mdw@git.distorted.org.uk Git - stgit/blob - stgit/git.py

   1 """Python GIT interface
   2 """
   3
   4 __copyright__ = """
   5 Copyright (C) 2005, Catalin Marinas <catalin.marinas@gmail.com>
   6
   7 This program is free software; you can redistribute it and/or modify
   8 it under the terms of the GNU General Public License version 2 as
   9 published by the Free Software Foundation.
  10
  11 This program is distributed in the hope that it will be useful,
  12 but WITHOUT ANY WARRANTY; without even the implied warranty of
  13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14 GNU General Public License for more details.
  15
  16 You should have received a copy of the GNU General Public License
  17 along with this program; if not, write to the Free Software
  18 Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
  19 """
  20
  21 import sys, os, popen2, re, gitmergeonefile
  22
  23 from stgit import basedir
  24 from stgit.utils import *
  25
  26 # git exception class
  27 class GitException(Exception):
  28     pass
  29
  30
  31
  32 #
  33 # Classes
  34 #
  35 class Commit:
  36     """Handle the commit objects
  37     """
  38     def __init__(self, id_hash):
  39         self.__id_hash = id_hash
  40
  41         lines = _output_lines('git-cat-file commit %s' % id_hash)
  42         for i in range(len(lines)):
  43             line = lines[i]
  44             if line == '\n':
  45                 break
  46             field = line.strip().split(' ', 1)
  47             if field[0] == 'tree':
  48                 self.__tree = field[1]
  49             if field[0] == 'author':
  50                 self.__author = field[1]
  51             if field[0] == 'committer':
  52                 self.__committer = field[1]
  53         self.__log = ''.join(lines[i+1:])
  54
  55     def get_id_hash(self):
  56         return self.__id_hash
  57
  58     def get_tree(self):
  59         return self.__tree
  60
  61     def get_parent(self):
  62         return self.get_parents()[0]
  63
  64     def get_parents(self):
  65         return _output_lines('git-rev-list --parents --max-count=1 %s'
  66                              % self.__id_hash)[0].split()[1:]
  67
  68     def get_author(self):
  69         return self.__author
  70
  71     def get_committer(self):
  72         return self.__committer
  73
  74     def get_log(self):
  75         return self.__log
  76
  77 # dictionary of Commit objects, used to avoid multiple calls to git
  78 __commits = dict()
  79
  80 #
  81 # Functions
  82 #
  83
  84 def get_commit(id_hash):
  85     """Commit objects factory. Save/look-up them in the __commits
  86     dictionary
  87     """
  88     global __commits
  89
  90     if id_hash in __commits:
  91         return __commits[id_hash]
  92     else:
  93         commit = Commit(id_hash)
  94         __commits[id_hash] = commit
  95         return commit
  96
  97 def get_conflicts():
  98     """Return the list of file conflicts
  99     """
 100     conflicts_file = os.path.join(basedir.get(), 'conflicts')
 101     if os.path.isfile(conflicts_file):
 102         f = file(conflicts_file)
 103         names = [line.strip() for line in f.readlines()]
 104         f.close()
 105         return names
 106     else:
 107         return None
 108
 109 def _input(cmd, file_desc):
 110     p = popen2.Popen3(cmd, True)
 111     while True:
 112         line = file_desc.readline()
 113         if not line:
 114             break
 115         p.tochild.write(line)
 116     p.tochild.close()
 117     if p.wait():
 118         raise GitException, '%s failed (%s)' % (str(cmd),
 119                                                 p.childerr.read().strip())
 120
 121 def _input_str(cmd, string):
 122     p = popen2.Popen3(cmd, True)
 123     p.tochild.write(string)
 124     p.tochild.close()
 125     if p.wait():
 126         raise GitException, '%s failed (%s)' % (str(cmd),
 127                                                 p.childerr.read().strip())
 128
 129 def _output(cmd):
 130     p=popen2.Popen3(cmd, True)
 131     output = p.fromchild.read()
 132     if p.wait():
 133         raise GitException, '%s failed (%s)' % (str(cmd),
 134                                                 p.childerr.read().strip())
 135     return output
 136
 137 def _output_one_line(cmd, file_desc = None):
 138     p=popen2.Popen3(cmd, True)
 139     if file_desc != None:
 140         for line in file_desc:
 141             p.tochild.write(line)
 142         p.tochild.close()
 143     output = p.fromchild.readline().strip()
 144     if p.wait():
 145         raise GitException, '%s failed (%s)' % (str(cmd),
 146                                                 p.childerr.read().strip())
 147     return output
 148
 149 def _output_lines(cmd):
 150     p=popen2.Popen3(cmd, True)
 151     lines = p.fromchild.readlines()
 152     if p.wait():
 153         raise GitException, '%s failed (%s)' % (str(cmd),
 154                                                 p.childerr.read().strip())
 155     return lines
 156
 157 def __run(cmd, args=None):
 158     """__run: runs cmd using spawnvp.
