##// END OF EJS Templates
rust: peek_mut optim for lazy ancestors...
rust: peek_mut optim for lazy ancestors This is one of the two optimizations that are also present in the Python code: replacing pairs of pop/push on the BinaryHeap by single updates, hence having it under the hood maintain its consistency (sift) only once. On Mozilla central, the measured gain (see details below) is around 7%. Creating the PeekMut object by calling peek_mut() right away instead of peek() first is less efficient (gain is only 4%, stats not included). Our interpretation is that its creation has a cost which is vasted in the cases where it ends by droping the value (Peekmut::pop() just does self.heap.pop() anyway). On the other hand, the immutable peek() is very fast: it's just taking a reference in the underlying vector. The Python version still has another optimization: if parent(current) == current-1, then the heap doesn't need to maintain its consistency, since we already know that it's bigger than all the others in the heap. Rust's BinaryHeap doesn't allow us to mutate its biggest element with no housekeeping, but we tried it anyway, with a copy of the BinaryHeap implementation with a dedicaded added method: it's not worth the technical debt in our opinion (we measured only a further 1.6% improvement). One possible explanation would be that the sift is really fast anyway in that case, whereas it's not in the case of Python, because it's at least partly done in slow Python code. Still it's possible that replacing BinaryHeap by something more dedicated to discrete ordered types could be faster. Measurements on mozilla-central: Three runs of 'hg perfancestors' on the parent changeset: Moyenne des médianes: 0.100587 ! wall 0.100062 comb 0.100000 user 0.100000 sys 0.000000 (best of 98) ! wall 0.135804 comb 0.130000 user 0.130000 sys 0.000000 (max of 98) ! wall 0.102864 comb 0.102755 user 0.099286 sys 0.003469 (avg of 98) ! wall 0.101486 comb 0.110000 user 0.110000 sys 0.000000 (median of 98) ! wall 0.096804 comb 0.090000 user 0.090000 sys 0.000000 (best of 100) ! wall 0.132235 comb 0.130000 user 0.120000 sys 0.010000 (max of 100) ! wall 0.100258 comb 0.100300 user 0.096000 sys 0.004300 (avg of 100) ! wall 0.098384 comb 0.100000 user 0.100000 sys 0.000000 (median of 100) ! wall 0.099925 comb 0.100000 user 0.100000 sys 0.000000 (best of 98) ! wall 0.133518 comb 0.140000 user 0.130000 sys 0.010000 (max of 98) ! wall 0.102381 comb 0.102449 user 0.098265 sys 0.004184 (avg of 98) ! wall 0.101891 comb 0.090000 user 0.090000 sys 0.000000 (median of 98) Mean of the medians: 0.100587 On the present changeset: ! wall 0.091344 comb 0.090000 user 0.090000 sys 0.000000 (best of 100) ! wall 0.122728 comb 0.120000 user 0.110000 sys 0.010000 (max of 100) ! wall 0.093268 comb 0.093300 user 0.089300 sys 0.004000 (avg of 100) ! wall 0.092567 comb 0.100000 user 0.090000 sys 0.010000 (median of 100) ! wall 0.093294 comb 0.080000 user 0.080000 sys 0.000000 (best of 100) ! wall 0.144887 comb 0.150000 user 0.140000 sys 0.010000 (max of 100) ! wall 0.097708 comb 0.097700 user 0.093400 sys 0.004300 (avg of 100) ! wall 0.094980 comb 0.100000 user 0.090000 sys 0.010000 (median of 100) ! wall 0.091262 comb 0.090000 user 0.080000 sys 0.010000 (best of 100) ! wall 0.123772 comb 0.130000 user 0.120000 sys 0.010000 (max of 100) ! wall 0.093188 comb 0.093200 user 0.089300 sys 0.003900 (avg of 100) ! wall 0.092364 comb 0.100000 user 0.090000 sys 0.010000 (median of 100) Mean of the medians is 0.0933 Differential Revision: https://phab.mercurial-scm.org/D5358

File last commit:

r38484:2c2e8246 @54 default
r40847:e13ab4ac default
Show More
darcs.py
224 lines | 8.3 KiB | text/x-python | PythonLexer
# darcs.py - darcs support for the convert extension
#
# Copyright 2007-2009 Matt Mackall <mpm@selenic.com> and others
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
from __future__ import absolute_import
import errno
import os
import re
import shutil
from mercurial.i18n import _
from mercurial import (
error,
pycompat,
util,
)
