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