hg-fast-export.sh should fail if git-fast-import fails
[fast-export/barak.git] / hg2git.py
blobbaa41cdabbe387dd2096152a9811a2167969791a
1 #!/usr/bin/env python
3 # Copyright (c) 2007, 2008 Rocco Rutte <pdmef@gmx.net> and others.
4 # License: MIT <http://www.opensource.org/licenses/mit-license.php>
6 from mercurial import repo,hg,cmdutil,util,ui,revlog,node
7 import re
8 import os
9 import sys
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('([^<]+) (<[^>]+>)$')
17 # silly regex to clean out user names
18 user_clean_re=re.compile('^["]([^"]+)["]$')
20 def set_default_branch(name):
21 global cfg_master
22 cfg_master = name
24 def set_origin_name(name):
25 global origin_name
26 origin_name = name
28 def setup_repo(url):
29 try:
30 myui=ui.ui(interactive=False)
31 except TypeError:
32 myui=ui.ui()
33 myui.setconfig('ui', 'interactive', 'off')
34 return myui,hg.repository(myui,url)
36 def fixup_user(user,authors):
37 user=user.strip("\"")
38 if authors!=None:
39 # if we have an authors table, try to get mapping
40 # by defaulting to the current value of 'user'
41 user=authors.get(user,user)
42 name,mail,m='','',user_re.match(user)
43 if m==None:
44 # if we don't have 'Name <mail>' syntax, use 'user
45 # <devnull@localhost>' if use contains no at and
46 # 'user <user>' otherwise
47 name=user
48 if '@' not in user:
49 mail='<devnull@localhost>'
50 else:
51 mail='<%s>' % user
52 else:
53 # if we have 'Name <mail>' syntax, everything is fine :)
54 name,mail=m.group(1),m.group(2)
56 # remove any silly quoting from username
57 m2=user_clean_re.match(name)
58 if m2!=None:
59 name=m2.group(1)
60 return '%s %s' % (name,mail)
62 def get_branch(name):
63 # 'HEAD' is the result of a bug in mutt's cvs->hg conversion,
64 # other CVS imports may need it, too
65 if name=='HEAD' or name=='default' or name=='':
66 name=cfg_master
67 if origin_name:
68 return origin_name + '/' + name
69 return name
71 def get_changeset(ui,repo,revision,authors={}):
72 node=repo.lookup(revision)
73 (manifest,user,(time,timezone),files,desc,extra)=repo.changelog.read(node)
74 tz="%+03d%02d" % (-timezone / 3600, ((-timezone % 3600) / 60))
75 branch=get_branch(extra.get('branch','master'))
76 return (node,manifest,fixup_user(user,authors),(time,tz),files,desc,branch,extra)
78 def mangle_key(key):
79 return key
81 def load_cache(filename,get_key=mangle_key):
82 cache={}
83 if not os.path.exists(filename):
84 return cache
85 f=open(filename,'r')
86 l=0
87 for line in f.readlines():
88 l+=1
89 fields=line.split(' ')
90 if fields==None or not len(fields)==2 or fields[0][0]!=':':
91 sys.stderr.write('Invalid file format in [%s], line %d\n' % (filename,l))
92 continue
93 # put key:value in cache, key without ^:
94 cache[get_key(fields[0][1:])]=fields[1].split('\n')[0]
95 f.close()
96 return cache
98 def save_cache(filename,cache):
99 f=open(filename,'w+')
100 map(lambda x: f.write(':%s %s\n' % (str(x),str(cache.get(x)))),cache.keys())
101 f.close()
103 def get_git_sha1(name,type='heads'):
104 try:
105 # use git-rev-parse to support packed refs
106 cmd="GIT_DIR='%s' git rev-parse --verify refs/%s/%s 2>/dev/null" % (os.getenv('GIT_DIR','/dev/null'),type,name)
107 p=os.popen(cmd)
108 l=p.readline()
109 p.close()
110 if l == None or len(l) == 0:
111 return None
112 return l[0:40]
113 except IOError:
114 return None