view contrib/convert-repo @ 1076:01db658cc78a

tarball support v0.3

Hello,

I'm slowly improving support for tarballs in Mercurial. Attached patch
is made against current tip in Matt's repository - f859e9cba1b9, and
contains everything done so far.

Changes:
- gzip and bzip2 tarballs are sent immediately without writing to
temporary files (I was wrong Matt, it can be done very easy)
- hgrc customization, you can choose which type (if any) you will support

There's no easy way to support compression levels, since TarFile open()
assume that it is 9. I tried to use gzopen(), and bz2open() methods
instead, but it seems that headers of generated archives, are missing or
wrong. We could eventually try to rewrite tarfile.py and include our own
version into Mercurial, but I don't know if it's good idea...

Wojtek

#!/usr/bin/env python
#
# This is a generalized framework for converting between SCM
# repository formats.
#
# In its current form, it's hardcoded to convert incrementally between
# git and Mercurial.
#
# To use, you must first import the first git version into Mercurial,
# and establish a mapping between the git commit hash and the hash in
# Mercurial for that version. This mapping is kept in a simple text
# file with lines like so:
#
# <git hash> <mercurial hash>
#
# To convert the rest of the repo, run:
#
# convert-repo <git-dir> <hg-dir> <mapfile>
#
# This updates the mapfile on each commit copied, so it can be
# interrupted and can be run repeatedly to copy new commits.

import sys, os, zlib, sha, time
from mercurial import hg, ui, util

class convert_git:
    def __init__(self, path):
        self.path = path

    def getheads(self):
        h = file(self.path + "/.git/HEAD").read()[:-1]
        return [h]

    def catfile(self, rev, type):
        if rev == "0" * 40: raise IOError()
        path = os.getcwd()
        os.chdir(self.path)
        fh = os.popen("git-cat-file %s %s 2>/dev/null" % (type, rev))
        os.chdir(path)
        return fh.read()

    def getfile(self, name, rev):
        return self.catfile(rev, "blob")

    def getchanges(self, version):
        path = os.getcwd()
        os.chdir(self.path)
        fh = os.popen("git-diff-tree --root -m -r %s" % (version))
        os.chdir(path)

        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")
            changes.append((f, h, p))
        return changes

    def getcommit(self, version):
        c = self.catfile(version, "commit") # read the commit hash
        end = c.find("\n\n")
        message = c[end+2:]
        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()
                date = " ".join(p[-2:])
                author = " ".join(p[:-2])
                if author[0] == "<": author = author[1:-1]
            if n == "committer":
                p = v.split()
                date = " ".join(p[-2:])
                committer = " ".join(p[:-2])
                if committer[0] == "<": committer = committer[1:-1]
                message += "\ncommitter: %s %s\n" % (committer, date)
            if n == "parent": parents.append(v)
        return (parents, author, date, message)

    def gettags(self):
        tags = {}
        for f in os.listdir(self.path + "/.git/refs/tags"):
            try:
                h = file(self.path + "/.git/refs/tags/" + f).read().strip()
                p, a, d, m = self.getcommit(h)
                if not p: p = [h] # git is ugly, don't blame me
                tags[f] = p[0]
            except:
                pass
        return tags

class convert_mercurial:
    def __init__(self, path):
        self.path = path
        u = ui.ui()
        self.repo = hg.repository(u, path)

    def getheads(self):
        h = self.repo.changelog.heads()
        h = [ hg.hex(x) for x in h ]
        return 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, author, dest, text):
        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)

        while parents:
            p1 = p2
            p2 = parents.pop(0)
            self.repo.rawcommit(files, text, author, dest,
                                hg.bin(p1), hg.bin(p2))
            text = "(octopus merge fixup)\n"

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

    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:
            print "updating tags"
            f = self.repo.wfile(".hgtags", "w")
            f.write("".join(newlines))
            f.close()
            if not oldlines: self.repo.add([".hgtags"])
            date = "%s 0" % time.mktime(time.gmtime())
            self.repo.rawcommit([".hgtags"], "update tags", "convert-repo",
                                date, self.repo.changelog.tip(), hg.nullid)

class convert:
    def __init__(self, source, dest, mapfile):
        self.source = source
        self.dest = dest
        self.mapfile = mapfile
        self.commitcache = {}

        self.map = {}
        for l in file(self.mapfile):
            sv, dv = l[:-1].split()
            self.map[sv] = dv

    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][0]
            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]:
                    if p in self.map: continue
                    if p not in removed:
                        # we're still dependent
                        visit.append(n)
                        dep = 1
                        break

            if not dep:
                # all n's parents are in the list
                removed[n] = 1
                s.append(n)
                if n in children:
                    for c in children[n]:
                        visit.insert(0, c)

        return s

    def copy(self, rev):
        p, a, d, t = self.commitcache[rev]
        files = self.source.getchanges(rev)

        for f,v,e in files:
            try:
                data = self.source.getfile(f, v)
            except IOError, inst:
                self.dest.delfile(f)
            else:
                self.dest.putfile(f, e, data)

        r = [self.map[v] for v in p]
        f = [f for f,v,e in files]
        self.map[rev] = self.dest.putcommit(f, r, a, d, t)
        file(self.mapfile, "a").write("%s %s\n" % (rev, self.map[rev]))

    def convert(self):
        heads = self.source.getheads()
        parents = self.walktree(heads)
        t = self.toposort(parents)
        num = len(t)

        for c in t:
            num -= 1
            if c in self.map: continue
            desc = self.commitcache[c][3].splitlines()[0]
            print num, desc
            self.copy(c)

        tags = self.source.gettags()
        ctags = {}
        for k in tags:
            v = tags[k]
            if v in self.map:
                ctags[k] = self.map[v]

        self.dest.puttags(ctags)

gitpath, hgpath, mapfile = sys.argv[1:]

c = convert(convert_git(gitpath), convert_mercurial(hgpath), mapfile)
c.convert()