 159
 160     Runs cmd using spawnvp.  The shell is avoided so it won't mess up
 161     our arguments.  If args is very large, the command is run multiple
 162     times; args is split xargs style: cmd is passed on each
 163     invocation.  Unlike xargs, returns immediately if any non-zero
 164     return code is received.
 165     """
 166
 167     args_l=cmd.split()
 168     if args is None:
 169         args = []
 170     for i in range(0, len(args)+1, 100):
 171         r=os.spawnvp(os.P_WAIT, args_l[0], args_l + args[i:min(i+100, len(args))])
 172     if r:
 173         return r
 174     return 0
 175
 176 def __tree_status(files = None, tree_id = 'HEAD', unknown = False,
 177                   noexclude = True):
 178     """Returns a list of pairs - [status, filename]
 179     """
 180     refresh_index()
 181
 182     if not files:
 183         files = []
 184     cache_files = []
 185
 186     # unknown files
 187     if unknown:
 188         exclude_file = os.path.join(basedir.get(), 'info', 'exclude')
 189         base_exclude = ['--exclude=%s' % s for s in
 190                         ['*.[ao]', '*.pyc', '.*', '*~', '#*', 'TAGS', 'tags']]
 191         base_exclude.append('--exclude-per-directory=.gitignore')
 192
 193         if os.path.exists(exclude_file):
 194             extra_exclude = ['--exclude-from=%s' % exclude_file]
 195         else:
 196             extra_exclude = []
 197         if noexclude:
 198             extra_exclude = base_exclude = []
 199
 200         lines = _output_lines(['git-ls-files', '--others', '--directory']
 201                         + base_exclude + extra_exclude)
 202         cache_files += [('?', line.strip()) for line in lines]
 203
 204     # conflicted files
 205     conflicts = get_conflicts()
 206     if not conflicts:
 207         conflicts = []
 208     cache_files += [('C', filename) for filename in conflicts]
 209
 210     # the rest
 211     for line in _output_lines(['git-diff-index', tree_id, '--'] + files):
 212         fs = tuple(line.rstrip().split(' ',4)[-1].split('\t',1))
 213         if fs[1] not in conflicts:
 214             cache_files.append(fs)
 215
 216     return cache_files
 217
 218 def local_changes():
 219     """Return true if there are local changes in the tree
 220     """
 221     return len(__tree_status()) != 0
 222
 223 # HEAD value cached
 224 __head = None
 225
 226 def get_head():
 227     """Verifies the HEAD and returns the SHA1 id that represents it
 228     """
 229     global __head
 230
 231     if not __head:
 232         __head = rev_parse('HEAD')
 233     return __head
 234
 235 def get_head_file():
 236     """Returns the name of the file pointed to by the HEAD link
 237     """
 238     return strip_prefix('refs/heads/',
 239                         _output_one_line('git-symbolic-ref HEAD'))
 240
 241 def set_head_file(ref):
 242     """Resets HEAD to point to a new ref
 243     """
 244     # head cache flushing is needed since we might have a different value
 245     # in the new head
 246     __clear_head_cache()
 247     if __run('git-symbolic-ref HEAD',
 248              [os.path.join('refs', 'heads', ref)]) != 0:
 249         raise GitException, 'Could not set head to "%s"' % ref
 250
 251 def __set_head(val):
 252     """Sets the HEAD value
 253     """
 254     global __head
 255
 256     if not __head or __head != val:
 257         if __run('git-update-ref HEAD', [val]) != 0:
 258             raise GitException, 'Could not update HEAD to "%s".' % val
 259         __head = val
 260
 261     # only allow SHA1 hashes
 262     assert(len(__head) == 40)
 263
 264 def __clear_head_cache():
 265     """Sets the __head to None so that a re-read is forced
 266     """
 267     global __head
 268
 269     __head = None
 270
 271 def refresh_index():
 272     """Refresh index with stat() information from the working directory.
