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
21
from bzrlib.lazy_import import lazy_import
22
lazy_import(globals(), """
37
# compatability - plugins import compare_trees from diff!!!
38
# deprecated as of 0.10
39
from bzrlib.delta import compare_trees
40
from bzrlib.symbol_versioning import (
44
from bzrlib.trace import mutter, warning
47
# TODO: Rather than building a changeset object, we should probably
48
# invoke callbacks on an object. That object can either accumulate a
49
# list, write them out directly, etc etc.
51
def internal_diff(old_filename, oldlines, new_filename, newlines, to_file,
52
allow_binary=False, sequence_matcher=None,
53
path_encoding='utf8'):
54
# FIXME: difflib is wrong if there is no trailing newline.
55
# The syntax used by patch seems to be "\ No newline at
56
# end of file" following the last diff line from that
57
# file. This is not trivial to insert into the
58
# unified_diff output and it might be better to just fix
59
# or replace that function.
61
# In the meantime we at least make sure the patch isn't
65
# Special workaround for Python2.3, where difflib fails if
66
# both sequences are empty.
67
if not oldlines and not newlines:
70
if allow_binary is False:
71
textfile.check_text_lines(oldlines)
72
textfile.check_text_lines(newlines)
74
if sequence_matcher is None:
75
sequence_matcher = patiencediff.PatienceSequenceMatcher
76
ud = patiencediff.unified_diff(oldlines, newlines,
77
fromfile=old_filename.encode(path_encoding),
78
tofile=new_filename.encode(path_encoding),
79
sequencematcher=sequence_matcher)
82
# work-around for difflib being too smart for its own good
83
# if /dev/null is "1,0", patch won't recognize it as /dev/null
85
ud[2] = ud[2].replace('-1,0', '-0,0')
87
ud[2] = ud[2].replace('+1,0', '+0,0')
88
# work around for difflib emitting random spaces after the label
89
ud[0] = ud[0][:-2] + '\n'
90
ud[1] = ud[1][:-2] + '\n'
94
if not line.endswith('\n'):
95
to_file.write("\n\\ No newline at end of file\n")
100
"""Set the env vars LANG=C and LC_ALL=C."""
101
osutils.set_or_unset_env('LANG', 'C')
102
osutils.set_or_unset_env('LC_ALL', 'C')
103
osutils.set_or_unset_env('LC_CTYPE', None)
104
osutils.set_or_unset_env('LANGUAGE', None)
107
def _spawn_external_diff(diffcmd, capture_errors=True):
108
"""Spawn the externall diff process, and return the child handle.
110
:param diffcmd: The command list to spawn
111
:param capture_errors: Capture stderr as well as setting LANG=C
112
and LC_ALL=C. This lets us read and understand the output of diff,
113
and respond to any errors.
114
:return: A Popen object.
117
if sys.platform == 'win32':
118
# Win32 doesn't support preexec_fn, but that is
119
# okay, because it doesn't support LANG and LC_ALL either.
122
preexec_fn = _set_lang_C
123
stderr = subprocess.PIPE
129
pipe = subprocess.Popen(diffcmd,
130
stdin=subprocess.PIPE,
131
stdout=subprocess.PIPE,
133
preexec_fn=preexec_fn)
135
if e.errno == errno.ENOENT:
136
raise errors.NoDiff(str(e))
142
def external_diff(old_filename, oldlines, new_filename, newlines, to_file,
144
"""Display a diff by calling out to the external diff program."""
145
# make sure our own output is properly ordered before the diff
148
oldtmp_fd, old_abspath = tempfile.mkstemp(prefix='bzr-diff-old-')
149
newtmp_fd, new_abspath = tempfile.mkstemp(prefix='bzr-diff-new-')
150
oldtmpf = os.fdopen(oldtmp_fd, 'wb')
151
newtmpf = os.fdopen(newtmp_fd, 'wb')
154
# TODO: perhaps a special case for comparing to or from the empty
155
# sequence; can just use /dev/null on Unix
157
# TODO: if either of the files being compared already exists as a
158
# regular named file (e.g. in the working directory) then we can
159
# compare directly to that, rather than copying it.
161
oldtmpf.writelines(oldlines)
162
newtmpf.writelines(newlines)
170
'--label', old_filename,
172
'--label', new_filename,
177
# diff only allows one style to be specified; they don't override.
178
# note that some of these take optargs, and the optargs can be
179
# directly appended to the options.
180
# this is only an approximate parser; it doesn't properly understand
182
for s in ['-c', '-u', '-C', '-U',
187
'-y', '--side-by-side',
199
diffcmd.extend(diff_opts)
201
pipe = _spawn_external_diff(diffcmd, capture_errors=True)
202
out,err = pipe.communicate()
