hg2git.py

   1 #!/usr/bin/env python2
   2
   3 # Copyright (c) 2007, 2008 Rocco Rutte <pdmef@gmx.net> and others.
   4 # License: MIT <http://www.opensource.org/licenses/mit-license.php>
   5
   6 from mercurial import hg,util,ui,templatefilters
   7 from mercurial import error as hgerror
   8 from mercurial.scmutil import revsymbol,binnode
   9
  10 import re
  11 import os
  12 import sys
  13 import subprocess
  14
  15 PY2 = sys.version_info.major < 3
  16 if PY2:
  17   str = unicode
  18   fsencode = lambda s: s.encode(sys.getfilesystemencoding())
  19 else:
  20   from os import fsencode
  21
  22 # default git branch name
  23 cfg_master=b'master'
  24 # default origin name
  25 origin_name=b''
  26 # silly regex to see if user field has email address
  27 user_re=re.compile(b'([^<]+) (<[^>]*>)$')
  28 # silly regex to clean out user names
  29 user_clean_re=re.compile(b'^["]([^"]+)["]$')
  30
  31 def set_default_branch(name):
  32   global cfg_master
  33   cfg_master = name.encode('utf8') if not isinstance(name, bytes) else name
  34
  35 def set_origin_name(name):
  36   global origin_name
  37   origin_name = name
  38
  39 def setup_repo(url):
  40   try:
  41     myui=ui.ui(interactive=False)
  42   except TypeError:
  43     myui=ui.ui()
  44     myui.setconfig(b'ui', b'interactive', b'off')
  45     # Avoids a warning when the repository has obsolete markers
  46     myui.setconfig(b'experimental', b'evolution.createmarkers', True)
  47   return myui,hg.repository(myui, fsencode(url)).unfiltered()
  48
  49 def fixup_user(user,authors):
  50   user=user.strip(b"\"")
  51   if authors!=None:
  52     # if we have an authors table, try to get mapping
  53     # by defaulting to the current value of 'user'
  54     user=authors.get(user,user)
  55   name,mail,m=b'',b'',user_re.match(user)
  56   if m==None:
  57     # if we don't have 'Name <mail>' syntax, extract name
  58     # and mail from hg helpers. this seems to work pretty well.
  59     # if email doesn't contain @, replace it with devnull@localhost
  60     name=templatefilters.person(user)
  61     mail=b'<%s>' % templatefilters.email(user)
  62     if b'@' not in mail:
  63       mail = b'<devnull@localhost>'
  64   else:
  65     # if we have 'Name <mail>' syntax, everything is fine :)
  66     name,mail=m.group(1),m.group(2)
  67
  68   # remove any silly quoting from username
  69   m2=user_clean_re.match(name)
  70   if m2!=None:
  71     name=m2.group(1)
  72   return b'%s %s' % (name,mail)
  73
  74 def get_branch(name):
  75   # 'HEAD' is the result of a bug in mutt's cvs->hg conversion,
  76   # other CVS imports may need it, too
  77   if name==b'HEAD' or name==b'default' or name==b'':
  78     name=cfg_master
  79   if origin_name:
  80     return origin_name + b'/' + name
  81   return name
  82
  83 def get_changeset(ui,repo,revision,authors={},encoding=''):
  84   # Starting with Mercurial 4.6 lookup no longer accepts raw hashes
  85   # for lookups. Work around it by changing our behaviour depending on
  86   # how it fails
  87   try:
  88     node=repo.lookup(revision)
  89   except (TypeError, hgerror.ProgrammingError):
  90     node=binnode(revsymbol(repo, b"%d" % revision)) # We were given a numeric rev
  91   except hgerror.RepoLookupError:
  92     node=revision # We got a raw hash
  93   (manifest,user,(time,timezone),files,desc,extra)=repo.changelog.read(node)
  94   if encoding:
  95     user=user.decode(encoding).encode('utf8')
  96     desc=desc.decode(encoding).encode('utf8')
  97   tz=b"%+03d%02d" % (-timezone // 3600, ((-timezone % 3600) // 60))
  98   branch=get_branch(extra.get(b'branch', b''))
  99   return (node,manifest,fixup_user(user,authors),(time,tz),files,desc,branch,extra)
 100
 101 def mangle_key(key):
 102   return key
 103
 104 def load_cache(filename,get_key=mangle_key):
 105   cache={}
 106   if not os.path.exists(filename):
 107     return cache
 108   f=open(filename,'rb')
 109   l=0
 110   for line in f.readlines():
 111     l+=1
 112     fields=line.split(b' ')
 113     if fields==None or not len(fields)==2 or fields[0][0:1]!=b':':
 114       sys.stderr.write('Invalid file format in [%s], line %d\n' % (filename,l))
 115       continue
 116     # put key:value in cache, key without ^:
 117     cache[get_key(fields[0][1:])]=fields[1].split(b'\n')[0]
 118   f.close()
 119   return cache
 120
 121 def save_cache(filename,cache):
 122   f=open(filename,'wb')
 123   for key, value in cache.items():
 124     if not isinstance(key, bytes):
 125       key = str(key).encode('utf8')
 126     if not isinstance(value, bytes):
 127       value = str(value).encode('utf8')
 128     f.write(b':%s %s\n' % (key, value))
 129   f.close()
 130
 131 def get_git_sha1(name,type='heads'):
 132   try:
 133     # use git-rev-parse to support packed refs
 134     ref="refs/%s/%s" % (type,name.decode('utf8'))
 135     l=subprocess.check_output(["git", "rev-parse", "--verify",
 136                                "--quiet", ref.encode('utf8')])
 137     if l == None or len(l) == 0:
 138       return None
 139     return l[0:40]
 140   except subprocess.CalledProcessError:
 141     return None