 273     """
 274     __run('git-update-index -q --unmerged --refresh')
 275
 276 def rev_parse(git_id):
 277     """Parse the string and return a verified SHA1 id
 278     """
 279     try:
 280         return _output_one_line(['git-rev-parse', '--verify', git_id])
 281     except GitException:
 282         raise GitException, 'Unknown revision: %s' % git_id
 283
 284 def branch_exists(branch):
 285     """Existence check for the named branch
 286     """
 287     branch = os.path.join('refs', 'heads', branch)
 288     for line in _output_lines('git-rev-parse --symbolic --all 2>&1'):
 289         if line.strip() == branch:
 290             return True
 291         if re.compile('[ |/]'+branch+' ').search(line):
 292             raise GitException, 'Bogus branch: %s' % line
 293     return False
 294
 295 def create_branch(new_branch, tree_id = None):
 296     """Create a new branch in the git repository
 297     """
 298     if branch_exists(new_branch):
 299         raise GitException, 'Branch "%s" already exists' % new_branch
 300
 301     current_head = get_head()
 302     set_head_file(new_branch)
 303     __set_head(current_head)
 304
 305     # a checkout isn't needed if new branch points to the current head
 306     if tree_id:
 307         switch(tree_id)
 308
 309     if os.path.isfile(os.path.join(basedir.get(), 'MERGE_HEAD')):
 310         os.remove(os.path.join(basedir.get(), 'MERGE_HEAD'))
 311
 312 def switch_branch(new_branch):
 313     """Switch to a git branch
 314     """
 315     global __head
 316
 317     if not branch_exists(new_branch):
 318         raise GitException, 'Branch "%s" does not exist' % new_branch
 319
 320     tree_id = rev_parse(os.path.join('refs', 'heads', new_branch)
 321                         + '^{commit}')
 322     if tree_id != get_head():
 323         refresh_index()
 324         if __run('git-read-tree -u -m', [get_head(), tree_id]) != 0:
 325             raise GitException, 'git-read-tree failed (local changes maybe?)'
 326         __head = tree_id
 327     set_head_file(new_branch)
 328
 329     if os.path.isfile(os.path.join(basedir.get(), 'MERGE_HEAD')):
 330         os.remove(os.path.join(basedir.get(), 'MERGE_HEAD'))
 331
 332 def delete_branch(name):
 333     """Delete a git branch
 334     """
 335     if not branch_exists(name):
 336         raise GitException, 'Branch "%s" does not exist' % name
 337     remove_file_and_dirs(os.path.join(basedir.get(), 'refs', 'heads'),
 338                          name)
 339
 340 def rename_branch(from_name, to_name):
 341     """Rename a git branch
 342     """
 343     if not branch_exists(from_name):
 344         raise GitException, 'Branch "%s" does not exist' % from_name
 345     if branch_exists(to_name):
 346         raise GitException, 'Branch "%s" already exists' % to_name
 347
 348     if get_head_file() == from_name:
 349         set_head_file(to_name)
 350     rename(os.path.join(basedir.get(), 'refs', 'heads'),
 351            from_name, to_name)
 352
 353 def add(names):
 354     """Add the files or recursively add the directory contents
 355     """
 356     # generate the file list
 357     files = []
 358     for i in names:
 359         if not os.path.exists(i):
 360             raise GitException, 'Unknown file or directory: %s' % i
 361
 362         if os.path.isdir(i):
 363             # recursive search. We only add files
 364             for root, dirs, local_files in os.walk(i):
 365                 for name in [os.path.join(root, f) for f in local_files]:
 366                     if os.path.isfile(name):
 367                         files.append(os.path.normpath(name))
 368         elif os.path.isfile(i):
 369             files.append(os.path.normpath(i))
 370         else:
 371             raise GitException, '%s is not a file or directory' % i
 372
 373     if files:
 374         if __run('git-update-index --add --', files):
 375             raise GitException, 'Unable to add file'
 376
 377 def rm(files, force = False):
 378     """Remove a file from the repository
 379     """
 380     if not force:
 381         for f in files:
 382             if os.path.exists(f):
 383                 raise GitException, '%s exists. Remove it first' %f
 384         if files:
