#!/usr/bin/env python

#

# This is a generalized framework for converting between SCM

# repository formats.

#

# To use, run:

#

# convert-repo <source> [<dest> [<mapfile>]]

#

# Currently accepted source formats: git, cvs

# Currently accepted destination formats: hg

#

# If destination isn't given, a new Mercurial repo named <src>-hg will

# be created. If <mapfile> isn't given, it will be put in a default

# location (<dest>/.hg/shamap by default)

#

# The <mapfile> is a simple text file that maps each source commit ID to

# the destination ID for that revision, like so:

#

# <source ID> <destination ID>

#

# If the file doesn't exist, it's automatically created.  It's updated

# on each commit copied, so convert-repo can be interrupted and can

# be run repeatedly to copy new commits.

import sys, os, zlib, sha, time, re, locale

os.environ["HGENCODING"] = "utf-8"

from mercurial import hg, ui, util, fancyopts

class Abort(Exception): pass

class NoRepo(Exception): pass

class commit:

    def __init__(self, **parts):

        for x in "author date desc parents".split():

            if not x in parts:

                abort("commit missing field %s\n" % x)

        self.__dict__.update(parts)

quiet = 0

def status(msg):

    if not quiet: sys.stdout.write(str(msg))

def warn(msg):

    sys.stderr.write(str(msg))

def abort(msg):

    raise Abort(msg)

def recode(s):

    try:

        return s.decode("utf-8").encode("utf-8")

    except:

        try:

            return s.decode("latin-1").encode("utf-8")

        except:

            return s.decode("utf-8", "replace").encode("utf-8")

# CVS conversion code inspired by hg-cvs-import and git-cvsimport

class convert_cvs:

    def __init__(self, path):

        self.path = path

        cvs = os.path.join(path, "CVS")

        if not os.path.exists(cvs):

            raise NoRepo("couldn't open CVS repo %s" % path)

        self.changeset = {}

        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()

        self._parse()

        self._connect()

    def _parse(self):

        if self.changeset:

            return

        d = os.getcwd()

        try:

            os.chdir(self.path)

            id = None

            state = 0

            for l in os.popen("cvsps -A"):

                if state == 0: # header

                    if l.startswith("PatchSet"):

                        id = l[9:-2]

                    elif l.startswith("Date"):

                        date = util.parsedate(l[6:-1], ["%Y/%m/%d %H:%M:%S"])

                        date = util.datestr(date)

                    elif l.startswith("Branch"):

                        branch = l[8:-1]

                        self.parent[id] = self.lastbranch.get(branch,'bad')

                        self.lastbranch[branch] = id

                    elif l.startswith("Ancestor branch"):

                        ancestor = l[17:-1]

                        self.parent[id] = self.lastbranch[ancestor]

                    elif l.startswith("Author"):

                        author = self.recode(l[8:-1])

                    elif l.startswith("Tag: "):

                        t = l[5:-1]

                        if t != "(none) ":

                            self.tags[t] = id

                    elif l.startswith("Log:"):

                        state = 1

                        log = ""

                elif state == 1: # log

                    if l == "Members: \n":

                        files = {}

                        log = self.recode(log[:-1])

                        if log.isspace():

                            log = "*** empty log message ***\n"

                        state = 2

                    else:

                        log += l

                elif state == 2:

                    if l == "\n": #

                        state = 0

                        p = [self.parent[id]]

                        if id == "1":

                            p = []

                        c = commit(author=author, date=date, parents=p,

                                   desc=log, branch=branch)

                        self.changeset[id] = c

                        self.files[id] = files

                    else:

                        file,rev = l[1:-2].rsplit(':',1)

                        rev = rev.split("->")[1]

                        files[file] = rev

            self.heads = self.lastbranch.values()

        finally:

            os.chdir(d)

    def _connect(self):

        root = self.cvsroot

        local = False

        user, host = None, None

        cmd = ['cvs', 'server']

        status("connecting to %s\n" % root)

        # only non-pserver for now

        if root.startswith(":pserver"):

            abort("can't handle pserver mode yet: %s\n" % root)

        if root.startswith(":local:"):

            local = True

            root = root[7:]

        else:

