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 or none at all if 0
  21 cfg_checkpoint_count=0
  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,authors):
  32   def get_branch(name):
  33     if name=='HEAD':
  34       name=cfg_master
  35     return name
  36   def fixup_user(user,authors):
  37     if authors!=None:
  38       # if we have an authors table, try to get mapping
  39       # by defaultung to the current value of 'user'
  40       user=authors.get(user,user)
  41     if user_re.match(user)==None:
  42       if '@' not in user:
  43         return user+' <none@none>'
  44       return user+' <'+user+'>'
  45     return user
  46   node=repo.lookup(revision)
  47   (manifest,user,(time,timezone),files,desc,extra)=repo.changelog.read(node)
  48   tz="%+03d%02d" % (-timezone / 3600, ((-timezone % 3600) / 60))
  49   branch=get_branch(extra.get('branch','master'))
  50   return (manifest,fixup_user(user,authors),(time,tz),files,desc,branch,extra)
  51
  52 def gitmode(x):
  53   return x and '100755' or '100644'
  54
  55 def wr(msg=''):
  56   print msg
  57   #map(lambda x: sys.stderr.write('\t[%s]\n' % x),msg.split('\n'))
  58
  59 def checkpoint(count):
  60   count=count+1
  61   if cfg_checkpoint_count>0 and count%cfg_checkpoint_count==0:
  62     sys.stderr.write("Checkpoint after %d commits\n" % count)
  63     wr('checkpoint')
  64     wr()
  65   return count
  66
  67 def get_parent_mark(parent,marks):
  68   """Get the mark for some parent.
  69   If we saw it in the current session, return :%d syntax and
  70   otherwise the SHA1 from the cache."""
  71   return marks.get(str(parent+1),':%d' % (parent+1))
  72
  73 def mismatch(f1,f2):
  74   """See if two revisions of a file are not equal."""
  75   return node.hex(f1)!=node.hex(f2)
  76
  77 def outer_set(dleft,dright,l,c,r):
  78   """Loop over our repository and find all changed and missing files."""
  79   for left in dleft.keys():
  80     right=dright.get(left,None)
  81     if right==None:
  82       # we have the file but our parent hasn't: add to left set
  83       l.append(left)
  84     elif mismatch(dleft[left],right):
  85       # we have it but checksums mismatch: add to center set
  86       c.append(left)
  87   for right in dright.keys():
  88     left=dleft.get(right,None)
  89     if left==None:
  90       # if parent has file but we don't: add to right set
  91       r.append(right)
  92     # change is already handled when comparing child against parent
  93   return l,c,r
  94
  95 def get_filechanges(repo,revision,parents,mleft):
  96   """Given some repository and revision, find all changed/deleted files."""
  97   l,c,r=[],[],[]
  98   for p in parents:
  99     if p<0: continue
 100     mright=repo.changectx(p).manifest()
 101     dleft=mleft.keys()
 102     dleft.sort()
 103     dright=mright.keys()
 104     dright.sort()
 105     l,c,r=outer_set(mleft,mright,l,c,r)
 106   return l,c,r
 107
 108 def export_commit(ui,repo,revision,marks,heads,last,max,count,authors):
 109   (_,user,(time,timezone),files,desc,branch,_)=get_changeset(ui,repo,revision,authors)
 110   parents=repo.changelog.parentrevs(revision)
 111
 112   wr('commit refs/heads/%s' % branch)
 113   wr('mark :%d' % (revision+1))
 114   wr('committer %s %d %s' % (user,time,timezone))
 115   wr('data %d' % (len(desc)+1)) # wtf?
 116   wr(desc)
 117   wr()
 118
 119   src=heads.get(branch,'')
 120   link=''
 121   if src!='':
 122     # if we have a cached head, this is an incremental import: initialize it
 123     # and kill reference so we won't init it again
 124     wr('from %s' % src)
 125     heads[branch]=''
 126     sys.stderr.write('Initializing branch [%s] to parent [%s]\n' %
 127         (branch,src))
 128     link=src # avoid making a merge commit for incremental import
 129   elif link=='' and not heads.has_key(branch) and revision>0:
 130     # newly created branch and not the first one: connect to parent
 131     tmp=get_parent_mark(parents[0],marks)
 132     wr('from %s' % tmp)
 133     sys.stderr.write('Link new branch [%s] to parent [%s]\n' %
 134         (branch,tmp))
 135     link=tmp # avoid making a merge commit for branch fork
 136
 137   if parents:
 138     l=last.get(branch,revision)
 139     for p in parents:
 140       # 1) as this commit implicitely is the child of the most recent
 141       #    commit of this branch, ignore this parent
 142       # 2) ignore nonexistent parents
 143       # 3) merge otherwise
 144       if p==l or p==revision or p<0:
 145         continue
