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 hg,ui
   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 # default email address when unknown
  16 unknown_addr='devnull@localhost'
  17 split_name_re = re.compile(r'^((?:[^<]|<at>)*?)(?:<(?!at>)(?:(?:([^>]*)|(?:(.*)>(.*)))))?$', re.S|re.I)
  18 git_crud = '\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0a\x0b\x0c\x0d\x0e\x0f' \
  19            '\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1a\x1b\x1c\x1d\x1e\x1f' \
  20            ' .,:;<>"\\' "'"
  21 git_delch_re = re.compile(r'[<>\n]+', re.S)
  22 spelled_at_re = re.compile(r'^(.*?)(?:@|(?:(?:^|\s)<at>(?:\s|$)))', re.I)
  23
  24 def set_default_branch(name):
  25   global cfg_master
  26   cfg_master = 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('ui', 'interactive', 'off')
  38   return myui,hg.repository(myui,url)
  39
  40 # Git strips "crud" characters off both the beginning and end of the user's name
  41 # and the user's email, then deletes any remaining '<', '>' and '\n' characters
  42 # before combining the user name with the user email surrounded by '<' and '>'.
  43 # This function provides the crud-stripping and deletion operation that is used
  44 # on both the name and email.
  45 def gitname(name):
  46   name = name.strip(git_crud)
  47   return git_delch_re.sub('', name)
  48
  49 def set_unknown_addr(addr):
  50   global unknown_addr
  51   ans=False
  52   if addr!=None:
  53     addr=gitname(addr)
  54     if addr!='':
  55       unknown_addr=addr
  56       ans=True
  57   return ans
  58
  59 # Split the combined name and email input into a separate name and email and
  60 # apply Git's rules to each part.  The idea is to use anything to the left of
  61 # the first '<' and to the right of the last '>' as the name.  Anything between
  62 # the first '<' and the last '>' is treated as the email.  If there is no '<'
  63 # but the name contains '@' (which may be spelled out) then treat the entire
  64 # thing as an email with no name.  If the detected name is empty then anything
  65 # up to the first '@' (which may be spelled out) in the email is used for the
  66 # name.  Failing that the entire email is used for the name.
  67 def split_name_email(combined):
  68   name = ''
  69   email,rawemail = '',''
  70   match = split_name_re.match(combined)
  71   if match:
  72     left,rest,mid,right = match.groups()
  73     if rest != None:
  74       name = gitname(left)
  75       rawemail = rest
  76       email = gitname(rawemail)
  77     elif mid != None:
  78       name = gitname(left.rstrip() + ' ' + right.lstrip())
  79       rawemail = mid
  80       email = gitname(rawemail)
  81     else:
  82       name = gitname(left)
  83   if email == '' and spelled_at_re.match(left):
  84     rawemail = left
  85     email = name
  86     name = ''
  87   if name == '':
  88     at = spelled_at_re.match(rawemail)
  89     if at:
  90       name = gitname(at.group(1))
  91     if name == '':
  92       name = email
  93     # We do this test to be compatible with the previous behavior of hg2git.py
  94     # When it's given a <email> without any name and email does not contain '@'
  95     # then it sets the email to the unknown address
  96     if (len(left) < 2 or left[-1] != ' ') and not at:
  97       email = ''
  98   return [name, email]
  99
 100 def fixup_user(user,authors):
 101   user=user.strip("\"")
 102   if authors!=None:
 103     # if we have an authors table, try to get mapping
 104     # by defaulting to the current value of 'user'
 105     user=authors.get(user,user)
 106   name,mail=split_name_email(user)
 107   if mail == '':
 108     # If we don't have an email address replace it with unknown_addr.
 109     mail = unknown_addr
 110   if name == '':
 111     # Git does not like an empty name either -- split_name_email can only
 112     # return an empty name if it also returns an empty email.  This probably
 113     # will never happen since the input would have to be empty or only "crud"
 114     # characters, but check just to be safe.
 115     name = '-'
 116   return '%s <%s>' % (name,mail)
 117
 118 def get_branch(name):
 119   # 'HEAD' is the result of a bug in mutt's cvs->hg conversion,
 120   # other CVS imports may need it, too
 121   if name=='HEAD' or name=='default' or name=='':
 122     name=cfg_master
 123   if origin_name:
 124     return origin_name + '/' + name
 125   return name
 126
 127 def get_changeset(ui,repo,revision,authors={}):
 128   node=repo.lookup(revision)
 129   (manifest,user,(time,timezone),files,desc,extra)=repo.changelog.read(node)
 130   tz="%+03d%02d" % (-timezone / 3600, ((-timezone % 3600) / 60))
 131   branch=get_branch(extra.get('branch','master'))
 132   return (node,manifest,fixup_user(user,authors),(time,tz),files,desc,branch,extra)
 133
 134 def mangle_key(key):
 135   return key
 136
 137 def load_cache(filename,get_key=mangle_key):
 138   cache={}
 139   if not os.path.exists(filename):
 140     return cache
 141   f=open(filename,'r')
 142   l=0
 143   for line in f.readlines():
 144     l+=1
 145     fields=line.split(' ')
 146     if fields==None or not len(fields)==2 or fields[0][0]!=':':
 147       sys.stderr.write('Invalid file format in [%s], line %d\n' % (filename,l))
 148       continue
 149     # put key:value in cache, key without ^:
 150     cache[get_key(fields[0][1:])]=fields[1].split('\n')[0]
 151   f.close()
 152   return cache
 153
 154 def save_cache(filename,cache):
 155   f=open(filename,'w+')
 156   map(lambda x: f.write(':%s %s\n' % (str(x),str(cache.get(x)))),cache.keys())
 157   f.close()
 158
 159 def get_git_sha1(name,type='heads'):
 160   try:
 161     # use git-rev-parse to support packed refs
 162     cmd="git rev-parse --verify refs/%s/%s 2>%s" % (type,name,os.devnull)
 163     p=os.popen(cmd)
 164     l=p.readline()
 165     p.close()
 166     if l == None or len(l) == 0:
 167       return None
 168     return l[0:40]
 169   except IOError:
 170     return None