            # :ext:user@host/home/user/path/to/cvsroot

            if root.startswith(":ext:"):

                root = root[5:]

            m = re.match(r'(?:([^@:/]+)@)?([^:/]+):?(.*)', root)

            if not m:

                local = True

            else:

                local = False

                user, host, root = m.group(1), m.group(2), m.group(3)

        if not local:

            rsh = os.environ.get("CVS_RSH" or "rsh")

            if user:

                cmd = [rsh, '-l', user, host] + cmd

            else:

                cmd = [rsh, host] + cmd

        self.writep, self.readp = os.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"):

            abort("server sucks\n")

        if "UseUnchanged" in r:

            self.writep.write("UseUnchanged\n")

            self.writep.flush()

            r = self.readp.readline()

    def getheads(self):

        return self.heads

        if rev.endswith("(DEAD)"):

            raise IOError

        args = ("-N -P -kk -r %s --" % rev).split()

        args.append(os.path.join(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 = ""

        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])

                data = self.readp.read(count)

            elif line.startswith(" "):

                data += line[1:]

            elif line.startswith("M "):

                pass

            elif line.startswith("Mbinary "):

                count = int(self.readp.readline()[:-1])

                data = self.readp.read(count)

            else:

                if line == "ok\n":

                elif line.startswith("E "):

                    warn("cvs server: %s\n" % line[2:])

                elif line.startswith("Remove"):

                    l = self.readp.readline()

                    l = self.readp.readline()

                    if l != "ok\n":

                        abort("unknown CVS response: %s\n" % l)

                else:

                    abort("unknown CVS response: %s\n" % line)

    def getchanges(self, rev):

        files = self.files[rev]

        cl.sort()

        return cl

    def recode(self, text):

        return text.decode(self.encoding, "replace").encode("utf-8")

    def getcommit(self, rev):

        return self.changeset[rev]

    def gettags(self):

        return self.tags

class convert_git:

    def __init__(self, path):

        if os.path.isdir(path + "/.git"):

            path += "/.git"

        self.path = path

        if not os.path.exists(path + "/HEAD"):

            raise NoRepo("couldn't open GIT repo %s" % path)

    def getheads(self):

        fh = os.popen("GIT_DIR=%s git-rev-parse --verify HEAD" % self.path)

        return [fh.read()[:-1]]

    def catfile(self, rev, type):

        if rev == "0" * 40: raise IOError()

        fh = os.popen("GIT_DIR=%s git-cat-file %s %s 2>/dev/null" % (self.path, type, rev))

        return fh.read()

    def getfile(self, name, rev):

        return self.catfile(rev, "blob")

    def getchanges(self, version):

        fh = os.popen("GIT_DIR=%s git-diff-tree --root -m -r %s" % (self.path, version))

        changes = []

        for l in fh:

            if "\t" not in l: continue

            m, f = l[:-1].split("\t")

            m = m.split()

            h = m[3]

            p = (m[1] == "100755")

        return changes

    def getcommit(self, version):

        c = self.catfile(version, "commit") # read the commit hash

        end = c.find("\n\n")

        message = c[end+2:]

        message = recode(message)

        l = c[:end].splitlines()

        manifest = l[0].split()[1]

        parents = []

        for e in l[1:]:

            n,v = e.split(" ", 1)

            if n == "author":

                p = v.split()

                tm, tz = p[-2:]

                author = " ".join(p[:-2])

                if author[0] == "<": author = author[1:-1]

                author = recode(author)

            if n == "committer":

                p = v.split()

                tm, tz = p[-2:]

                committer = " ".join(p[:-2])

                if committer[0] == "<": committer = committer[1:-1]

                committer = recode(committer)

                message += "\ncommitter: %s\n" % committer

            if n == "parent": parents.append(v)

        tzs, tzh, tzm = tz[-5:-4] + "1", tz[-4:-2], tz[-2:]

        tz = -int(tzs) * (int(tzh) * 3600 + int(tzm))

        date = tm + " " + str(tz)

        c = commit(parents=parents, date=date, author=author, desc=message)

        return c

    def gettags(self):

        tags = {}

        for f in os.listdir(self.path + "/refs/tags"):

            try:

                h = file(self.path + "/refs/tags/" + f).read().strip()