 146       tmp=get_parent_mark(p,marks)
 147       # if we fork off a branch, don't merge with our parent via 'merge'
 148       # as we have 'from' already above
 149       if tmp==link:
 150         continue
 151       sys.stderr.write('Merging branch [%s] with parent [%s] from [r%d]\n' %
 152           (branch,tmp,p))
 153       wr('merge %s' % tmp)
 154
 155   last[branch]=revision
 156   heads[branch]=''
 157   # we need this later to write out tags
 158   marks[str(revision)]=':%d'%(revision+1)
 159
 160   ctx=repo.changectx(str(revision))
 161   man=ctx.manifest()
 162   added,changed,removed=get_filechanges(repo,revision,parents,man)
 163
 164   sys.stderr.write('Exporting revision %d with %d/%d/%d added/changed/removed files\n' %
 165       (revision,len(added),len(changed),len(removed)))
 166
 167   for a in added+changed:
 168     fctx=ctx.filectx(a)
 169     d=fctx.data()
 170     wr('M %s inline %s' % (gitmode(man.execf(a)),a))
 171     wr('data %d' % len(d)) # had some trouble with size()
 172     wr(d)
 173
 174   for r in removed:
 175     wr('D %s' % r)
 176
 177   wr()
 178   return checkpoint(count)
 179
 180 def export_tags(ui,repo,marks_cache,start,end,count,authors):
 181   l=repo.tagslist()
 182   for tag,node in l:
 183     # ignore latest revision
 184     if tag=='tip': continue
 185     rev=repo.changelog.rev(node)
 186     # ignore those tags not in our import range
 187     if rev<start or rev>=end: continue
 188
 189     ref=marks_cache.get(str(rev),None)
 190     if ref==None:
 191       sys.stderr.write('Failed to find reference for creating tag'
 192           ' %s at r%d\n' % (tag,rev))
 193       continue
 194     (_,user,(time,timezone),_,desc,branch,_)=get_changeset(ui,repo,rev,authors)
 195     sys.stderr.write('Exporting tag [%s] at [hg r%d] [git %s]\n' % (tag,rev,ref))
 196     wr('tag %s' % tag)
 197     wr('from %s' % ref)
 198     wr('tagger %s %d %s' % (user,time,timezone))
 199     msg='hg2git created tag %s for hg revision %d on branch %s on (summary):\n\t%s' % (tag,
 200         rev,branch,desc.split('\n')[0])
 201     wr('data %d' % (len(msg)+1))
 202     wr(msg)
 203     wr()
 204     count=checkpoint(count)
 205   return count
 206
 207 def load_cache(filename):
 208   cache={}
 209   if not os.path.exists(filename):
 210     return cache
 211   f=open(filename,'r')
 212   l=0
 213   for line in f.readlines():
 214     l+=1
 215     fields=line.split(' ')
 216     if fields==None or not len(fields)==2 or fields[0][0]!=':':
 217       sys.stderr.write('Invalid file format in [%s], line %d\n' % (filename,l))
 218       continue
 219     # put key:value in cache, key without ^:
 220     cache[fields[0][1:]]=fields[1].split('\n')[0]
 221   f.close()
 222   return cache
 223
 224 def save_cache(filename,cache):
 225   f=open(filename,'w+')
 226   map(lambda x: f.write(':%s %s\n' % (str(x),str(cache.get(x)))),cache.keys())
 227   f.close()
 228
 229 def verify_heads(ui,repo,cache):
 230   def getsha1(branch):
 231     f=open(os.getenv('GIT_DIR','/dev/null')+'/refs/heads/'+branch)
 232     sha1=f.readlines()[0].split('\n')[0]
 233     f.close()
 234     return sha1
 235
 236   for b in cache.keys():
 237     sys.stderr.write('Verifying branch [%s]\n' % b)
 238     sha1=getsha1(b)
 239     c=cache.get(b)
 240     if sha1!=c:
 241       sys.stderr.write('Warning: Branch [%s] modified outside hg2git:'
 242         '\n%s (repo) != %s (cache)\n' % (b,sha1,c))
 243   return True
 244
 245 def hg2git(repourl,m,marksfile,headsfile,tipfile,authors={}):
 246   _max=int(m)
 247
 248   marks_cache=load_cache(marksfile)
 249   heads_cache=load_cache(headsfile)
 250   state_cache=load_cache(tipfile)
 251
 252   ui,repo=setup_repo(repourl)
 253
 254   if not verify_heads(ui,repo,heads_cache):
 255     return 1
 256
 257   tip=repo.changelog.count()
 258
 259   min=int(state_cache.get('tip',0))
 260   max=_max
 261   if _max<0:
 262     max=tip
 263
 264   c=0
 265   last={}
 266   for rev in range(min,max):
 267     c=export_commit(ui,repo,rev,marks_cache,heads_cache,last,tip,c,authors)
 268
 269   c=export_tags(ui,repo,marks_cache,min,max,c,authors)
 270
 271   sys.stderr.write('Issued %d commands\n' % c)
 272
 273   state_cache['tip']=max
 274   state_cache['repo']=repourl
 275   save_cache(tipfile,state_cache)
 276
 277   return 0
 278
 279 if __name__=='__main__':
 280   if len(sys.argv)!=6: sys.exit(usage(1))
 281   repourl,m,marksfile,headsfile,tipfile=sys.argv[1:]
 282   sys.exit(hg2git(repourl,m,marksfile,headsfile,tipfile))