1#!/usr/bin/env python
2#
3#===- git-clang-format - ClangFormat Git Integration ---------*- python -*--===#
4#
5#                     The LLVM Compiler Infrastructure
6#
7# This file is distributed under the University of Illinois Open Source
8# License. See LICENSE.TXT for details.
9#
10#===------------------------------------------------------------------------===#
11
12r"""
13clang-format git integration
14============================
15
16This file provides a clang-format integration for git. Put it somewhere in your
17path and ensure that it is executable. Then, "git clang-format" will invoke
18clang-format on the changes in current files or a specific commit.
19
20For further details, run:
21git clang-format -h
22
23Requires Python 2.7
24"""
25
26import argparse
27import collections
28import contextlib
29import errno
30import os
31import re
32import subprocess
33import sys
34
35usage = 'git clang-format [OPTIONS] [<commit>] [--] [<file>...]'
36
37desc = '''
38Run clang-format on all lines that differ between the working directory
39and <commit>, which defaults to HEAD.  Changes are only applied to the working
40directory.
41
42The following git-config settings set the default of the corresponding option:
43  clangFormat.binary
44  clangFormat.commit
45  clangFormat.extension
46  clangFormat.style
47'''
48
49# Name of the temporary index file in which save the output of clang-format.
50# This file is created within the .git directory.
51temp_index_basename = 'clang-format-index'
52
53
54Range = collections.namedtuple('Range', 'start, count')
55
56
57def main():
58  config = load_git_config()
59
60  # In order to keep '--' yet allow options after positionals, we need to
61  # check for '--' ourselves.  (Setting nargs='*' throws away the '--', while
62  # nargs=argparse.REMAINDER disallows options after positionals.)
63  argv = sys.argv[1:]
64  try:
65    idx = argv.index('--')
66  except ValueError:
67    dash_dash = []
68  else:
69    dash_dash = argv[idx:]
70    argv = argv[:idx]
71
72  default_extensions = ','.join([
73      # From clang/lib/Frontend/FrontendOptions.cpp, all lower case
74      'c', 'h',  # C
75      'm',  # ObjC
76      'mm',  # ObjC++
77      'cc', 'cp', 'cpp', 'c++', 'cxx', 'hpp',  # C++
78      # Other languages that clang-format supports
79      'proto', 'protodevel',  # Protocol Buffers
80      'js',  # JavaScript
81      'ts',  # TypeScript
82      ])
83
84  p = argparse.ArgumentParser(
85    usage=usage, formatter_class=argparse.RawDescriptionHelpFormatter,
86    description=desc)
87  p.add_argument('--binary',
88                 default=config.get('clangformat.binary', 'clang-format'),
89                 help='path to clang-format'),
90  p.add_argument('--commit',
91                 default=config.get('clangformat.commit', 'HEAD'),
92                 help='default commit to use if none is specified'),
93  p.add_argument('--diff', action='store_true',
94                 help='print a diff instead of applying the changes')
95  p.add_argument('--extensions',
96                 default=config.get('clangformat.extensions',
97                                    default_extensions),
98                 help=('comma-separated list of file extensions to format, '
99                       'excluding the period and case-insensitive')),
100  p.add_argument('-f', '--force', action='store_true',
101                 help='allow changes to unstaged files')
102  p.add_argument('-p', '--patch', action='store_true',
103                 help='select hunks interactively')
104  p.add_argument('-q', '--quiet', action='count', default=0,
105                 help='print less information')
106  p.add_argument('--style',
107                 default=config.get('clangformat.style', None),
108                 help='passed to clang-format'),
109  p.add_argument('-v', '--verbose', action='count', default=0,
110                 help='print extra information')
111  # We gather all the remaining positional arguments into 'args' since we need
112  # to use some heuristics to determine whether or not <commit> was present.
113  # However, to print pretty messages, we make use of metavar and help.
114  p.add_argument('args', nargs='*', metavar='<commit>',
115                 help='revision from which to compute the diff')
116  p.add_argument('ignored', nargs='*', metavar='<file>...',
117                 help='if specified, only consider differences in these files')
118  opts = p.parse_args(argv)
119
120  opts.verbose -= opts.quiet
121  del opts.quiet
122
123  commit, files = interpret_args(opts.args, dash_dash, opts.commit)
124  changed_lines = compute_diff_and_extract_lines(commit, files)
125  if opts.verbose >= 1:
126    ignored_files = set(changed_lines)
127  filter_by_extension(changed_lines, opts.extensions.lower().split(','))
128  if opts.verbose >= 1:
129    ignored_files.difference_update(changed_lines)
130    if ignored_files:
131      print 'Ignoring changes in the following files (wrong extension):'
132      for filename in ignored_files:
133        print '   ', filename
134    if changed_lines:
135      print 'Running clang-format on the following files:'
136      for filename in changed_lines:
137        print '   ', filename
138  if not changed_lines:
139    print 'no modified files to format'
140    return
141  # The computed diff outputs absolute paths, so we must cd before accessing
142  # those files.