 385             __run('git-update-index --remove --', files)
 386     else:
 387         if files:
 388             __run('git-update-index --force-remove --', files)
 389
 390 def update_cache(files = None, force = False):
 391     """Update the cache information for the given files
 392     """
 393     if not files:
 394         files = []
 395
 396     cache_files = __tree_status(files)
 397
 398     # everything is up-to-date
 399     if len(cache_files) == 0:
 400         return False
 401
 402     # check for unresolved conflicts
 403     if not force and [x for x in cache_files
 404                       if x[0] not in ['M', 'N', 'A', 'D']]:
 405         raise GitException, 'Updating cache failed: unresolved conflicts'
 406
 407     # update the cache
 408     add_files = [x[1] for x in cache_files if x[0] in ['N', 'A']]
 409     rm_files =  [x[1] for x in cache_files if x[0] in ['D']]
 410     m_files =   [x[1] for x in cache_files if x[0] in ['M']]
 411
 412     if add_files and __run('git-update-index --add --', add_files) != 0:
 413         raise GitException, 'Failed git-update-index --add'
 414     if rm_files and __run('git-update-index --force-remove --', rm_files) != 0:
 415         raise GitException, 'Failed git-update-index --rm'
 416     if m_files and __run('git-update-index --', m_files) != 0:
 417         raise GitException, 'Failed git-update-index'
 418
 419     return True
 420
 421 def commit(message, files = None, parents = None, allowempty = False,
 422            cache_update = True, tree_id = None,
 423            author_name = None, author_email = None, author_date = None,
 424            committer_name = None, committer_email = None):
 425     """Commit the current tree to repository
 426     """
 427     if not files:
 428         files = []
 429     if not parents:
 430         parents = []
 431
 432     # Get the tree status
 433     if cache_update and parents != []:
 434         changes = update_cache(files)
 435         if not changes and not allowempty:
 436             raise GitException, 'No changes to commit'
 437
 438     # get the commit message
 439     if not message:
 440         message = '\n'
 441     elif message[-1:] != '\n':
 442         message += '\n'
 443
 444     must_switch = True
 445     # write the index to repository
 446     if tree_id == None:
 447         tree_id = _output_one_line('git-write-tree')
 448     else:
 449         must_switch = False
 450
 451     # the commit
 452     cmd = ''
 453     if author_name:
 454         cmd += 'GIT_AUTHOR_NAME="%s" ' % author_name
 455     if author_email:
 456         cmd += 'GIT_AUTHOR_EMAIL="%s" ' % author_email
 457     if author_date:
 458         cmd += 'GIT_AUTHOR_DATE="%s" ' % author_date
 459     if committer_name:
 460         cmd += 'GIT_COMMITTER_NAME="%s" ' % committer_name
 461     if committer_email:
 462         cmd += 'GIT_COMMITTER_EMAIL="%s" ' % committer_email
 463     cmd += 'git-commit-tree %s' % tree_id
 464
 465     # get the parents
 466     for p in parents:
 467         cmd += ' -p %s' % p
 468
 469     commit_id = _output_one_line(cmd, message)
 470     if must_switch:
 471         __set_head(commit_id)
 472
 473     return commit_id
 474
 475 def apply_diff(rev1, rev2, check_index = True, files = None):
 476     """Apply the diff between rev1 and rev2 onto the current
 477     index. This function doesn't need to raise an exception since it
 478     is only used for fast-pushing a patch. If this operation fails,
 479     the pushing would fall back to the three-way merge.
 480     """
 481     if check_index:
 482         index_opt = '--index'
 483     else:
 484         index_opt = ''
 485
 486     if not files:
 487         files = []
 488
 489     diff_str = diff(files, rev1, rev2)
 490     if diff_str:
 491         try:
 492             _input_str('git-apply %s' % index_opt, diff_str)
 493         except GitException:
 494             return False
 495
 496     return True
 497
 498 def merge(base, head1, head2):
 499     """Perform a 3-way merge between base, head1 and head2 into the
 500     local tree
 501     """
 502     refresh_index()
 503     if __run('git-read-tree -u -m --aggressive', [base, head1, head2]) != 0:
 504         raise GitException, 'git-read-tree failed (local changes maybe?)'
