cvs.py
367 lines
| 14.2 KiB
| text/x-python
|
PythonLexer
|
r8250 | # 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, incorporated herein by reference. | ||||
|
r4536 | |||
|
r7444 | import os, locale, re, socket, errno | ||
|
r5539 | from cStringIO import StringIO | ||
|
r4536 | from mercurial import util | ||
|
r6690 | from mercurial.i18n import _ | ||
|
r4536 | |||
|
r5497 | from common import NoRepo, commit, converter_source, checktool | ||
|
r6690 | import cvsps | ||
|
r4536 | |||
class convert_cvs(converter_source): | ||||
|
r4760 | def __init__(self, ui, path, rev=None): | ||
|
r4807 | super(convert_cvs, self).__init__(ui, path, rev=rev) | ||
|
r4536 | cvs = os.path.join(path, "CVS") | ||
if not os.path.exists(cvs): | ||||
|
r5521 | raise NoRepo("%s does not look like a CVS checkout" % path) | ||
|
r4536 | |||
|
r6690 | checktool('cvs') | ||
|
r7101 | self.cmd = ui.config('convert', 'cvsps', 'builtin') | ||
|
r6318 | cvspsexe = self.cmd.split(None, 1)[0] | ||
|
r6690 | self.builtin = cvspsexe == 'builtin' | ||
if not self.builtin: | ||||
checktool(cvspsexe) | ||||
|
r5497 | |||
|
r8048 | self.changeset = None | ||
|
r4536 | self.files = {} | ||
self.tags = {} | ||||
self.lastbranch = {} | ||||
self.parent = {} | ||||
self.socket = None | ||||
self.cvsroot = file(os.path.join(cvs, "Root")).read()[:-1] | ||||
self.cvsrepo = file(os.path.join(cvs, "Repository")).read()[:-1] | ||||
self.encoding = locale.getpreferredencoding() | ||||
|
r6690 | |||
|
r4536 | self._connect() | ||
|
r8048 | def _parse(self): | ||
if self.changeset is not None: | ||||
|
r4536 | return | ||
|
r8048 | self.changeset = {} | ||
|
r4536 | |||
|
r4760 | maxrev = 0 | ||
|
r6318 | cmd = self.cmd | ||
|
r4760 | if self.rev: | ||
# TODO: handle tags | ||||
try: | ||||
# patchset number? | ||||
maxrev = int(self.rev) | ||||
except ValueError: | ||||
try: | ||||
# date | ||||
util.parsedate(self.rev, ['%Y/%m/%d %H:%M:%S']) | ||||
|
r5308 | cmd = '%s -d "1970/01/01 00:00:01" -d "%s"' % (cmd, self.rev) | ||
|
r4760 | except util.Abort: | ||
|
r6956 | raise util.Abort(_('revision %s is not a patchset number or date') % self.rev) | ||
|
r4760 | |||
|
r4536 | d = os.getcwd() | ||
try: | ||||
os.chdir(self.path) | ||||
id = None | ||||
state = 0 | ||||
|
r5920 | filerevids = {} | ||
|
r6690 | |||
if self.builtin: | ||||
# builtin cvsps code | ||||
|
r8048 | self.ui.status(_('using builtin cvsps\n')) | ||
|
r6690 | |||
|
r8169 | cache = 'update' | ||
if not self.ui.configbool('convert', 'cvsps.cache', True): | ||||
cache = None | ||||
db = cvsps.createlog(self.ui, cache=cache) | ||||
|
r8048 | 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)) | ||||
|
r6690 | |||
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) | ||||
date = util.datestr(cs.date) | ||||
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]) | ||||
|
r5920 | |||
|
r6690 | # 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 | ||||
else: | ||||
# external cvsps | ||||
for l in util.popen(cmd): | ||||
if state == 0: # header | ||||
if l.startswith("PatchSet"): | ||||
id = l[9:-2] | ||||
if maxrev and int(id) > maxrev: | ||||
# ignore everything | ||||
state = 3 | ||||
|
r7441 | elif l.startswith("Date:"): | ||
|
r6690 | date = util.parsedate(l[6:-1], ["%Y/%m/%d %H:%M:%S"]) | ||
date = util.datestr(date) | ||||
|
r7441 | elif l.startswith("Branch:"): | ||
|
