hg2git.py

   1 #!/usr/bin/env python3
   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 binnode
   9
  10 import re
  11 import os
  12 import sys
  13 import subprocess
  14
  15 # default git branch name
  16 cfg_master=b'master'
  17 # default origin name
  18 origin_name=b''
  19 # silly regex to see if user field has email address
  20 user_re=re.compile(b'([^<]+) (<[^>]*>)$')
  21 # silly regex to clean out user names
  22 user_clean_re=re.compile(b'^["]([^"]+)["]$')
  23
  24 def set_default_branch(name):
  25   global cfg_master
  26   cfg_master = name.encode('utf8') if not isinstance(name, bytes) else name
  27
  28 def set_origin_name(name):
  29   global origin_name
  30   origin_name = name
  31
  32 def setup_repo(url):
  33   try:
  34     myui=ui.ui(interactive=False)
  35   except TypeError:
  36     myui=ui.ui()
  37     myui.setconfig(b'ui', b'interactive', b'off')
  38     # Avoids a warning when the repository has obsolete markers
  39     myui.setconfig(b'experimental', b'evolution.createmarkers', True)
  40   return myui,hg.repository(myui, os.fsencode(url)).unfiltered()
  41
  42 def fixup_user(user,authors):
  43   user=user.strip(b"\"")
  44   if authors!=None:
  45     # if we have an authors table, try to get mapping
  46     # by defaulting to the current value of 'user'
  47     user=authors.get(user,user)
  48   name,mail,m=b'',b'',user_re.match(user)
  49   if m==None:
  50     # if we don't have 'Name <mail>' syntax, extract name
  51     # and mail from hg helpers. this seems to work pretty well.
  52     # if email doesn't contain @, replace it with devnull@localhost
  53     name=templatefilters.person(user)
  54     mail=b'<%s>' % templatefilters.email(user)
  55     if b'@' not in mail:
  56       mail = b'<devnull@localhost>'
  57   else:
  58     # if we have 'Name <mail>' syntax, everything is fine :)
  59     name,mail=m.group(1),m.group(2)
  60
  61   # remove any silly quoting from username
  62   m2=user_clean_re.match(name)
  63   if m2!=None:
  64     name=m2.group(1)
  65   return b'%s %s' % (name,mail)
  66
  67 def get_branch(name):
  68   # 'HEAD' is the result of a bug in mutt's cvs->hg conversion,
  69   # other CVS imports may need it, too
  70   if name==b'HEAD' or name==b'default' or name==b'':
  71     name=cfg_master
  72   if origin_name:
  73     return origin_name + b'/' + name
  74   return name
  75
  76 def get_changeset(ui,repo,revision,authors={},encoding=''):
  77   (manifest,user,(time,timezone),files,desc,extra)=repo.changelog.read(revision)
  78   if encoding:
  79     user=user.decode(encoding).encode('utf8')
  80     desc=desc.decode(encoding).encode('utf8')
  81   tz=b"%+03d%02d" % (-timezone // 3600, ((-timezone % 3600) // 60))
  82   branch=get_branch(extra.get(b'branch', b''))
  83   return (manifest,fixup_user(user,authors),(time,tz),files,desc,branch,extra)
  84
  85 def mangle_key(key):
  86   return key
  87
  88 def load_cache(filename,get_key=mangle_key):
  89   cache={}
  90   if not os.path.exists(filename):
  91     return cache
  92   f=open(filename,'rb')
  93   l=0
  94   for line in f.readlines():
  95     l+=1
  96     fields=line.split(b' ')
  97     if fields==None or not len(fields)==2 or fields[0][0:1]!=b':':
  98       sys.stderr.write('Invalid file format in [%s], line %d\n' % (filename,l))
  99       continue
 100     # put key:value in cache, key without ^:
 101     cache[get_key(fields[0][1:])]=fields[1].split(b'\n')[0]
 102   f.close()
 103   return cache
 104
 105 def save_cache(filename,cache):
 106   f=open(filename,'wb')
 107   for key, value in cache.items():
 108     if not isinstance(key, bytes):
 109       key = str(key).encode('utf8')
 110     if not isinstance(value, bytes):
 111       value = str(value).encode('utf8')
 112     f.write(b':%s %s\n' % (key, value))
 113   f.close()
 114
 115 def get_git_sha1(name,type='heads'):
 116   try:
 117     # use git-rev-parse to support packed refs
 118     ref="refs/%s/%s" % (type,name.decode('utf8'))
 119     l=subprocess.check_output(["git", "rev-parse", "--verify",
 120                                "--quiet", ref.encode('utf8')])
 121     if l == None or len(l) == 0:
 122       return None
 123     return l[0:40]
 124   except subprocess.CalledProcessError:
 125     return None