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(), """
36
# compatability - plugins import compare_trees from diff!!!
37
# deprecated as of 0.10
38
from bzrlib.delta import compare_trees
39
from bzrlib.symbol_versioning import (
43
from bzrlib.trace import mutter, warning
46
# TODO: Rather than building a changeset object, we should probably
47
# invoke callbacks on an object. That object can either accumulate a
48
# list, write them out directly, etc etc.
50
def internal_diff(old_filename, oldlines, new_filename, newlines, to_file,
51
allow_binary=False, sequence_matcher=None,
52
path_encoding='utf8'):
53
# FIXME: difflib is wrong if there is no trailing newline.
54
# The syntax used by patch seems to be "\ No newline at
55
# end of file" following the last diff line from that
56
# file. This is not trivial to insert into the
57
# unified_diff output and it might be better to just fix
58
# or replace that function.
60
# In the meantime we at least make sure the patch isn't
64
# Special workaround for Python2.3, where difflib fails if
65
# both sequences are empty.
66
if not oldlines and not newlines:
69
if allow_binary is False:
70
textfile.check_text_lines(oldlines)
71
textfile.check_text_lines(newlines)
73
if sequence_matcher is None:
74
sequence_matcher = patiencediff.PatienceSequenceMatcher
75
ud = patiencediff.unified_diff(oldlines, newlines,
76
fromfile=old_filename.encode(path_encoding),
77
tofile=new_filename.encode(path_encoding),
78
sequencematcher=sequence_matcher)
81
# work-around for difflib being too smart for its own good
82
# if /dev/null is "1,0", patch won't recognize it as /dev/null
84
ud[2] = ud[2].replace('-1,0', '-0,0')
86
ud[2] = ud[2].replace('+1,0', '+0,0')
87
# work around for difflib emitting random spaces after the label
88
ud[0] = ud[0][:-2] + '\n'
89
ud[1] = ud[1][:-2] + '\n'
93
if not line.endswith('\n'):
94
to_file.write("\n\\ No newline at end of file\n")
98
def _spawn_external_diff(diffcmd, capture_errors=True):
99
"""Spawn the externall diff process, and return the child handle.
101
:param diffcmd: The command list to spawn
102
:param capture_errors: Capture stderr as well as setting LANG=C
103
and LC_ALL=C. This lets us read and understand the output of diff,
104
and respond to any errors.
105
:return: A Popen object.
108
# construct minimal environment
110
path = os.environ.get('PATH')
113
env['LANGUAGE'] = 'C' # on win32 only LANGUAGE has effect
116
stderr = subprocess.PIPE
122
pipe = subprocess.Popen(diffcmd,
123
stdin=subprocess.PIPE,
124
stdout=subprocess.PIPE,
128
if e.errno == errno.ENOENT:
129
raise errors.NoDiff(str(e))
135
def external_diff(old_filename, oldlines, new_filename, newlines, to_file,
137
"""Display a diff by calling out to the external diff program."""
138
# make sure our own output is properly ordered before the diff
141
oldtmp_fd, old_abspath = tempfile.mkstemp(prefix='bzr-diff-old-')
142
newtmp_fd, new_abspath = tempfile.mkstemp(prefix='bzr-diff-new-')
143
oldtmpf = os.fdopen(oldtmp_fd, 'wb')
144
newtmpf = os.fdopen(newtmp_fd, 'wb')
147
# TODO: perhaps a special case for comparing to or from the empty
148
# sequence; can just use /dev/null on Unix
150
# TODO: if either of the files being compared already exists as a
151
# regular named file (e.g. in the working directory) then we can
152
# compare directly to that, rather than copying it.
154
oldtmpf.writelines(oldlines)
155
newtmpf.writelines(newlines)
163
'--label', old_filename,
165
'--label', new_filename,
170
# diff only allows one style to be specified; they don't override.
171
# note that some of these take optargs, and the optargs can be
172
# directly appended to the options.
173
# this is only an approximate parser; it doesn't properly understand
175
for s in ['-c', '-u', '-C', '-U',
180
'-y', '--side-by-side',
192
diffcmd.extend(diff_opts)
194
pipe = _spawn_external_diff(diffcmd, capture_errors=True)
195
out,err = pipe.communicate()
198
# internal_diff() adds a trailing newline, add one here for consistency
201
# 'diff' gives retcode == 2 for all sorts of errors
202
# one of those is 'Binary files differ'.
203
# Bad options could also be the problem.
204
# 'Binary files' is not a real error, so we suppress that error.
207
# Since we got here, we want to make sure to give an i18n error
208
pipe = _spawn_external_diff(diffcmd, capture_errors=False)
209
out, err = pipe.communicate()
211
# Write out the new i18n diff response
212
to_file.write(out+'\n')
213
if pipe.returncode != 2:
214
raise errors.BzrError(
215
'external diff failed with exit code 2'
216
' when run with LANG=C and LC_ALL=C,'
217
' but not when run natively: %r' % (diffcmd,))
219
first_line = lang_c_out.split('\n', 1)[0]