r6690 | branch = l[8:-1] | ||
self.parent[id] = self.lastbranch.get(branch, 'bad') | ||||
self.lastbranch[branch] = id | ||||
|
r7441 | elif l.startswith("Ancestor branch:"): | ||
|
r6690 | ancestor = l[17:-1] | ||
# figure out the parent later | ||||
self.parent[id] = self.lastbranch[ancestor] | ||||
|
r7441 | elif l.startswith("Author:"): | ||
|
r6690 | author = self.recode(l[8:-1]) | ||
elif l.startswith("Tag:") or l.startswith("Tags:"): | ||||
t = l[l.index(':')+1:] | ||||
t = [ut.strip() for ut in t.split(',')] | ||||
if (len(t) > 1) or (t[0] and (t[0] != "(none)")): | ||||
self.tags.update(dict.fromkeys(t, id)) | ||||
elif l.startswith("Log:"): | ||||
# switch to gathering log | ||||
state = 1 | ||||
log = "" | ||||
elif state == 1: # log | ||||
if l == "Members: \n": | ||||
# switch to gathering members | ||||
files = {} | ||||
oldrevs = [] | ||||
log = self.recode(log[:-1]) | ||||
state = 2 | ||||
else: | ||||
# gather log | ||||
log += l | ||||
elif state == 2: # members | ||||
if l == "\n": # start of next entry | ||||
state = 0 | ||||
p = [self.parent[id]] | ||||
if id == "1": | ||||
p = [] | ||||
if branch == "HEAD": | ||||
branch = "" | ||||
if branch: | ||||
|
r7472 | latest = 0 | ||
|
r6690 | # the last changeset that contains a base | ||
# file is our parent | ||||
for r in oldrevs: | ||||
|
r7472 | latest = max(filerevids.get(r, 0), latest) | ||
|
r6690 | if latest: | ||
p = [latest] | ||||
|
r5920 | |||
|
r6690 | # add current commit to set | ||
c = commit(author=author, date=date, parents=p, | ||||
desc=log, branch=branch) | ||||
self.changeset[id] = c | ||||
self.files[id] = files | ||||
else: | ||||
colon = l.rfind(':') | ||||
file = l[1:colon] | ||||
rev = l[colon+1:-2] | ||||
oldrev, rev = rev.split("->") | ||||
files[file] = rev | ||||
# save some information for identifying branch points | ||||
oldrevs.append("%s:%s" % (oldrev, file)) | ||||
filerevids["%s:%s" % (rev, file)] = id | ||||
elif state == 3: | ||||
# swallow all input | ||||
continue | ||||
|
r4536 | |||
self.heads = self.lastbranch.values() | ||||
finally: | ||||
os.chdir(d) | ||||
def _connect(self): | ||||
root = self.cvsroot | ||||
conntype = None | ||||
user, host = None, None | ||||
cmd = ['cvs', 'server'] | ||||
|
r6956 | self.ui.status(_("connecting to %s\n") % root) | ||
|
r4536 | |||
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" | ||||
|
r5082 | if not port: | ||
port = 2401 | ||||
|
r4536 | else: | ||
|
r5082 | port = int(port) | ||
format0 = ":pserver:%s@%s:%s" % (user, serv, root) | ||||
format1 = ":pserver:%s@%s:%d%s" % (user, serv, port, root) | ||||
|
r4536 | |||
if not passw: | ||||
passw = "A" | ||||
|
r7442 | cvspass = os.path.expanduser("~/.cvspass") | ||
|
r7444 | try: | ||
|
r7442 | pf = open(cvspass) | ||
for line in pf.read().splitlines(): | ||||
part1, part2 = line.split(' ', 1) | ||||
if part1 == '/1': | ||||
# /1 :pserver:user@example.com:2401/cvsroot/foo Ah<Z | ||||
part1, part2 = part2.split(' ', 1) | ||||
format = format1 | ||||
else: | ||||
# :pserver:user@example.com:/cvsroot/foo Ah<Z | ||||
format = format0 | ||||
if part1 == format: | ||||
passw = part2 | ||||
break | ||||
pf.close() | ||||
|
r7444 | except IOError, inst: | ||
if inst.errno != errno.ENOENT: | ||||
if not getattr(inst, 'filename', None): | ||||
inst.filename = cvspass | ||||
raise | ||||
|
r4536 | |||
