view mercurial/filelog.py @ 1335:bea6356b8bca

git -> hg conversion script

contrib/convert-repo changes:
- do not print verbose output so that error messages are seen more easily
- Output the date as integer and not as floating point number.
- Do not require a ".git" subdirectory to work on, but use the GIT_DIR
  environment var to specify the git repository.

Change is otherwise compatible to the current version and I have tested it
by converting the kernel and several git respositories from kernel.org.
(Btw, the udev test dir contains a /sys dir with entries which should not
be normal dirs and not be normal files. ;-)

Thanks again for mercurial,

Florian La Roche



--- a/contrib/convert-repo
+++ b/contrib/convert-repo
@@ -28,26 +28,18 @@
         self.path = path

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

     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)
+        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):
-        path = os.getcwd()
-        os.chdir(self.path)
-        fh = os.popen("git-diff-tree --root -m -r %s" % (version))
-        os.chdir(path)
-
+        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
@@ -83,9 +75,9 @@

     def gettags(self):
         tags = {}
-        for f in os.listdir(self.path + "/.git/refs/tags"):
+        for f in os.listdir(self.path + "/refs/tags"):
             try:
-                h = file(self.path + "/.git/refs/tags/" + f).read().strip()
+                h = file(self.path + "/refs/tags/" + f).read().strip()
                 tags[f] = h
             except:
                 pass
@@ -99,8 +91,7 @@

     def getheads(self):
         h = self.repo.changelog.heads()
-        h = [ hg.hex(x) for x in h ]
-        return h
+        return [ hg.hex(x) for x in h ]

     def putfile(self, f, e, data):
         self.repo.wfile(f, "w").write(data)
@@ -155,12 +146,12 @@
         newlines.sort()

         if newlines != oldlines:
-            print "updating tags"
+            #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())
+            date = "%s 0" % int(time.mktime(time.gmtime()))
             self.repo.rawcommit([".hgtags"], "update tags", "convert-repo",
                                 date, self.repo.changelog.tip(), hg.nullid)

@@ -262,7 +253,7 @@
             num -= 1
             if c in self.map: continue
             desc = self.commitcache[c][3].splitlines()[0]
-            print num, desc
+            #print num, desc
             self.copy(c)

         tags = self.source.gettags()
@@ -275,6 +266,8 @@
         self.dest.puttags(ctags)

 gitpath, hgpath, mapfile = sys.argv[1:]
+if os.path.isdir(gitpath + "/.git"):
+    gitpath += "/.git"

 c = convert(convert_git(gitpath), convert_mercurial(hgpath), mapfile)
 c.convert()
_______________________________________________
Mercurial mailing list
[e-mail]
http://selenic.com/mailman/listinfo/mercurial

# filelog.py - file history class for mercurial
#
# Copyright 2005 Matt Mackall <[e-mail]>
#
# This software may be used and distributed according to the terms
# of the GNU General Public License, incorporated herein by reference.

import os
from revlog import *
from demandload import *
demandload(globals(), "bdiff")

class filelog(revlog):
    def __init__(self, opener, path):
        revlog.__init__(self, opener,
                        os.path.join("data", self.encodedir(path + ".i")),
                        os.path.join("data", self.encodedir(path + ".d")))

    # This avoids a collision between a file named foo and a dir named
    # foo.i or foo.d
    def encodedir(self, path):
        return (path
                .replace(".hg/", ".hg.hg/")
                .replace(".i/", ".i.hg/")
                .replace(".d/", ".d.hg/"))

    def decodedir(self, path):
        return (path
                .replace(".d.hg/", ".d/")
                .replace(".i.hg/", ".i/")
                .replace(".hg.hg/", ".hg/"))

    def read(self, node):
        t = self.revision(node)
        if not t.startswith('\1\n'):
            return t
        s = t.find('\1\n', 2)
        return t[s+2:]

    def readmeta(self, node):
        t = self.revision(node)
        if not t.startswith('\1\n'):
            return {}
        s = t.find('\1\n', 2)
        mt = t[2:s]
        m = {}
        for l in mt.splitlines():
            k, v = l.split(": ", 1)
            m[k] = v
        return m

    def add(self, text, meta, transaction, link, p1=None, p2=None):
        if meta or text.startswith('\1\n'):
            mt = ""
            if meta:
                mt = [ "%s: %s\n" % (k, v) for k,v in meta.items() ]
            text = "\1\n" + "".join(mt) + "\1\n" + text
        return self.addrevision(text, transaction, link, p1, p2)

    def renamed(self, node):
        if 0 and self.parents(node)[0] != nullid:
            return False
        m = self.readmeta(node)
        if m and m.has_key("copy"):
            return (m["copy"], bin(m["copyrev"]))
        return False

    def annotate(self, node):

        def decorate(text, rev):
            return ([rev] * len(text.splitlines()), text)

        def pair(parent, child):
            for a1, a2, b1, b2 in bdiff.blocks(parent[1], child[1]):
                child[0][b1:b2] = parent[0][a1:a2]
            return child

        # find all ancestors
        needed = {node:1}
        visit = [node]
        while visit:
            n = visit.pop(0)
            for p in self.parents(n):
                if p not in needed:
                    needed[p] = 1
                    visit.append(p)
                else:
                    # count how many times we'll use this
                    needed[p] += 1

        # sort by revision which is a topological order
        visit = [ (self.rev(n), n) for n in needed.keys() ]
        visit.sort()
        hist = {}

        for r,n in visit:
            curr = decorate(self.read(n), self.linkrev(n))
            for p in self.parents(n):
                if p != nullid:
                    curr = pair(hist[p], curr)
                    # trim the history of unneeded revs
                    needed[p] -= 1
                    if not needed[p]:
                        del hist[p]
            hist[n] = curr

        return zip(hist[n][0], hist[n][1].splitlines(1))