from mercurial.utils import dateutil
from . import common
NoRepo = common.NoRepo
# The naming drift of ElementTree is fun!
try:
import xml.etree.cElementTree.ElementTree as ElementTree
import xml.etree.cElementTree.XMLParser as XMLParser
except ImportError:
try:
import xml.etree.ElementTree.ElementTree as ElementTree
import xml.etree.ElementTree.XMLParser as XMLParser
except ImportError:
try:
import elementtree.cElementTree.ElementTree as ElementTree
import elementtree.cElementTree.XMLParser as XMLParser
except ImportError:
try:
import elementtree.ElementTree.ElementTree as ElementTree
import elementtree.ElementTree.XMLParser as XMLParser
except ImportError:
pass
class darcs_source(common.converter_source, common.commandline):
def __init__(self, ui, repotype, path, revs=None):
common.converter_source.__init__(self, ui, repotype, path, revs=revs)
common.commandline.__init__(self, ui, 'darcs')
# check for _darcs, ElementTree so that we can easily skip
# test-convert-darcs if ElementTree is not around
if not os.path.exists(os.path.join(path, '_darcs')):
raise NoRepo(_("%s does not look like a darcs repository") % path)
common.checktool('darcs')
version = self.run0('--version').splitlines()[0].strip()
if version < '2.1':
raise error.Abort(_('darcs version 2.1 or newer needed (found %r)')
% version)
if "ElementTree" not in globals():
raise error.Abort(_("Python ElementTree module is not available"))
self.path = os.path.realpath(path)
self.lastrev = None
self.changes = {}
self.parents = {}
self.tags = {}
# Check darcs repository format
format = self.format()
if format:
if format in ('darcs-1.0', 'hashed'):
raise NoRepo(_("%s repository format is unsupported, "
"please upgrade") % format)
else:
self.ui.warn(_('failed to detect repository format!'))
def before(self):
self.tmppath = pycompat.mkdtemp(
prefix='convert-' + os.path.basename(self.path) + '-')
output, status = self.run('init', repodir=self.tmppath)
self.checkexit(status)
tree = self.xml('changes', xml_output=True, summary=True,
repodir=self.path)
tagname = None
child = None
for elt in tree.findall('patch'):
node = elt.get('hash')
name = elt.findtext('name', '')
if name.startswith('TAG '):
tagname = name[4:].strip()
elif tagname is not None:
self.tags[tagname] = node
tagname = None
self.changes[node] = elt
self.parents[child] = [node]
child = node
self.parents[child] = []
def after(self):
self.ui.debug('cleaning up %s\n' % self.tmppath)
shutil.rmtree(self.tmppath, ignore_errors=True)
def recode(self, s, encoding=None):
if isinstance(s, pycompat.unicode):
# XMLParser returns unicode objects for anything it can't
# encode into ASCII. We convert them back to str to get
# recode's normal conversion behavior.
s = s.encode('latin-1')
return super(darcs_source, self).recode(s, encoding)
def xml(self, cmd, **kwargs):
# NOTE: darcs is currently encoding agnostic and will print
# patch metadata byte-for-byte, even in the XML changelog.
etree = ElementTree()
# While we are decoding the XML as latin-1 to be as liberal as
# possible, etree will still raise an exception if any
# non-printable characters are in the XML changelog.
parser = XMLParser(encoding='latin-1')
p = self._run(cmd, **kwargs)
etree.parse(p.stdout, parser=parser)
p.wait()
self.checkexit(p.returncode)
return etree.getroot()
def format(self):
output, status = self.run('show', 'repo', repodir=self.path)
self.checkexit(status)
m = re.search(r'^\s*Format:\s*(.*)$', output, re.MULTILINE)
if not m:
return None
return ','.join(sorted(f.strip() for f in m.group(1).split(',')))
def manifest(self):
man = []
output, status = self.run('show', 'files', no_directories=True,
repodir=self.tmppath)
self.checkexit(status)
for line in output.split('\n'):
path = line[2:]
if path:
man.append(path)
return man
def getheads(self):
return self.parents[None]
def getcommit(self, rev):
elt = self.changes[rev]
dateformat = '%a %b %d %H:%M:%S %Z %Y'
date = dateutil.strdate(elt.get('local_date'), dateformat)
desc = elt.findtext('name') + '\n' + elt.findtext('comment', '')
# etree can return unicode objects for name, comment, and author,
# so recode() is used to ensure str objects are emitted.
newdateformat = '%Y-%m-%d %H:%M:%S %1%2'
return common.commit(author=self.recode(elt.get('author')),
date=dateutil.datestr(date, newdateformat),
desc=self.recode(desc).strip(),
parents=self.parents[rev])
def pull(self, rev):
output, status = self.run('pull', self.path, all=True,
match='hash %s' % rev,
no_test=True, no_posthook=True,
external_merge='/bin/false',
repodir=self.tmppath)
if status:
if output.find('We have conflicts in') == -1:
self.checkexit(status, output)
output, status = self.run('revert', all=True, repodir=self.tmppath)
self.checkexit(status, output)
def getchanges(self, rev, full):
if full:
raise error.Abort(_("convert from darcs does not support --full"))
copies = {}
changes = []
man = None
for elt in self.changes[rev].find('summary').getchildren():
if elt.tag in ('add_directory', 'remove_directory'):
continue
if elt.tag == 'move':
if man is None:
man = self.manifest()
source, dest = elt.get('from'), elt.get('to')
if source in man:
# File move
changes.append((source, rev))
changes.append((dest, rev))
copies[dest] = source
else:
# Directory move, deduce file moves from manifest
source = source + '/'
for f in man:
if not f.startswith(source):
continue
fdest = dest + '/' + f[len(source):]
changes.append((f, rev))
changes.append((fdest, rev))
copies[fdest] = f
else:
changes.append((elt.text.strip(), rev))
self.pull(rev)
self.lastrev = rev
return sorted(changes), copies, set()
def getfile(self, name, rev):
if rev != self.lastrev:
raise error.Abort(_('internal calling inconsistency'))
path = os.path.join(self.tmppath, name)
try:
data = util.readfile(path)
mode = os.lstat(path).st_mode
except IOError as inst:
if inst.errno == errno.ENOENT:
return None, None
raise
mode = (mode & 0o111) and 'x' or ''
return data, mode
def gettags(self):
return self.tags