 505
 506     # check the index for unmerged entries
 507     files = {}
 508     stages_re = re.compile('^([0-7]+) ([0-9a-f]{40}) ([1-3])\t(.*)$', re.S)
 509
 510     for line in _output('git-ls-files --unmerged --stage -z').split('\0'):
 511         if not line:
 512             continue
 513
 514         mode, hash, stage, path = stages_re.findall(line)[0]
 515
 516         if not path in files:
 517             files[path] = {}
 518             files[path]['1'] = ('', '')
 519             files[path]['2'] = ('', '')
 520             files[path]['3'] = ('', '')
 521
 522         files[path][stage] = (mode, hash)
 523
 524     # merge the unmerged files
 525     errors = False
 526     for path in files:
 527         stages = files[path]
 528         if gitmergeonefile.merge(stages['1'][1], stages['2'][1],
 529                                  stages['3'][1], path, stages['1'][0],
 530                                  stages['2'][0], stages['3'][0]) != 0:
 531             errors = True
 532
 533     if errors:
 534         raise GitException, 'GIT index merging failed (possible conflicts)'
 535
 536 def status(files = None, modified = False, new = False, deleted = False,
 537            conflict = False, unknown = False, noexclude = False):
 538     """Show the tree status
 539     """
 540     if not files:
 541         files = []
 542
 543     cache_files = __tree_status(files, unknown = True, noexclude = noexclude)
 544     all = not (modified or new or deleted or conflict or unknown)
 545
 546     if not all:
 547         filestat = []
 548         if modified:
 549             filestat.append('M')
 550         if new:
 551             filestat.append('A')
 552             filestat.append('N')
 553         if deleted:
 554             filestat.append('D')
 555         if conflict:
 556             filestat.append('C')
 557         if unknown:
 558             filestat.append('?')
 559         cache_files = [x for x in cache_files if x[0] in filestat]
 560
 561     for fs in cache_files:
 562         if all:
 563             print '%s %s' % (fs[0], fs[1])
 564         else:
 565             print '%s' % fs[1]
 566
 567 def diff(files = None, rev1 = 'HEAD', rev2 = None, out_fd = None):
 568     """Show the diff between rev1 and rev2
 569     """
 570     if not files:
 571         files = []
 572
 573     if rev1 and rev2:
 574         diff_str = _output(['git-diff-tree', '-p', rev1, rev2, '--'] + files)
 575     elif rev1 or rev2:
 576         refresh_index()
 577         if rev2:
 578             diff_str = _output(['git-diff-index', '-p', '-R', rev2, '--'] + files)
 579         else:
 580             diff_str = _output(['git-diff-index', '-p', rev1, '--'] + files)
 581     else:
 582         diff_str = ''
 583
 584     if out_fd:
 585         out_fd.write(diff_str)
 586     else:
 587         return diff_str
 588
 589 def diffstat(files = None, rev1 = 'HEAD', rev2 = None):
 590     """Return the diffstat between rev1 and rev2
 591     """
 592     if not files:
 593         files = []
 594
 595     p=popen2.Popen3('git-apply --stat')
 596     diff(files, rev1, rev2, p.tochild)
 597     p.tochild.close()
 598     diff_str = p.fromchild.read().rstrip()
 599     if p.wait():
 600         raise GitException, 'git.diffstat failed'
 601     return diff_str
 602
 603 def files(rev1, rev2):
 604     """Return the files modified between rev1 and rev2
 605     """
 606
 607     result = ''
 608     for line in _output_lines('git-diff-tree -r %s %s' % (rev1, rev2)):
 609         result += '%s %s\n' % tuple(line.rstrip().split(' ',4)[-1].split('\t',1))
 610
 611     return result.rstrip()
 612
 613 def barefiles(rev1, rev2):
 614     """Return the files modified between rev1 and rev2, without status info
 615     """
 616
 617     result = ''
 618     for line in _output_lines('git-diff-tree -r %s %s' % (rev1, rev2)):
 619         result += '%s\n' % line.rstrip().split(' ',4)[-1].split('\t',1)[-1]
 620
 621     return result.rstrip()
 622
 623 def pretty_commit(commit_id = 'HEAD'):