143  cd_to_toplevel()
144  old_tree = create_tree_from_workdir(changed_lines)
145  new_tree = run_clang_format_and_save_to_tree(changed_lines,
146                                               binary=opts.binary,
147                                               style=opts.style)
148  if opts.verbose >= 1:
149    print 'old tree:', old_tree
150    print 'new tree:', new_tree
151  if old_tree == new_tree:
152    if opts.verbose >= 0:
153      print 'clang-format did not modify any files'
154  elif opts.diff:
155    print_diff(old_tree, new_tree)
156  else:
157    changed_files = apply_changes(old_tree, new_tree, force=opts.force,
158                                  patch_mode=opts.patch)
159    if (opts.verbose >= 0 and not opts.patch) or opts.verbose >= 1:
160      print 'changed files:'
161      for filename in changed_files:
162        print '   ', filename
163
164
165def load_git_config(non_string_options=None):
166  """Return the git configuration as a dictionary.
167
168  All options are assumed to be strings unless in `non_string_options`, in which
169  is a dictionary mapping option name (in lower case) to either "--bool" or
170  "--int"."""
171  if non_string_options is None:
172    non_string_options = {}
173  out = {}
174  for entry in run('git', 'config', '--list', '--null').split('\0'):
175    if entry:
176      name, value = entry.split('\n', 1)
177      if name in non_string_options:
178        value = run('git', 'config', non_string_options[name], name)
179      out[name] = value
180  return out
181
182
183def interpret_args(args, dash_dash, default_commit):
184  """Interpret `args` as "[commit] [--] [files...]" and return (commit, files).
185
186  It is assumed that "--" and everything that follows has been removed from
187  args and placed in `dash_dash`.
188
189  If "--" is present (i.e., `dash_dash` is non-empty), the argument to its
190  left (if present) is taken as commit.  Otherwise, the first argument is
191  checked if it is a commit or a file.  If commit is not given,
192  `default_commit` is used."""
193  if dash_dash:
194    if len(args) == 0:
195      commit = default_commit
196    elif len(args) > 1:
197      die('at most one commit allowed; %d given' % len(args))
198    else:
199      commit = args[0]
200    object_type = get_object_type(commit)
201    if object_type not in ('commit', 'tag'):
202      if object_type is None:
203        die("'%s' is not a commit" % commit)
204      else:
205        die("'%s' is a %s, but a commit was expected" % (commit, object_type))
206    files = dash_dash[1:]
207  elif args:
208    if disambiguate_revision(args[0]):
209      commit = args[0]
210      files = args[1:]
211    else:
212      commit = default_commit
213      files = args
214  else:
215    commit = default_commit
216    files = []
217  return commit, files
218
219
220def disambiguate_revision(value):
221  """Returns True if `value` is a revision, False if it is a file, or dies."""
222  # If `value` is ambiguous (neither a commit nor a file), the following
223  # command will die with an appropriate error message.
224  run('git', 'rev-parse', value, verbose=False)
225  object_type = get_object_type(value)
226  if object_type is None:
227    return False
228  if object_type in ('commit', 'tag'):
229    return True
230  die('`%s` is a %s, but a commit or filename was expected' %
231      (value, object_type))
232
233
234def get_object_type(value):
235  """Returns a string description of an object's type, or None if it is not
236  a valid git object."""
237  cmd = ['git', 'cat-file', '-t', value]
238  p = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
239  stdout, stderr = p.communicate()
240  if p.returncode != 0:
241    return None
242  return stdout.strip()
243
244
245def compute_diff_and_extract_lines(commit, files):
246  """Calls compute_diff() followed by extract_lines()."""
247  diff_process = compute_diff(commit, files)
248  changed_lines = extract_lines(diff_process.stdout)
249  diff_process.stdout.close()
250  diff_process.wait()
251  if diff_process.returncode != 0:
252    # Assume error was already printed to stderr.
253    sys.exit(2)
254  return changed_lines
255
256
257def compute_diff(commit, files):
258  """Return a subprocess object producing the diff from `commit`.
259
260  The return value's `stdin` file object will produce a patch with the
261  differences between the working directory and `commit`, filtered on `files`
262  (if non-empty).  Zero context lines are used in the patch."""