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": | ||||
|
r6956 | raise util.Abort(_("CVS pserver authentication failed")) | ||
|
r4536 | |||
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) | ||||
|
r5304 | # Do not take Windows path "c:\foo\bar" for a connection strings | ||
if os.path.isdir(root) or not m: | ||||
|
r4536 | conntype = "local" | ||
else: | ||||
conntype = "rsh" | ||||
user, host, root = m.group(1), m.group(2), m.group(3) | ||||
if conntype != "pserver": | ||||
if conntype == "rsh": | ||||
|
r5860 | rsh = os.environ.get("CVS_RSH") or "ssh" | ||
|
r4536 | if user: | ||
cmd = [rsh, '-l', user, host] + cmd | ||||
else: | ||||
cmd = [rsh, host] + cmd | ||||
|
r5303 | # popen2 does not support argument lists under Windows | ||
cmd = [util.shellquote(arg) for arg in cmd] | ||||
cmd = util.quotecommand(' '.join(cmd)) | ||||
|
r8339 | self.writep, self.readp = util.popen2(cmd) | ||
|
r4536 | |||
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"): | ||||
|
r6956 | raise util.Abort(_("server sucks")) | ||
|
r4536 | if "UseUnchanged" in r: | ||
self.writep.write("UseUnchanged\n") | ||||
self.writep.flush() | ||||
r = self.readp.readline() | ||||
def getheads(self): | ||||
|
r8048 | self._parse() | ||
|
r4536 | return self.heads | ||
def _getfile(self, name, rev): | ||||
|
r5539 | |||
def chunkedread(fp, count): | ||||
# file-objects returned by socked.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: | ||||
|
r6956 | raise util.Abort(_("%d bytes missing from remote file") % count) | ||
|
r5539 | count -= len(data) | ||
output.write(data) | ||||
return output.getvalue() | ||||
|
r4536 | if rev.endswith("(DEAD)"): | ||
raise IOError | ||||
args = ("-N -P -kk -r %s --" % rev).split() | ||||
|
r5305 | args.append(self.cvsrepo + '/' + name) | ||
|
r4536 | 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 = "" | ||||
while 1: | ||||
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]) | ||||
|
r5539 | data = chunkedread(self.readp, count) | ||
|
r4536 | elif line.startswith(" "): | ||
data += line[1:] | ||||
elif line.startswith("M "): | ||||
pass | ||||
elif line.startswith("Mbinary "): | ||||
count = int(self.readp.readline()[:-1]) | ||||
|
r5539 | data = chunkedread(self.readp, count) | ||
|
r4536 | else: | ||
if line == "ok\n": | ||||
return (data, "x" in mode and "x" or "") | ||||
elif line.startswith("E "): | ||||
|
r6956 | self.ui.warn(_("cvs server: %s\n") % line[2:]) | ||
|
r4536 | elif line.startswith("Remove"): | ||
|
r7874 | self.readp.readline() | ||
|
r4536 | else: | ||
|
r6956 | raise util.Abort(_("unknown CVS response: %s") % line) | ||
|
r4536 | |||
def getfile(self, file, rev): | ||||
|
r8048 | self._parse() | ||
|
r4536 | data, mode = self._getfile(file, rev) | ||
self.modecache[(file, rev)] = mode | ||||
return data | ||||
def getmode(self, file, rev): | ||||
return self.modecache[(file, rev)] | ||||
def getchanges(self, rev): | ||||
|
r8048 | self._parse() | ||
|
r4536 | self.modecache = {} | ||
|
r8209 | return sorted(self.files[rev].iteritems()), {} | ||
|
r4536 | |||
def getcommit(self, rev): | ||||
|
r8048 | self._parse() | ||
|
r4536 | return self.changeset[rev] | ||
def gettags(self): | ||||
|
r8048 | self._parse() | ||
|
r4536 | return self.tags | ||
|
r5381 | |||
def getchangedfiles(self, rev, i): | ||||
|
r8048 | self._parse() | ||
|
r8209 | return sorted(self.files[rev]) | ||