 624     """Return a given commit (log + diff)
 625     """
 626     return _output(['git-diff-tree', '--cc', '--always', '--pretty', '-r',
 627                     commit_id])
 628
 629 def checkout(files = None, tree_id = None, force = False):
 630     """Check out the given or all files
 631     """
 632     if not files:
 633         files = []
 634
 635     if tree_id and __run('git-read-tree --reset', [tree_id]) != 0:
 636         raise GitException, 'Failed git-read-tree --reset %s' % tree_id
 637
 638     checkout_cmd = 'git-checkout-index -q -u'
 639     if force:
 640         checkout_cmd += ' -f'
 641     if len(files) == 0:
 642         checkout_cmd += ' -a'
 643     else:
 644         checkout_cmd += ' --'
 645
 646     if __run(checkout_cmd, files) != 0:
 647         raise GitException, 'Failed git-checkout-index'
 648
 649 def switch(tree_id, keep = False):
 650     """Switch the tree to the given id
 651     """
 652     if not keep:
 653         refresh_index()
 654         if __run('git-read-tree -u -m', [get_head(), tree_id]) != 0:
 655             raise GitException, 'git-read-tree failed (local changes maybe?)'
 656
 657     __set_head(tree_id)
 658
 659 def reset(files = None, tree_id = None, check_out = True):
 660     """Revert the tree changes relative to the given tree_id. It removes
 661     any local changes
 662     """
 663     if not tree_id:
 664         tree_id = get_head()
 665
 666     if check_out:
 667         cache_files = __tree_status(files, tree_id)
 668         # files which were added but need to be removed
 669         rm_files =  [x[1] for x in cache_files if x[0] in ['A']]
 670
 671         checkout(files, tree_id, True)
 672         # checkout doesn't remove files
 673         map(os.remove, rm_files)
 674
 675     # if the reset refers to the whole tree, switch the HEAD as well
 676     if not files:
 677         __set_head(tree_id)
 678
 679 def pull(repository = 'origin', refspec = None):
 680     """Pull changes from the remote repository. At the moment, just
 681     use the 'git-pull' command
 682     """
 683     # 'git-pull' updates the HEAD
 684     __clear_head_cache()
 685
 686     args = [repository]
 687     if refspec:
 688         args.append(refspec)
 689
 690     if __run('git-pull', args) != 0:
 691         raise GitException, 'Failed "git-pull %s"' % repository
 692
 693 def apply_patch(filename = None, base = None):
 694     """Apply a patch onto the current or given index. There must not
 695     be any local changes in the tree, otherwise the command fails
 696     """
 697     def __apply_patch():
 698         if filename:
 699             return __run('git-apply --index', [filename]) == 0
 700         else:
 701             try:
 702                 _input('git-apply --index', sys.stdin)
 703             except GitException:
 704                 return False
 705             return True
 706
 707     if base:
 708         orig_head = get_head()
 709         switch(base)
 710     else:
 711         refresh_index()         # needed since __apply_patch() doesn't do it
 712
 713     if not __apply_patch():
 714         if base:
 715             switch(orig_head)
 716         raise GitException, 'Patch does not apply cleanly'
 717     elif base:
 718         top = commit(message = 'temporary commit used for applying a patch',
 719                      parents = [base])
 720         switch(orig_head)
 721         merge(base, orig_head, top)
 722
 723 def clone(repository, local_dir):
 724     """Clone a remote repository. At the moment, just use the
 725     'git-clone' script
 726     """
 727     if __run('git-clone', [repository, local_dir]) != 0:
 728         raise GitException, 'Failed "git-clone %s %s"' \
 729               % (repository, local_dir)
 730
 731 def modifying_revs(files, base_rev):
 732     """Return the revisions from the list modifying the given files
 733     """
 734     cmd = ['git-rev-list', '%s..' % base_rev, '--']
 735     revs = [line.strip() for line in _output_lines(cmd + files)]
 736
 737     return revs