hg2git.py

   1 #!/usr/bin/env python
   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 repo,hg,cmdutil,util,ui,revlog,node
   7 import re
   8 import os
   9 import sys
  10
  11 # default git branch name
  12 cfg_master='master'
  13 # default origin name
  14 origin_name=''
  15 # silly regex to see if user field has email address
  16 user_re=re.compile('([^<]+) (<[^>]+>)$')
  17 # silly regex to clean out user names
  18 user_clean_re=re.compile('^["]([^"]+)["]$')
  19
  20 def set_default_branch(name):
  21   global cfg_master
  22   cfg_master = name
  23
  24 def set_origin_name(name):
  25   global origin_name
  26   origin_name = name
  27
  28 def setup_repo(url):
  29   myui=ui.ui()
  30   myui.setconfig('ui', 'interactive', 'off')
  31   return myui,hg.repository(myui,url)
  32
  33 def fixup_user(user,authors):
  34   if authors!=None:
  35     # if we have an authors table, try to get mapping
  36     # by defaulting to the current value of 'user'
  37     user=authors.get(user,user)
  38   name,mail,m='','',user_re.match(user)
  39   if m==None:
  40     # if we don't have 'Name <mail>' syntax, use 'user
  41     # <devnull@localhost>' if use contains no at and
  42     # 'user <user>' otherwise
  43     name=user
  44     if '@' not in user:
  45       mail='<devnull@localhost>'
  46     else:
  47       mail='<%s>' % user
  48   else:
  49     # if we have 'Name <mail>' syntax, everything is fine :)
  50     name,mail=m.group(1),m.group(2)
  51
  52   # remove any silly quoting from username
  53   m2=user_clean_re.match(name)
  54   if m2!=None:
  55     name=m2.group(1)
  56   return '%s %s' % (name,mail)
  57
  58 def get_branch(name):
  59   # 'HEAD' is the result of a bug in mutt's cvs->hg conversion,
  60   # other CVS imports may need it, too
  61   if name=='HEAD' or name=='default' or name=='':
  62     name=cfg_master
  63   if origin_name:
  64     return origin_name + '/' + name
  65   return name
  66
  67 def get_changeset(ui,repo,revision,authors={}):
  68   node=repo.lookup(revision)
  69   (manifest,user,(time,timezone),files,desc,extra)=repo.changelog.read(node)
  70   tz="%+03d%02d" % (-timezone / 3600, ((-timezone % 3600) / 60))
  71   branch=get_branch(extra.get('branch','master'))
  72   return (node,manifest,fixup_user(user,authors),(time,tz),files,desc,branch,extra)
  73
  74 def mangle_key(key):
  75   return key
  76
  77 def load_cache(filename,get_key=mangle_key):
  78   cache={}
  79   if not os.path.exists(filename):
  80     return cache
  81   f=open(filename,'r')
  82   l=0
  83   for line in f.readlines():
  84     l+=1
  85     fields=line.split(' ')
  86     if fields==None or not len(fields)==2 or fields[0][0]!=':':
  87       sys.stderr.write('Invalid file format in [%s], line %d\n' % (filename,l))
  88       continue
  89     # put key:value in cache, key without ^:
  90     cache[get_key(fields[0][1:])]=fields[1].split('\n')[0]
  91   f.close()
  92   return cache
  93
  94 def save_cache(filename,cache):
  95   f=open(filename,'w+')
  96   map(lambda x: f.write(':%s %s\n' % (str(x),str(cache.get(x)))),cache.keys())
  97   f.close()
  98
  99 def get_git_sha1(name,type='heads'):
 100   try:
 101     # use git-rev-parse to support packed refs
 102     cmd="GIT_DIR='%s' git rev-parse --verify refs/%s/%s 2>/dev/null" % (os.getenv('GIT_DIR','/dev/null'),type,name)
 103     p=os.popen(cmd)
 104     l=p.readline()
 105     p.close()
 106     if l == None or len(l) == 0:
 107       return None
 108     return l[0:40]
 109   except IOError:
 110     return None