##// END OF EJS Templates
phase: improve retractboundary perf...
phase: improve retractboundary perf The existing retractboundary implementation computed the new boundary by walking all descendants of all existing roots and computing the new roots. This is O(commits since first root), which on long repos can be hundreds of thousands of commits. The new algorithm only updates roots that are greater than the new root locations. For common operations like commit on a repo with the earliest root several hundred thousand commits ago, this makes retractboundary go from 1 second to 0.008 seconds. I tested it by running the test suite with both implementations and checking that the root results were always the identical. There was some discussion on IRC about the safety of this (i.e. what if the new nodes are already part of the phase, etc). I've looked into it and believe this patch is safe: 1) The old existing code already filters the input nodes to only contain nodes that require retracting (i.e. we only make node X a new root if the old phase is less than the target phase), so there's no chance of us adding a unnecessary root to the phase (unless the input root is made unnecessary by another root in the same input, but see point #3). 2) Another way of thinking about this is: the only way the new algorithm would be different from the old algorithm is if it added a root that is a descendant of an old root (since the old algorithm would've caught this in the big "roots(%ln::)". At the beginning of the function, when we filter out roots that already meet the phase criteria, the *definition* of meeting the phase criteria is "not being a descendant of an existing root". Therefore, by definition none of the new roots we are processing are descendants of an existing root. 3) If two nodes are passed in as input, and one node is an ancestor of the other (and therefore the later node should not be a root), this is still caught by the 'roots(%ln::)' revset. So there's no chance of an extra root being introduced that way either.

File last commit:

r26779:aaa33ec3 default
r26909:e3611881 default
Show More
cvs.py
280 lines | 10.3 KiB | text/x-python | PythonLexer
# cvs.py: CVS conversion code inspired by hg-cvs-import and git-cvsimport
#
# Copyright 2005-2009 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.
import os, re, socket, errno
from cStringIO import StringIO
from mercurial import encoding, util, error
from mercurial.i18n import _
from common import NoRepo, commit, converter_source, checktool
from common import makedatetimestamp
import cvsps
class convert_cvs(converter_source):
def __init__(self, ui, path, revs=None):
super(convert_cvs, self).__init__(ui, path, revs=revs)
cvs = os.path.join(path, "CVS")
if not os.path.exists(cvs):
raise NoRepo(_("%s does not look like a CVS checkout") % path)
checktool('cvs')
self.changeset = None
self.files = {}
self.tags = {}
self.lastbranch = {}
self.socket = None
self.cvsroot = open(os.path.join(cvs, "Root")).read()[:-1]
self.cvsrepo = open(os.path.join(cvs, "Repository")).read()[:-1]
self.encoding = encoding.encoding
self._connect()
def _parse(self):
if self.changeset is not None:
return
self.changeset = {}
maxrev = 0
if self.revs:
if len(self.revs) > 1:
raise error.Abort(_('cvs source does not support specifying '
'multiple revs'))
# TODO: handle tags
try:
# patchset number?
maxrev = int(self.revs[0])
except ValueError:
raise error.Abort(_('revision %s is not a patchset number')
% self.revs[0])
d = os.getcwd()
try:
os.chdir(self.path)
id = None
cache = 'update'
if not self.ui.configbool('convert', 'cvsps.cache', True):
cache = None
db = cvsps.createlog(self.ui, cache=cache)
db = cvsps.createchangeset(self.ui, db,
fuzz=int(self.ui.config('convert', 'cvsps.fuzz', 60)),
mergeto=self.ui.config('convert', 'cvsps.mergeto', None),
mergefrom=self.ui.config('convert', 'cvsps.mergefrom', None))
for cs in db:
if maxrev and cs.id > maxrev:
break
id = str(cs.id)
cs.author = self.recode(cs.author)
self.lastbranch[cs.branch] = id
cs.comment = self.recode(cs.comment)
if self.ui.configbool('convert', 'localtimezone'):
cs.date = makedatetimestamp(cs.date[0])
date = util.datestr(cs.date, '%Y-%m-%d %H:%M:%S %1%2')
self.tags.update(dict.fromkeys(cs.tags, id))
files = {}
for f in cs.entries:
files[f.file] = "%s%s" % ('.'.join([str(x)
for x in f.revision]),
['', '(DEAD)'][f.dead])
# add current commit to set
c = commit(author=cs.author, date=date,
parents=[str(p.id) for p in cs.parents],
desc=cs.comment, branch=cs.branch or '')
self.changeset[id] = c
self.files[id] = files