263  cmd = ['git', 'diff-index', '-p', '-U0', commit, '--']
264  cmd.extend(files)
265  p = subprocess.Popen(cmd, stdin=subprocess.PIPE, stdout=subprocess.PIPE)
266  p.stdin.close()
267  return p
268
269
270def extract_lines(patch_file):
271  """Extract the changed lines in `patch_file`.
272
273  The return value is a dictionary mapping filename to a list of (start_line,
274  line_count) pairs.
275
276  The input must have been produced with ``-U0``, meaning unidiff format with
277  zero lines of context.  The return value is a dict mapping filename to a
278  list of line `Range`s."""
279  matches = {}
280  for line in patch_file:
281    match = re.search(r'^\+\+\+\ [^/]+/(.*)', line)
282    if match:
283      filename = match.group(1).rstrip('\r\n')
284    match = re.search(r'^@@ -[0-9,]+ \+(\d+)(,(\d+))?', line)
285    if match:
286      start_line = int(match.group(1))
287      line_count = 1
288      if match.group(3):
289        line_count = int(match.group(3))
290      if line_count > 0:
291        matches.setdefault(filename, []).append(Range(start_line, line_count))
292  return matches
293
294
295def filter_by_extension(dictionary, allowed_extensions):
296  """Delete every key in `dictionary` that doesn't have an allowed extension.
297
298  `allowed_extensions` must be a collection of lowercase file extensions,
299  excluding the period."""
300  allowed_extensions = frozenset(allowed_extensions)
301  for filename in dictionary.keys():
302    base_ext = filename.rsplit('.', 1)
303    if len(base_ext) == 1 or base_ext[1].lower() not in allowed_extensions:
304      del dictionary[filename]
305
306
307def cd_to_toplevel():
308  """Change to the top level of the git repository."""
309  toplevel = run('git', 'rev-parse', '--show-toplevel')
310  os.chdir(toplevel)
311
312
313def create_tree_from_workdir(filenames):
314  """Create a new git tree with the given files from the working directory.
315
316  Returns the object ID (SHA-1) of the created tree."""
317  return create_tree(filenames, '--stdin')
318
319
320def run_clang_format_and_save_to_tree(changed_lines, binary='clang-format',
321                                      style=None):
322  """Run clang-format on each file and save the result to a git tree.
323
324  Returns the object ID (SHA-1) of the created tree."""
325  def index_info_generator():
326    for filename, line_ranges in changed_lines.iteritems():
327      mode = oct(os.stat(filename).st_mode)
328      blob_id = clang_format_to_blob(filename, line_ranges, binary=binary,
329                                     style=style)
330      yield '%s %s\t%s' % (mode, blob_id, filename)
331  return create_tree(index_info_generator(), '--index-info')
332
333
334def create_tree(input_lines, mode):
335  """Create a tree object from the given input.
336
337  If mode is '--stdin', it must be a list of filenames.  If mode is
338  '--index-info' is must be a list of values suitable for "git update-index
339  --index-info", such as "<mode> <SP> <sha1> <TAB> <filename>".  Any other mode
340  is invalid."""
341  assert mode in ('--stdin', '--index-info')
342  cmd = ['git', 'update-index', '--add', '-z', mode]
343  with temporary_index_file():
344    p = subprocess.Popen(cmd, stdin=subprocess.PIPE)
345    for line in input_lines:
346      p.stdin.write('%s\0' % line)
347    p.stdin.close()
348    if p.wait() != 0:
349      die('`%s` failed' % ' '.join(cmd))
350    tree_id = run('git', 'write-tree')
351    return tree_id
352
353
354def clang_format_to_blob(filename, line_ranges, binary='clang-format',
355                         style=None):
356  """Run clang-format on the given file and save the result to a git blob.
357
358  Returns the object ID (SHA-1) of the created blob."""
359  clang_format_cmd = [binary, filename]
360  if style:
361    clang_format_cmd.extend(['-style='+style])
362  clang_format_cmd.extend([
363      '-lines=%s:%s' % (start_line, start_line+line_count-1)
364      for start_line, line_count in line_ranges])
365  try:
366    clang_format = subprocess.Popen(clang_format_cmd, stdin=subprocess.PIPE,
367                                    stdout=subprocess.PIPE)
368  except OSError as e:
369    if e.errno == errno.ENOENT:
370      die('cannot find executable "%s"' % binary)
371    else:
372      raise
373  clang_format.stdin.close()
374  hash_object_cmd = ['git', 'hash-object', '-w', '--path='+filename, '--stdin']
375  hash_object = subprocess.Popen(hash_object_cmd, stdin=clang_format.stdout,
376                                 stdout=subprocess.PIPE)
377  clang_format.stdout.close()
378  stdout = hash_object.communicate()[0]
379  if hash_object.returncode != 0:
380    die('`%s` failed' % ' '.join(hash_object_cmd))
381  if clang_format.wait() != 0:
382    die('`%s` failed' % ' '.join(clang_format_cmd))
383  return stdout.rstrip('\r\n')
384
385
386@contextlib.contextmanager
387def temporary_index_file(tree=None):
388  """Context manager for setting GIT_INDEX_FILE to a temporary file and deleting
389  the file afterward."""
390  index_path = create_temporary_index(tree)
391  old_index_path = os.environ.get('GIT_INDEX_FILE')
392  os.environ['GIT_INDEX_FILE'] = index_path
393  try:
394    yield
395  finally:
396    if old_index_path is None:
397      del os.environ['GIT_INDEX_FILE']
398    else:
399      os.environ['GIT_INDEX_FILE'] = old_index_path
400    os.remove(index_path)
401
402
403def create_temporary_index(tree=None):
404  """Create a temporary index file and return the created file's path.
405
406  If `tree` is not None, use that as the tree to read in.  Otherwise, an
407  empty index is created."""
408  gitdir = run('git', 'rev-parse', '--git-dir')
409  path = os.path.join(gitdir, temp_index_basename)
410  if tree is None:
411    tree = '--empty'
412  run('git', 'read-tree', '--index-output='+path, tree)
413  return path
414
415
416def print_diff(old_tree, new_tree):
417  """Print the diff between the two trees to stdout."""
418  # We use the porcelain 'diff' and not plumbing 'diff-tree' because the output
419  # is expected to be viewed by the user, and only the former does nice things
420  # like color and pagination.
421  subprocess.check_call(['git', 'diff', old_tree, new_tree, '--'])
422
423
424def apply_changes(old_tree, new_tree, force=False, patch_mode=False):
425  """Apply the changes in `new_tree` to the working directory.
426
427  Bails if there are local changes in those files and not `force`.  If
428  `patch_mode`, runs `git checkout --patch` to select hunks interactively."""
429  changed_files = run('git', 'diff-tree', '-r', '-z', '--name-only', old_tree,
430                      new_tree).rstrip('\0').split('\0')
431  if not force:
432    unstaged_files = run('git', 'diff-files', '--name-status', *changed_files)
433    if unstaged_files:
434      print >>sys.stderr, ('The following files would be modified but '
435                           'have unstaged changes:')
436      print >>sys.stderr, unstaged_files
437      print >>sys.stderr, 'Please commit, stage, or stash them first.'
438      sys.exit(2)
439  if patch_mode:
440    # In patch mode, we could just as well create an index from the new tree
441    # and checkout from that, but then the user will be presented with a
442    # message saying "Discard ... from worktree".  Instead, we use the old
443    # tree as the index and checkout from new_tree, which gives the slightly
444    # better message, "Apply ... to index and worktree".  This is not quite
445    # right, since it won't be applied to the user's index, but oh well.
446    with temporary_index_file(old_tree):
447      subprocess.check_call(['git', 'checkout', '--patch', new_tree])
448    index_tree = old_tree
449  else:
450    with temporary_index_file(new_tree):
451      run('git', 'checkout-index', '-a', '-f')
452  return changed_files
453
454
455def run(*args, **kwargs):
456  stdin = kwargs.pop('stdin', '')
457  verbose = kwargs.pop('verbose', True)
458  strip = kwargs.pop('strip', True)
459  for name in kwargs:
460    raise TypeError("run() got an unexpected keyword argument '%s'" % name)
461  p = subprocess.Popen(args, stdout=subprocess.PIPE, stderr=subprocess.PIPE,
462                       stdin=subprocess.PIPE)
463  stdout, stderr = p.communicate(input=stdin)
464  if p.returncode == 0:
465    if stderr:
466      if verbose:
467        print >>sys.stderr, '`%s` printed to stderr:' % ' '.join(args)
468      print >>sys.stderr, stderr.rstrip()
469    if strip:
470      stdout = stdout.rstrip('\r\n')
471    return stdout
472  if verbose:
473    print >>sys.stderr, '`%s` returned %s' % (' '.join(args), p.returncode)
474  if stderr:
475    print >>sys.stderr, stderr.rstrip()
476  sys.exit(2)
477
478
479def die(message):
480  print >>sys.stderr, 'error:', message
481  sys.exit(2)
482
483
484if __name__ == '__main__':
485  main()
486