hgext/convert/darcs.py

match: convert O(n) to O(log n) in exactmatcher.visitchildrenset

When using narrow, during rebase this is called (at least) once per directory in
the set of files in the commit being rebased. Every time it's called, we did the
set arithmetic (now extracted and cached), which was probably pretty cheap but
not necessary to repeat each time, looped over every item in the matcher and
kept things that started with the directory we were querying.

With very large narrowspecs, and a commit that touched a file in a large number
of directories, this was slow. In a pathological repo, the rebase of a single
commit (that touched over 17k files, I believe in approximately as many
directories) with a narrowspec that had >32k entries took 8,246s of profiled
time, with 5,007s of that spent in visitchildrenset (transitively). With this
change, the time spent in visitchildrenset is less than 34s (which is where my
profile cut off). Most of the remaining time was network access due to our
custom remotefilelog-based setup not properly prefetching.

Differential Revision: https://phab.mercurial-scm.org/D10294

# darcs.py - darcs support for the convert extension
#
#  Copyright 2007-2009 Olivia Mackall <[e-mail]> 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, b'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, b'_darcs')):
            raise NoRepo(_(b"%s does not look like a darcs repository") % path)

        common.checktool(b'darcs')
        version = self.run0(b'--version').splitlines()[0].strip()
        if version < b'2.1':
            raise error.Abort(
                _(b'darcs version 2.1 or newer needed (found %r)') % version
            )

        if b"ElementTree" not in globals():
            raise error.Abort(_(b"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 (b'darcs-1.0', b'hashed'):
                raise NoRepo(
                    _(
                        b"%s repository format is unsupported, "
                        b"please upgrade"
                    )
                    % format
                )
        else:
            self.ui.warn(_(b'failed to detect repository format!'))

    def before(self):
        self.tmppath = pycompat.mkdtemp(
            prefix=b'convert-' + os.path.basename(self.path) + b'-'
        )
        output, status = self.run(b'init', repodir=self.tmppath)
        self.checkexit(status)

        tree = self.xml(
            b'changes', xml_output=True, summary=True, repodir=self.path
        )
        tagname = None
        child = None
        for elt in tree.findall(b'patch'):
            node = elt.get(b'hash')
            name = elt.findtext(b'name', b'')
            if name.startswith(b'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(b'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=b'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(b'show', b'repo', repodir=self.path)
        self.checkexit(status)
        m = re.search(r'^\s*Format:\s*(.*)$', output, re.MULTILINE)
        if not m:
            return None
        return b','.join(sorted(f.strip() for f in m.group(1).split(b',')))

    def manifest(self):
        man = []
        output, status = self.run(
            b'show', b'files', no_directories=True, repodir=self.tmppath
        )
        self.checkexit(status)
        for line in output.split(b'\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 = b'%a %b %d %H:%M:%S %Z %Y'
        date = dateutil.strdate(elt.get(b'local_date'), dateformat)
        desc = elt.findtext(b'name') + b'\n' + elt.findtext(b'comment', b'')
        # etree can return unicode objects for name, comment, and author,
        # so recode() is used to ensure str objects are emitted.
        newdateformat = b'%Y-%m-%d %H:%M:%S %1%2'
        return common.commit(
            author=self.recode(elt.get(b'author')),
            date=dateutil.datestr(date, newdateformat),
            desc=self.recode(desc).strip(),
            parents=self.parents[rev],
        )

    def pull(self, rev):
        output, status = self.run(
            b'pull',
            self.path,
            all=True,
            match=b'hash %s' % rev,
            no_test=True,
            no_posthook=True,
            external_merge=b'/bin/false',
            repodir=self.tmppath,
        )
        if status:
            if output.find(b'We have conflicts in') == -1:
                self.checkexit(status, output)
            output, status = self.run(b'revert', all=True, repodir=self.tmppath)
            self.checkexit(status, output)

    def getchanges(self, rev, full):
        if full:
            raise error.Abort(_(b"convert from darcs does not support --full"))
        copies = {}
        changes = []
        man = None
        for elt in self.changes[rev].find(b'summary').getchildren():
            if elt.tag in (b'add_directory', b'remove_directory'):
                continue
            if elt.tag == b'move':
                if man is None:
                    man = self.manifest()
                source, dest = elt.get(b'from'), elt.get(b'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 + b'/'
                    for f in man:
                        if not f.startswith(source):
                            continue
                        fdest = dest + b'/' + 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(_(b'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 b'x' or b''
        return data, mode

    def gettags(self):
        return self.tags