205
# internal_diff() adds a trailing newline, add one here for consistency
208
# 'diff' gives retcode == 2 for all sorts of errors
209
# one of those is 'Binary files differ'.
210
# Bad options could also be the problem.
211
# 'Binary files' is not a real error, so we suppress that error.
214
# Since we got here, we want to make sure to give an i18n error
215
pipe = _spawn_external_diff(diffcmd, capture_errors=False)
216
out, err = pipe.communicate()
218
# Write out the new i18n diff response
219
to_file.write(out+'\n')
220
if pipe.returncode != 2:
221
raise errors.BzrError(
222
'external diff failed with exit code 2'
223
' when run with LANG=C and LC_ALL=C,'
224
' but not when run natively: %r' % (diffcmd,))
226
first_line = lang_c_out.split('\n', 1)[0]
227
# Starting with diffutils 2.8.4 the word "binary" was dropped.
228
m = re.match('^(binary )?files.*differ$', first_line, re.I)
230
raise errors.BzrError('external diff failed with exit code 2;'
231
' command: %r' % (diffcmd,))
233
# Binary files differ, just return
236
# If we got to here, we haven't written out the output of diff
240
# returns 1 if files differ; that's OK
242
msg = 'signal %d' % (-rc)
244
msg = 'exit code %d' % rc
246
raise errors.BzrError('external diff failed with %s; command: %r'
251
oldtmpf.close() # and delete
253
# Clean up. Warn in case the files couldn't be deleted
254
# (in case windows still holds the file open, but not
255
# if the files have already been deleted)
257
os.remove(old_abspath)
259
if e.errno not in (errno.ENOENT,):
260
warning('Failed to delete temporary file: %s %s',
263
os.remove(new_abspath)
265
if e.errno not in (errno.ENOENT,):
266
warning('Failed to delete temporary file: %s %s',
270
@deprecated_function(zero_eight)
271
def show_diff(b, from_spec, specific_files, external_diff_options=None,
272
revision2=None, output=None, b2=None):
273
"""Shortcut for showing the diff to the working tree.
275
Please use show_diff_trees instead.
281
None for 'basis tree', or otherwise the old revision to compare against.
283
The more general form is show_diff_trees(), where the caller
284
supplies any two trees.
289
if from_spec is None:
290
old_tree = b.bzrdir.open_workingtree()
292
old_tree = old_tree = old_tree.basis_tree()
294
old_tree = b.repository.revision_tree(from_spec.in_history(b).rev_id)
296
if revision2 is None:
298
new_tree = b.bzrdir.open_workingtree()
300
new_tree = b2.bzrdir.open_workingtree()
302
new_tree = b.repository.revision_tree(revision2.in_history(b).rev_id)
304
return show_diff_trees(old_tree, new_tree, output, specific_files,
305
external_diff_options)
308
def diff_cmd_helper(tree, specific_files, external_diff_options,
309
old_revision_spec=None, new_revision_spec=None,
311
old_label='a/', new_label='b/'):
312
"""Helper for cmd_diff.
317
:param specific_files:
318
The specific files to compare, or None
320
:param external_diff_options:
321
If non-None, run an external diff, and pass it these options
323
:param old_revision_spec:
324
If None, use basis tree as old revision, otherwise use the tree for
325
the specified revision.
327
:param new_revision_spec:
328
If None, use working tree as new revision, otherwise use the tree for
329
the specified revision.
331
:param revision_specs:
332
Zero, one or two RevisionSpecs from the command line, saying what revisions
333
to compare. This can be passed as an alternative to the old_revision_spec
334
and new_revision_spec parameters.
336
The more general form is show_diff_trees(), where the caller
337
supplies any two trees.
340
# TODO: perhaps remove the old parameters old_revision_spec and
341
# new_revision_spec, since this is only really for use from cmd_diff and
342
# it now always passes through a sequence of revision_specs -- mbp
347
revision = spec.in_store(tree.branch)
349
revision = spec.in_store(None)
350
revision_id = revision.rev_id
351
branch = revision.branch
352
return branch.repository.revision_tree(revision_id)
354
if revision_specs is not None:
355
assert (old_revision_spec is None
356
and new_revision_spec is None)
357
if len(revision_specs) > 0:
358
old_revision_spec = revision_specs[0]
359
if len(revision_specs) > 1:
360
new_revision_spec = revision_specs[1]
362
if old_revision_spec is None:
363
old_tree = tree.basis_tree()
365
old_tree = spec_tree(old_revision_spec)
367
if (new_revision_spec is None
368
or new_revision_spec.spec is None):
371
new_tree = spec_tree(new_revision_spec)
373
if new_tree is not tree:
374
extra_trees = (tree,)
378
return show_diff_trees(old_tree, new_tree, sys.stdout, specific_files,
379
external_diff_options,
380
old_label=old_label, new_label=new_label,
381
extra_trees=extra_trees)
384
def show_diff_trees(old_tree, new_tree, to_file, specific_files=None,
385
external_diff_options=None,
386
old_label='a/', new_label='b/',
388
"""Show in text form the changes from one tree to another.
