ||
|
r21031 | nodecount += 1 | ||
|
r20185 | if node in self._closednodes: | ||
state = 'c' | ||||
else: | ||||
state = 'o' | ||||
f.write("%s %s %s\n" % (hex(node), state, | ||||
encoding.fromlocal(label))) | ||||
|
r18128 | f.close() | ||
|
r21031 | repo.ui.log('branchcache', | ||
'wrote %s branch cache with %d labels and %d nodes\n', | ||||
repo.filtername, len(self), nodecount) | ||||
|
r21788 | except (IOError, OSError, util.Abort), inst: | ||
repo.ui.debug("couldn't write branch cache: %s\n" % inst) | ||||
|
r18214 | # Abort may be raise by read only opener | ||
|
r18128 | pass | ||
|
r18131 | |||
|
r18305 | def update(self, repo, revgen): | ||
|
r18131 | """Given a branchhead cache, self, that may have extra nodes or be | ||
|
r20263 | missing heads, and a generator of nodes that are strictly a superset of | ||
|
r18131 | heads missing, this function updates self to be correct. | ||
""" | ||||
|
r21031 | starttime = time.time() | ||
|
r18131 | cl = repo.changelog | ||
# collect new branch entries | ||||
newbranches = {} | ||||
|
r20185 | getbranchinfo = cl.branchinfo | ||
|
r18307 | for r in revgen: | ||
|
r20185 | branch, closesbranch = getbranchinfo(r) | ||
|
r20262 | newbranches.setdefault(branch, []).append(r) | ||
|
r20185 | if closesbranch: | ||
|
r20262 | self._closednodes.add(cl.node(r)) | ||
|
r18131 | # if older branchheads are reachable from new ones, they aren't | ||
# really branchheads. Note checking parents is insufficient: | ||||
# 1 (branch a) -> 2 (branch b) -> 3 (branch a) | ||||
|
r20262 | for branch, newheadrevs in newbranches.iteritems(): | ||
|
r18131 | bheads = self.setdefault(branch, []) | ||
|
r20264 | bheadset = set(cl.rev(node) for node in bheads) | ||
|
r18131 | |||
|
r20263 | # This have been tested True on all internal usage of this function. | ||
# run it again in case of doubt | ||||
# assert not (set(bheadrevs) & set(newheadrevs)) | ||||
newheadrevs.sort() | ||||
|
r20264 | bheadset.update(newheadrevs) | ||
|
r18131 | |||
# This loop prunes out two kinds of heads - heads that are | ||||
# superseded by a head in newheadrevs, and newheadrevs that are not | ||||
# heads because an existing head is their descendant. | ||||
|
r20263 | while newheadrevs: | ||
latest = newheadrevs.pop() | ||||
|
r20264 | if latest not in bheadset: | ||
|
r18131 | continue | ||
|
r20264 | ancestors = set(cl.ancestors([latest], min(bheadset))) | ||
bheadset -= ancestors | ||||
bheadrevs = sorted(bheadset) | ||||
|
r18131 | self[branch] = [cl.node(rev) for rev in bheadrevs] | ||
|
r20263 | tiprev = bheadrevs[-1] | ||
|
r18131 | if tiprev > self.tiprev: | ||
self.tipnode = cl.node(tiprev) | ||||
self.tiprev = tiprev | ||||
|
r19838 | if not self.validfor(repo): | ||
|
r18131 | # cache key are not valid anymore | ||
self.tipnode = nullid | ||||
self.tiprev = nullrev | ||||
for heads in self.values(): | ||||
tiprev = max(cl.rev(node) for node in heads) | ||||
if tiprev > self.tiprev: | ||||
self.tipnode = cl.node(tiprev) | ||||
self.tiprev = tiprev | ||||
|
r18168 | self.filteredhash = self._hashfiltered(repo) | ||
|
r21031 | |||
duration = time.time() - starttime | ||||
repo.ui.log('branchcache', 'updated %s branch cache in %.4f seconds\n', | ||||
repo.filtername, duration) | ||||