self.heads = self.lastbranch.values()
finally:
os.chdir(d)
def _connect(self):
root = self.cvsroot
conntype = None
user, host = None, None
cmd = ['cvs', 'server']
self.ui.status(_("connecting to %s\n") % root)
if root.startswith(":pserver:"):
root = root[9:]
m = re.match(r'(?:(.*?)(?::(.*?))?@)?([^:\/]*)(?::(\d*))?(.*)',
root)
if m:
conntype = "pserver"
user, passw, serv, port, root = m.groups()
if not user:
user = "anonymous"
if not port:
port = 2401
else:
port = int(port)
format0 = ":pserver:%s@%s:%s" % (user, serv, root)
format1 = ":pserver:%s@%s:%d%s" % (user, serv, port, root)
if not passw:
passw = "A"
cvspass = os.path.expanduser("~/.cvspass")
try:
pf = open(cvspass)
for line in pf.read().splitlines():
part1, part2 = line.split(' ', 1)
# /1 :pserver:user@example.com:2401/cvsroot/foo
# Ah<Z
if part1 == '/1':
part1, part2 = part2.split(' ', 1)
format = format1
# :pserver:user@example.com:/cvsroot/foo Ah<Z
else:
format = format0
if part1 == format:
passw = part2
break
pf.close()
except IOError as inst:
if inst.errno != errno.ENOENT:
if not getattr(inst, 'filename', None):
inst.filename = cvspass
raise
sck = socket.socket()
sck.connect((serv, port))
sck.send("\n".join(["BEGIN AUTH REQUEST", root, user, passw,
"END AUTH REQUEST", ""]))
if sck.recv(128) != "I LOVE YOU\n":
raise error.Abort(_("CVS pserver authentication failed"))
self.writep = self.readp = sck.makefile('r+')
if not conntype and root.startswith(":local:"):
conntype = "local"
root = root[7:]
if not conntype:
# :ext:user@host/home/user/path/to/cvsroot
if root.startswith(":ext:"):
root = root[5:]
m = re.match(r'(?:([^@:/]+)@)?([^:/]+):?(.*)', root)
# Do not take Windows path "c:\foo\bar" for a connection strings
if os.path.isdir(root) or not m:
conntype = "local"
else:
conntype = "rsh"
user, host, root = m.group(1), m.group(2), m.group(3)
if conntype != "pserver":
if conntype == "rsh":
rsh = os.environ.get("CVS_RSH") or "ssh"
if user:
cmd = [rsh, '-l', user, host] + cmd
else:
cmd = [rsh, host] + cmd
# popen2 does not support argument lists under Windows
cmd = [util.shellquote(arg) for arg in cmd]
cmd = util.quotecommand(' '.join(cmd))
self.writep, self.readp = util.popen2(cmd)
self.realroot = root
self.writep.write("Root %s\n" % root)
self.writep.write("Valid-responses ok error Valid-requests Mode"
" M Mbinary E Checked-in Created Updated"
" Merged Removed\n")
self.writep.write("valid-requests\n")
self.writep.flush()
r = self.readp.readline()
if not r.startswith("Valid-requests"):
raise error.Abort(_('unexpected response from CVS server '
'(expected "Valid-requests", but got %r)')
% r)
if "UseUnchanged" in r:
self.writep.write("UseUnchanged\n")
self.writep.flush()
r = self.readp.readline()
def getheads(self):
self._parse()
return self.heads
def getfile(self, name, rev):
def chunkedread(fp, count):
# file-objects returned by socket.makefile() do not handle
# large read() requests very well.
chunksize = 65536
output = StringIO()
while count > 0:
data = fp.read(min(count, chunksize))
if not data:
raise error.Abort(_("%d bytes missing from remote file")
% count)
count -= len(data)
output.write(data)
return output.getvalue()
self._parse()
if rev.endswith("(DEAD)"):
return None, None
args = ("-N -P -kk -r %s --" % rev).split()
args.append(self.cvsrepo + '/' + name)
for x in args:
self.writep.write("Argument %s\n" % x)
self.writep.write("Directory .\n%s\nco\n" % self.realroot)
self.writep.flush()
data = ""
mode = None
while True:
line = self.readp.readline()
if line.startswith("Created ") or line.startswith("Updated "):
self.readp.readline() # path
self.readp.readline() # entries
mode = self.readp.readline()[:-1]
count = int(self.readp.readline()[:-1])
data = chunkedread(self.readp, count)
elif line.startswith(" "):
data += line[1:]
elif line.startswith("M "):
pass
elif line.startswith("Mbinary "):
count = int(self.readp.readline()[:-1])
data = chunkedread(self.readp, count)
else:
if line == "ok\n":
if mode is None:
raise error.Abort(_('malformed response from CVS'))
return (data, "x" in mode and "x" or "")
elif line.startswith("E "):
self.ui.warn(_("cvs server: %s\n") % line[2:])
elif line.startswith("Remove"):
self.readp.readline()
else:
raise error.Abort(_("unknown CVS response: %s") % line)
def getchanges(self, rev, full):
if full:
raise error.Abort(_("convert from cvs does not support --full"))
self._parse()
return sorted(self.files[rev].iteritems()), {}, set()
def getcommit(self, rev):
self._parse()
return self.changeset[rev]
def gettags(self):
self._parse()
return self.tags
def getchangedfiles(self, rev, i):
self._parse()
return sorted(self.files[rev])