391
If set, include only changes to these files.
393
external_diff_options
394
If set, use an external GNU diff and pass these options.
397
If set, more Trees to use for looking up file ids
401
if extra_trees is not None:
402
for tree in extra_trees:
406
return _show_diff_trees(old_tree, new_tree, to_file,
407
specific_files, external_diff_options,
408
old_label=old_label, new_label=new_label,
409
extra_trees=extra_trees)
412
if extra_trees is not None:
413
for tree in extra_trees:
419
def _show_diff_trees(old_tree, new_tree, to_file,
420
specific_files, external_diff_options,
421
old_label='a/', new_label='b/', extra_trees=None):
423
# GNU Patch uses the epoch date to detect files that are being added
424
# or removed in a diff.
425
EPOCH_DATE = '1970-01-01 00:00:00 +0000'
427
# TODO: Generation of pseudo-diffs for added/deleted files could
428
# be usefully made into a much faster special case.
430
if external_diff_options:
431
assert isinstance(external_diff_options, basestring)
432
opts = external_diff_options.split()
433
def diff_file(olab, olines, nlab, nlines, to_file):
434
external_diff(olab, olines, nlab, nlines, to_file, opts)
436
diff_file = internal_diff
438
delta = new_tree.changes_from(old_tree,
439
specific_files=specific_files,
440
extra_trees=extra_trees, require_versioned=True)
443
for path, file_id, kind in delta.removed:
445
print >>to_file, '=== removed %s %r' % (kind, path.encode('utf8'))
446
old_name = '%s%s\t%s' % (old_label, path,
447
_patch_header_date(old_tree, file_id, path))
448
new_name = '%s%s\t%s' % (new_label, path, EPOCH_DATE)
449
old_tree.inventory[file_id].diff(diff_file, old_name, old_tree,
450
new_name, None, None, to_file)
451
for path, file_id, kind in delta.added:
453
print >>to_file, '=== added %s %r' % (kind, path.encode('utf8'))
454
old_name = '%s%s\t%s' % (old_label, path, EPOCH_DATE)
455
new_name = '%s%s\t%s' % (new_label, path,
456
_patch_header_date(new_tree, file_id, path))
457
new_tree.inventory[file_id].diff(diff_file, new_name, new_tree,
458
old_name, None, None, to_file,
460
for (old_path, new_path, file_id, kind,
461
text_modified, meta_modified) in delta.renamed:
463
prop_str = get_prop_change(meta_modified)
464
print >>to_file, '=== renamed %s %r => %r%s' % (
465
kind, old_path.encode('utf8'),
466
new_path.encode('utf8'), prop_str)
467
old_name = '%s%s\t%s' % (old_label, old_path,
468
_patch_header_date(old_tree, file_id,
470
new_name = '%s%s\t%s' % (new_label, new_path,
471
_patch_header_date(new_tree, file_id,
473
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
475
text_modified, kind, to_file, diff_file)
476
for path, file_id, kind, text_modified, meta_modified in delta.modified:
478
prop_str = get_prop_change(meta_modified)
479
print >>to_file, '=== modified %s %r%s' % (kind, path.encode('utf8'), prop_str)
480
old_name = '%s%s\t%s' % (old_label, path,
481
_patch_header_date(old_tree, file_id, path))
482
new_name = '%s%s\t%s' % (new_label, path,
483
_patch_header_date(new_tree, file_id, path))
485
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
487
True, kind, to_file, diff_file)
492
def _patch_header_date(tree, file_id, path):
493
"""Returns a timestamp suitable for use in a patch header."""
494
return timestamp.format_patch_date(tree.get_file_mtime(file_id, path))
497
def _raise_if_nonexistent(paths, old_tree, new_tree):
498
"""Complain if paths are not in either inventory or tree.
500
It's OK with the files exist in either tree's inventory, or
501
if they exist in the tree but are not versioned.
503
This can be used by operations such as bzr status that can accept
504
unknown or ignored files.
506
mutter("check paths: %r", paths)
509
s = old_tree.filter_unversioned_files(paths)
510
s = new_tree.filter_unversioned_files(s)
511
s = [path for path in s if not new_tree.has_filename(path)]
513
raise errors.PathsDoNotExist(sorted(s))
516
def get_prop_change(meta_modified):
518
return " (properties changed)"
523
def _maybe_diff_file_or_symlink(old_path, old_tree, file_id,
524
new_path, new_tree, text_modified,
525
kind, to_file, diff_file):
527
new_entry = new_tree.inventory[file_id]
528
old_tree.inventory[file_id].diff(diff_file,
added
removed
bzrlib/diff.py
