hg2git.py

   1 #!/usr/bin/env python
   2
   3 # Copyright (c) 2007 Rocco Rutte <pdmef@gmx.net>
   4 # License: GPLv2
   5
   6 """hg2git.py - A mercurial-to-git filter for git-fast-import(1)
   7 Usage: hg2git.py <hg repo url> <marks file> <heads file> <tip file>
   8 """
   9
  10 from mercurial import repo,hg,cmdutil,util,ui,revlog,node
  11 from tempfile import mkstemp
  12 import re
  13 import sys
  14 import os
  15
  16 # silly regex to see if user field has email address
  17 user_re=re.compile('[^<]+ <[^>]+>$')
  18 # git branch for hg's default 'HEAD' branch
  19 cfg_master='master'
  20 # insert 'checkpoint' command after this many commits
  21 cfg_checkpoint_count=1000
  22
  23 def usage(ret):
  24   sys.stderr.write(__doc__)
  25   return ret
  26
  27 def setup_repo(url):
  28   myui=ui.ui()
  29   return myui,hg.repository(myui,url)
  30
  31 def get_changeset(ui,repo,revision):
  32   def get_branch(name):
  33     if name=='HEAD':
  34       name=cfg_master
  35     return name
  36   def fixup_user(user):
  37     if user_re.match(user)==None:
  38       if '@' not in user:
  39         return user+' <none@none>'
  40       return user+' <'+user+'>'
  41     return user
  42   node=repo.lookup(revision)
  43   (manifest,user,(time,timezone),files,desc,extra)=repo.changelog.read(node)
  44   tz="%+03d%02d" % (-timezone / 3600, ((-timezone % 3600) / 60))
  45   branch=get_branch(extra.get('branch','master'))
  46   return (manifest,fixup_user(user),(time,tz),files,desc,branch,extra)
  47
  48 def gitmode(x):
  49   return x and '100755' or '100644'
  50
  51 def wr(msg=''):
  52   print msg
  53   #map(lambda x: sys.stderr.write('\t[%s]\n' % x),msg.split('\n'))
  54
  55 def checkpoint(count):
  56   count=count+1
  57   if count%cfg_checkpoint_count==0:
  58     sys.stderr.write("Checkpoint after %d commits\n" % count)
  59     wr('checkpoint')
  60     wr()
  61   return count
  62
  63 def get_parent_mark(parent,marks):
  64   """Get the mark for some parent.
  65   If we saw it in the current session, return :%d syntax and
  66   otherwise the SHA1 from the cache."""
  67   return marks.get(str(parent+1),':%d' % (parent+1))
  68
  69 def mismatch(x,f1,f2):
  70   """See if two revisions of a file are not equal."""
  71   return node.hex(f1)!=node.hex(f2)
  72
  73 def outer_set(dleft,dright,l,r):
  74   """Loop over our repository in and find all changed and missing files."""
  75   for left in dleft.keys():
  76     right=dright.get(left,None)
  77     if right==None or mismatch('A',dleft[left],right):
  78       # if either have the current file not in parent or the
  79       # checksums differ: add it to changed files
  80       l.append(left)
  81   for right in dright.keys():
  82     left=dleft.get(right,None)
  83     if left==None:
  84       # if we have a file in the parent but not our manifest,
  85       # add it to deleted files; checksums are checked earlier
  86       r.append(right)
  87   return l,r
  88
  89 def get_filechanges(repo,revision,parents,mleft):
  90   """Given some repository and revision, find all changed/deleted files."""
  91   l,r=[],[]
  92   for p in parents:
  93     if p<0: continue
  94     mright=repo.changectx(p).manifest()
  95     dleft=mleft.keys()
  96     dleft.sort()
  97     dright=mright.keys()
  98     dright.sort()
  99     l,r=outer_set(mleft,mright,l,r)
 100   return l,r
 101
 102 def export_commit(ui,repo,revision,marks,heads,last,max,count):
 103   (_,user,(time,timezone),files,desc,branch,_)=get_changeset(ui,repo,revision)
 104   parents=repo.changelog.parentrevs(revision)
 105
 106   wr('commit refs/heads/%s' % branch)
 107   wr('mark :%d' % (revision+1))
 108   wr('committer %s %d %s' % (user,time,timezone))
 109   wr('data %d' % (len(desc)+1)) # wtf?