                c = self.catfile(h, "tag") # read the commit hash

                h = c.splitlines()[0].split()[1]

                tags[f] = h

            except:

                pass

        return tags

class convert_mercurial:

    def __init__(self, path):

        self.path = path

        u = ui.ui()

        try:

            self.repo = hg.repository(u, path)

        except:

            raise NoRepo("could open hg repo %s" % path)

    def mapfile(self):

        return os.path.join(self.path, ".hg", "shamap")

    def getheads(self):

        h = self.repo.changelog.heads()

        return [ hg.hex(x) for x in h ]

    def putfile(self, f, e, data):

        self.repo.wfile(f, "w").write(data)

        if self.repo.dirstate.state(f) == '?':

            self.repo.dirstate.update([f], "a")

        util.set_exec(self.repo.wjoin(f), e)

    def delfile(self, f):

        try:

            os.unlink(self.repo.wjoin(f))

            #self.repo.remove([f])

        except:

            pass

    def putcommit(self, files, parents, commit):

        seen = {}

        pl = []

        for p in parents:

            if p not in seen:

                pl.append(p)

                seen[p] = 1

        parents = pl

        if len(parents) < 2: parents.append("0" * 40)

        if len(parents) < 2: parents.append("0" * 40)

        p2 = parents.pop(0)

        text = commit.desc

        extra = {}

        try:

            extra["branch"] = commit.branch

        except AttributeError:

            pass

        while parents:

            p1 = p2

            p2 = parents.pop(0)

            a = self.repo.rawcommit(files, text, commit.author, commit.date,

                                    hg.bin(p1), hg.bin(p2), extra=extra)

            text = "(octopus merge fixup)\n"

            p2 = hg.hex(self.repo.changelog.tip())

        return p2

    def puttags(self, tags):

        try:

            old = self.repo.wfile(".hgtags").read()

            oldlines = old.splitlines(1)

            oldlines.sort()

        except:

            oldlines = []

        k = tags.keys()

        k.sort()

        newlines = []

        for tag in k:

            newlines.append("%s %s\n" % (tags[tag], tag))

        newlines.sort()

        if newlines != oldlines:

            status("updating tags\n")

            f = self.repo.wfile(".hgtags", "w")

            f.write("".join(newlines))

            f.close()

            if not oldlines: self.repo.add([".hgtags"])

            date = "%s 0" % int(time.mktime(time.gmtime()))

            self.repo.rawcommit([".hgtags"], "update tags", "convert-repo",

                                date, self.repo.changelog.tip(), hg.nullid)

            return hg.hex(self.repo.changelog.tip())

converters = [convert_cvs, convert_git, convert_mercurial]

def converter(path):

    if not os.path.isdir(path):

        abort("%s: not a directory\n" % path)

    for c in converters:

        try:

            return c(path)

        except NoRepo:

            pass

    abort("%s: unknown repository type\n" % path)

class convert:

    def __init__(self, source, dest, mapfile):

        self.source = source

        self.dest = dest

        self.mapfile = mapfile

        self.commitcache = {}

        self.map = {}

        try:

            for l in file(self.mapfile):

                sv, dv = l[:-1].split()

                self.map[sv] = dv

        except IOError:

            pass

    def walktree(self, heads):

        visit = heads

        known = {}

        parents = {}

        while visit:

            n = visit.pop(0)

            if n in known or n in self.map: continue

            known[n] = 1

            self.commitcache[n] = self.source.getcommit(n)

            cp = self.commitcache[n].parents

            for p in cp:

                parents.setdefault(n, []).append(p)

                visit.append(p)

        return parents

    def toposort(self, parents):

        visit = parents.keys()

        seen = {}

        children = {}

        while visit:

            n = visit.pop(0)

            if n in seen: continue

            seen[n] = 1

            pc = 0

            if n in parents:

                for p in parents[n]:

                    if p not in self.map: pc += 1

                    visit.append(p)

                    children.setdefault(p, []).append(n)

            if not pc: root = n

        s = []

        removed = {}

        visit = children.keys()

        while visit:

            n = visit.pop(0)

            if n in removed: continue

            dep = 0

            if n in parents:

                for p in parents[n]: