1
# Copyright (C) 2004, 2005, 2006 Canonical Ltd.
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
25
# compatability - plugins import compare_trees from diff!!!
26
# deprecated as of 0.10
27
from bzrlib.delta import compare_trees
28
from bzrlib.errors import BzrError
29
import bzrlib.errors as errors
31
from bzrlib.patiencediff import unified_diff
32
import bzrlib.patiencediff
33
from bzrlib.symbol_versioning import (deprecated_function,
35
from bzrlib.textfile import check_text_lines
36
from bzrlib.trace import mutter, warning
39
# TODO: Rather than building a changeset object, we should probably
40
# invoke callbacks on an object. That object can either accumulate a
41
# list, write them out directly, etc etc.
43
def internal_diff(old_filename, oldlines, new_filename, newlines, to_file,
44
allow_binary=False, sequence_matcher=None,
45
path_encoding='utf8'):
46
# FIXME: difflib is wrong if there is no trailing newline.
47
# The syntax used by patch seems to be "\ No newline at
48
# end of file" following the last diff line from that
49
# file. This is not trivial to insert into the
50
# unified_diff output and it might be better to just fix
51
# or replace that function.
53
# In the meantime we at least make sure the patch isn't
57
# Special workaround for Python2.3, where difflib fails if
58
# both sequences are empty.
59
if not oldlines and not newlines:
62
if allow_binary is False:
63
check_text_lines(oldlines)
64
check_text_lines(newlines)
66
if sequence_matcher is None:
67
sequence_matcher = bzrlib.patiencediff.PatienceSequenceMatcher
68
ud = unified_diff(oldlines, newlines,
69
fromfile=old_filename.encode(path_encoding),
70
tofile=new_filename.encode(path_encoding),
71
sequencematcher=sequence_matcher)
74
# work-around for difflib being too smart for its own good
75
# if /dev/null is "1,0", patch won't recognize it as /dev/null
77
ud[2] = ud[2].replace('-1,0', '-0,0')
79
ud[2] = ud[2].replace('+1,0', '+0,0')
80
# work around for difflib emitting random spaces after the label
81
ud[0] = ud[0][:-2] + '\n'
82
ud[1] = ud[1][:-2] + '\n'
86
if not line.endswith('\n'):
87
to_file.write("\n\\ No newline at end of file\n")
92
"""Set the env var LANG=C"""
93
os.environ['LANG'] = 'C'
96
def _spawn_external_diff(diffcmd, capture_errors=True):
97
"""Spawn the externall diff process, and return the child handle.
99
:param diffcmd: The command list to spawn
100
:param capture_errors: Capture stderr as well as setting LANG=C.
101
This lets us read and understand the output of diff, and respond
103
:return: A Popen object.
106
preexec_fn = _set_lang_C
107
stderr = subprocess.PIPE
113
pipe = subprocess.Popen(diffcmd,
114
stdin=subprocess.PIPE,
115
stdout=subprocess.PIPE,
117
preexec_fn=preexec_fn)
119
if e.errno == errno.ENOENT:
120
raise errors.NoDiff(str(e))
126
def external_diff(old_filename, oldlines, new_filename, newlines, to_file,
128
"""Display a diff by calling out to the external diff program."""
129
# make sure our own output is properly ordered before the diff
132
oldtmp_fd, old_abspath = tempfile.mkstemp(prefix='bzr-diff-old-')
133
newtmp_fd, new_abspath = tempfile.mkstemp(prefix='bzr-diff-new-')
134
oldtmpf = os.fdopen(oldtmp_fd, 'wb')
135
newtmpf = os.fdopen(newtmp_fd, 'wb')
138
# TODO: perhaps a special case for comparing to or from the empty
139
# sequence; can just use /dev/null on Unix
141
# TODO: if either of the files being compared already exists as a
142
# regular named file (e.g. in the working directory) then we can
143
# compare directly to that, rather than copying it.
145
oldtmpf.writelines(oldlines)
146
newtmpf.writelines(newlines)
154
'--label', old_filename,
156
'--label', new_filename,
161
# diff only allows one style to be specified; they don't override.
162
# note that some of these take optargs, and the optargs can be
163
# directly appended to the options.
164
# this is only an approximate parser; it doesn't properly understand
166
for s in ['-c', '-u', '-C', '-U',
171
'-y', '--side-by-side',
183
diffcmd.extend(diff_opts)
185
pipe = _spawn_external_diff(diffcmd, capture_errors=True)
186
out,err = pipe.communicate()
189
# internal_diff() adds a trailing newline, add one here for consistency
192
# 'diff' gives retcode == 2 for all sorts of errors
193
# one of those is 'Binary files differ'.
194
# Bad options could also be the problem.
195
# 'Binary files' is not a real error, so we suppress that error.
198
# Since we got here, we want to make sure to give an i18n error
199
pipe = _spawn_external_diff(diffcmd, capture_errors=False)
200
out, err = pipe.communicate()
202
# Write out the new i18n diff response
203
to_file.write(out+'\n')
204
if pipe.returncode != 2:
205
raise BzrError('external diff failed with exit code 2'
206
' when run with LANG=C, but not when run'
207
' natively: %r' % (diffcmd,))
209
first_line = lang_c_out.split('\n', 1)[0]
210
# Starting with diffutils 2.8.4 the word "binary" was dropped.
211
m = re.match('^(binary )?files.*differ$', first_line, re.I)
213
raise BzrError('external diff failed with exit code 2;'
214
' command: %r' % (diffcmd,))
216
# Binary files differ, just return
219
# If we got to here, we haven't written out the output of diff
223
# returns 1 if files differ; that's OK
225
msg = 'signal %d' % (-rc)
227
msg = 'exit code %d' % rc
229
raise BzrError('external diff failed with %s; command: %r'
234
oldtmpf.close() # and delete
236
# Clean up. Warn in case the files couldn't be deleted
237
# (in case windows still holds the file open, but not
238
# if the files have already been deleted)
240
os.remove(old_abspath)
242
if e.errno not in (errno.ENOENT,):
243
warning('Failed to delete temporary file: %s %s',
246
os.remove(new_abspath)
248
if e.errno not in (errno.ENOENT,):
249
warning('Failed to delete temporary file: %s %s',
253
@deprecated_function(zero_eight)
254
def show_diff(b, from_spec, specific_files, external_diff_options=None,
255
revision2=None, output=None, b2=None):
256
"""Shortcut for showing the diff to the working tree.
258
Please use show_diff_trees instead.
264
None for 'basis tree', or otherwise the old revision to compare against.
266
The more general form is show_diff_trees(), where the caller
267
supplies any two trees.
272
if from_spec is None:
273
old_tree = b.bzrdir.open_workingtree()
275
old_tree = old_tree = old_tree.basis_tree()
277
old_tree = b.repository.revision_tree(from_spec.in_history(b).rev_id)
279
if revision2 is None:
281
new_tree = b.bzrdir.open_workingtree()
283
new_tree = b2.bzrdir.open_workingtree()
285
new_tree = b.repository.revision_tree(revision2.in_history(b).rev_id)
287
return show_diff_trees(old_tree, new_tree, output, specific_files,
288
external_diff_options)
291
def diff_cmd_helper(tree, specific_files, external_diff_options,
292
old_revision_spec=None, new_revision_spec=None,
293
old_label='a/', new_label='b/'):
294
"""Helper for cmd_diff.
300
The specific files to compare, or None
302
external_diff_options
303
If non-None, run an external diff, and pass it these options
306
If None, use basis tree as old revision, otherwise use the tree for
307
the specified revision.
310
If None, use working tree as new revision, otherwise use the tree for
311
the specified revision.
313
The more general form is show_diff_trees(), where the caller
314
supplies any two trees.
318
revision = spec.in_store(tree.branch)
320
revision = spec.in_store(None)
321
revision_id = revision.rev_id
322
branch = revision.branch
323
return branch.repository.revision_tree(revision_id)
324
if old_revision_spec is None:
325
old_tree = tree.basis_tree()
327
old_tree = spec_tree(old_revision_spec)
329
if new_revision_spec is None:
332
new_tree = spec_tree(new_revision_spec)
333
if new_tree is not tree:
334
extra_trees = (tree,)
338
return show_diff_trees(old_tree, new_tree, sys.stdout, specific_files,
339
external_diff_options,
340
old_label=old_label, new_label=new_label,
341
extra_trees=extra_trees)
344
def show_diff_trees(old_tree, new_tree, to_file, specific_files=None,
345
external_diff_options=None,
346
old_label='a/', new_label='b/',
348
"""Show in text form the changes from one tree to another.
351
If set, include only changes to these files.
353
external_diff_options
354
If set, use an external GNU diff and pass these options.
357
If set, more Trees to use for looking up file ids
363
return _show_diff_trees(old_tree, new_tree, to_file,
364
specific_files, external_diff_options,
365
old_label=old_label, new_label=new_label,
366
extra_trees=extra_trees)
373
def _show_diff_trees(old_tree, new_tree, to_file,
374
specific_files, external_diff_options,
375
old_label='a/', new_label='b/', extra_trees=None):
377
# GNU Patch uses the epoch date to detect files that are being added
378
# or removed in a diff.
379
EPOCH_DATE = '1970-01-01 00:00:00 +0000'
381
# TODO: Generation of pseudo-diffs for added/deleted files could
382
# be usefully made into a much faster special case.
384
if external_diff_options:
385
assert isinstance(external_diff_options, basestring)
386
opts = external_diff_options.split()
387
def diff_file(olab, olines, nlab, nlines, to_file):
388
external_diff(olab, olines, nlab, nlines, to_file, opts)
390
diff_file = internal_diff
392
delta = new_tree.changes_from(old_tree,
393
specific_files=specific_files,
394
extra_trees=extra_trees, require_versioned=True)
397
for path, file_id, kind in delta.removed:
399
print >>to_file, '=== removed %s %r' % (kind, path.encode('utf8'))
400
old_name = '%s%s\t%s' % (old_label, path,
401
_patch_header_date(old_tree, file_id, path))
402
new_name = '%s%s\t%s' % (new_label, path, EPOCH_DATE)
403
old_tree.inventory[file_id].diff(diff_file, old_name, old_tree,
404
new_name, None, None, to_file)
405
for path, file_id, kind in delta.added:
407
print >>to_file, '=== added %s %r' % (kind, path.encode('utf8'))
408
old_name = '%s%s\t%s' % (old_label, path, EPOCH_DATE)
409
new_name = '%s%s\t%s' % (new_label, path,
410
_patch_header_date(new_tree, file_id, path))
411
new_tree.inventory[file_id].diff(diff_file, new_name, new_tree,
412
old_name, None, None, to_file,
414
for (old_path, new_path, file_id, kind,
415
text_modified, meta_modified) in delta.renamed:
417
prop_str = get_prop_change(meta_modified)
418
print >>to_file, '=== renamed %s %r => %r%s' % (
419
kind, old_path.encode('utf8'),
420
new_path.encode('utf8'), prop_str)
421
old_name = '%s%s\t%s' % (old_label, old_path,
422
_patch_header_date(old_tree, file_id,
424
new_name = '%s%s\t%s' % (new_label, new_path,
425
_patch_header_date(new_tree, file_id,
427
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
429
text_modified, kind, to_file, diff_file)
430
for path, file_id, kind, text_modified, meta_modified in delta.modified:
432
prop_str = get_prop_change(meta_modified)
433
print >>to_file, '=== modified %s %r%s' % (kind, path.encode('utf8'), prop_str)
434
old_name = '%s%s\t%s' % (old_label, path,
435
_patch_header_date(old_tree, file_id, path))
436
new_name = '%s%s\t%s' % (new_label, path,
437
_patch_header_date(new_tree, file_id, path))
439
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
441
True, kind, to_file, diff_file)
446
def _patch_header_date(tree, file_id, path):
447
"""Returns a timestamp suitable for use in a patch header."""
448
tm = time.gmtime(tree.get_file_mtime(file_id, path))
449
return time.strftime('%Y-%m-%d %H:%M:%S +0000', tm)
452
def _raise_if_nonexistent(paths, old_tree, new_tree):
453
"""Complain if paths are not in either inventory or tree.
455
It's OK with the files exist in either tree's inventory, or
456
if they exist in the tree but are not versioned.
458
This can be used by operations such as bzr status that can accept
459
unknown or ignored files.
461
mutter("check paths: %r", paths)
464
s = old_tree.filter_unversioned_files(paths)
465
s = new_tree.filter_unversioned_files(s)
466
s = [path for path in s if not new_tree.has_filename(path)]
468
raise errors.PathsDoNotExist(sorted(s))
471
def get_prop_change(meta_modified):
473
return " (properties changed)"
478
def _maybe_diff_file_or_symlink(old_path, old_tree, file_id,
479
new_path, new_tree, text_modified,
480
kind, to_file, diff_file):
482
new_entry = new_tree.inventory[file_id]
483
old_tree.inventory[file_id].diff(diff_file,
added
removed
bzrlib/diff.py