 110   wr(desc)
 111   wr()
 112
 113   src=heads.get(branch,'')
 114   link=''
 115   if src!='':
 116     # if we have a cached head, this is an incremental import: initialize it
 117     # and kill reference so we won't init it again
 118     wr('from %s' % src)
 119     heads[branch]=''
 120     sys.stderr.write('Initializing branch [%s] to parent [%s]\n' %
 121         (branch,src))
 122     link=src # avoid making a merge commit for incremental import
 123   elif link=='' and not heads.has_key(branch) and revision>0:
 124     # newly created branch and not the first one: connect to parent
 125     tmp=get_parent_mark(parents[0],marks)
 126     wr('from %s' % tmp)
 127     sys.stderr.write('Link new branch [%s] to parent [%s]\n' %
 128         (branch,tmp))
 129     link=tmp # avoid making a merge commit for branch fork
 130
 131   if parents:
 132     l=last.get(branch,revision)
 133     for p in parents:
 134       # 1) as this commit implicitely is the child of the most recent
 135       #    commit of this branch, ignore this parent
 136       # 2) ignore nonexistent parents
 137       # 3) merge otherwise
 138       if p==l or p==revision or p<0:
 139         continue
 140       tmp=get_parent_mark(p,marks)
 141       # if we fork off a branch, don't merge with our parent via 'merge'
 142       # as we have 'from' already above
 143       if tmp==link:
 144         continue
 145       sys.stderr.write('Merging branch [%s] with parent [%s] from [r%d]\n' %
 146           (branch,tmp,p))
 147       wr('merge %s' % tmp)
 148
 149   last[branch]=revision
 150   heads[branch]=''
 151   # we need this later to write out tags
 152   marks[str(revision)]=':%d'%(revision+1)
 153
 154   ctx=repo.changectx(str(revision))
 155   man=ctx.manifest()
 156   added,removed=get_filechanges(repo,revision,parents,man)
 157
 158   sys.stderr.write('Exporting revision %d with %d changed/%d removed files\n' %
 159       (revision,len(added),len(removed)))
 160
 161   for a in added:
 162     fctx=ctx.filectx(a)
 163     d=fctx.data()
 164     wr('M %s inline %s' % (gitmode(man.execf(a)),a))
 165     wr('data %d' % len(d)) # had some trouble with size()
 166     wr(d)
 167
 168   for r in removed:
 169     wr('D %s' % r)
 170
 171   wr()
 172   return checkpoint(count)
 173
 174 def export_tags(ui,repo,cache,count):
 175   l=repo.tagslist()
 176   for tag,node in l:
 177     if tag=='tip':
 178       continue
 179     rev=repo.changelog.rev(node)
 180     ref=cache.get(str(rev),None)
 181     if ref==None:
 182       #sys.stderr.write('Failed to find reference for creating tag'
 183       #    ' %s at r%d\n' % (tag,rev))
 184       continue
 185     (_,user,(time,timezone),_,desc,branch,_)=get_changeset(ui,repo,rev)
 186     sys.stderr.write('Exporting tag [%s] at [hg r%d] [git %s]\n' % (tag,rev,ref))
 187     wr('tag %s' % tag)
 188     wr('from %s' % ref)
 189     wr('tagger %s %d %s' % (user,time,timezone))
 190     msg='hg2git created tag %s for hg revision %d on branch %s on (summary):\n\t%s' % (tag,
 191         rev,branch,desc.split('\n')[0])
 192     wr('data %d' % (len(msg)+1))
 193     wr(msg)
 194     wr()
 195     count=checkpoint(count)
 196   return count
 197
 198 def load_cache(filename):
 199   cache={}
 200   if not os.path.exists(filename):
 201     return cache
 202   f=open(filename,'r')
 203   l=0
 204   for line in f.readlines():
 205     l+=1
 206     fields=line.split(' ')
 207     if fields==None or not len(fields)==2 or fields[0][0]!=':':
 208       sys.stderr.write('Invalid file format in [%s], line %d\n' % (filename,l))
 209       continue
 210     # put key:value in cache, key without ^:
 211     cache[fields[0][1:]]=fields[1].split('\n')[0]
 212   f.close()
 213   return cache
 214
 215 def save_cache(filename,cache):
 216   f=open(filename,'w+')
 217   map(lambda x: f.write(':%s %s\n' % (str(x),str(cache.get(x)))),cache.keys())
 218   f.close()
 219
 220 def verify_heads(ui,repo,cache):
 221   def getsha1(branch):
 222     f=open(os.getenv('GIT_DIR','/dev/null')+'/refs/heads/'+branch)
 223     sha1=f.readlines()[0].split('\n')[0]
 224     f.close()
 225     return sha1
 226
 227   for b in cache.keys():
 228     sys.stderr.write('Verifying branch [%s]\n' % b)
 229     sha1=getsha1(b)
 230     c=cache.get(b)
 231     if sha1!=c:
 232       sys.stderr.write('Warning: Branch [%s] modified outside hg2git:'
 233         '\n%s (repo) != %s (cache)\n' % (b,sha1,c))
 234   return True
 235
 236 if __name__=='__main__':
 237   if len(sys.argv)!=6: sys.exit(usage(1))
 238   repourl,m,marksfile,headsfile,tipfile=sys.argv[1:]
 239   _max=int(m)
 240
 241   marks_cache=load_cache(marksfile)
 242   heads_cache=load_cache(headsfile)
 243   state_cache=load_cache(tipfile)
 244
 245   ui,repo=setup_repo(repourl)
 246
 247   if not verify_heads(ui,repo,heads_cache):
 248     sys.exit(1)
 249
 250   tip=repo.changelog.count()
 251
 252   min=int(state_cache.get('tip',0))
 253   max=_max
 254   if _max<0:
 255     max=tip
 256
 257   c=int(state_cache.get('count',0))
 258   last={}
 259   for rev in range(min,max):
 260     c=export_commit(ui,repo,rev,marks_cache,heads_cache,last,tip,c)
 261
 262   c=export_tags(ui,repo,marks_cache,c)
 263
 264   state_cache['tip']=max
 265   state_cache['count']=c
 266   state_cache['repo']=repourl
 267   save_cache(tipfile,state_cache)