220
# Starting with diffutils 2.8.4 the word "binary" was dropped.
221
m = re.match('^(binary )?files.*differ$', first_line, re.I)
223
raise errors.BzrError('external diff failed with exit code 2;'
224
' command: %r' % (diffcmd,))
226
# Binary files differ, just return
229
# If we got to here, we haven't written out the output of diff
233
# returns 1 if files differ; that's OK
235
msg = 'signal %d' % (-rc)
237
msg = 'exit code %d' % rc
239
raise errors.BzrError('external diff failed with %s; command: %r'
244
oldtmpf.close() # and delete
246
# Clean up. Warn in case the files couldn't be deleted
247
# (in case windows still holds the file open, but not
248
# if the files have already been deleted)
250
os.remove(old_abspath)
252
if e.errno not in (errno.ENOENT,):
253
warning('Failed to delete temporary file: %s %s',
256
os.remove(new_abspath)
258
if e.errno not in (errno.ENOENT,):
259
warning('Failed to delete temporary file: %s %s',
263
@deprecated_function(zero_eight)
264
def show_diff(b, from_spec, specific_files, external_diff_options=None,
265
revision2=None, output=None, b2=None):
266
"""Shortcut for showing the diff to the working tree.
268
Please use show_diff_trees instead.
274
None for 'basis tree', or otherwise the old revision to compare against.
276
The more general form is show_diff_trees(), where the caller
277
supplies any two trees.
282
if from_spec is None:
283
old_tree = b.bzrdir.open_workingtree()
285
old_tree = old_tree = old_tree.basis_tree()
287
old_tree = b.repository.revision_tree(from_spec.in_history(b).rev_id)
289
if revision2 is None:
291
new_tree = b.bzrdir.open_workingtree()
293
new_tree = b2.bzrdir.open_workingtree()
295
new_tree = b.repository.revision_tree(revision2.in_history(b).rev_id)
297
return show_diff_trees(old_tree, new_tree, output, specific_files,
298
external_diff_options)
301
def diff_cmd_helper(tree, specific_files, external_diff_options,
302
old_revision_spec=None, new_revision_spec=None,
304
old_label='a/', new_label='b/'):
305
"""Helper for cmd_diff.
310
:param specific_files:
311
The specific files to compare, or None
313
:param external_diff_options:
314
If non-None, run an external diff, and pass it these options
316
:param old_revision_spec:
317
If None, use basis tree as old revision, otherwise use the tree for
318
the specified revision.
320
:param new_revision_spec:
321
If None, use working tree as new revision, otherwise use the tree for
322
the specified revision.
324
:param revision_specs:
325
Zero, one or two RevisionSpecs from the command line, saying what revisions
326
to compare. This can be passed as an alternative to the old_revision_spec
327
and new_revision_spec parameters.
329
The more general form is show_diff_trees(), where the caller
330
supplies any two trees.
333
# TODO: perhaps remove the old parameters old_revision_spec and
334
# new_revision_spec, since this is only really for use from cmd_diff and
335
# it now always passes through a sequence of revision_specs -- mbp
340
revision = spec.in_store(tree.branch)
342
revision = spec.in_store(None)
343
revision_id = revision.rev_id
344
branch = revision.branch
345
return branch.repository.revision_tree(revision_id)
347
if revision_specs is not None:
348
assert (old_revision_spec is None
349
and new_revision_spec is None)
350
if len(revision_specs) > 0:
351
old_revision_spec = revision_specs[0]
352
if len(revision_specs) > 1:
353
new_revision_spec = revision_specs[1]
355
if old_revision_spec is None:
356
old_tree = tree.basis_tree()
358
old_tree = spec_tree(old_revision_spec)
360
if (new_revision_spec is None
361
or new_revision_spec.spec is None):
364
new_tree = spec_tree(new_revision_spec)
366
if new_tree is not tree:
367
extra_trees = (tree,)
371
return show_diff_trees(old_tree, new_tree, sys.stdout, specific_files,
372
external_diff_options,
373
old_label=old_label, new_label=new_label,
374
extra_trees=extra_trees)
377
def show_diff_trees(old_tree, new_tree, to_file, specific_files=None,
378
external_diff_options=None,
379
old_label='a/', new_label='b/',
381
"""Show in text form the changes from one tree to another.
384
If set, include only changes to these files.
386
external_diff_options
387
If set, use an external GNU diff and pass these options.
390
If set, more Trees to use for looking up file ids
396
return _show_diff_trees(old_tree, new_tree, to_file,
397
specific_files, external_diff_options,
398
old_label=old_label, new_label=new_label,
399
extra_trees=extra_trees)
406
def _show_diff_trees(old_tree, new_tree, to_file,
407
specific_files, external_diff_options,
408
old_label='a/', new_label='b/', extra_trees=None):
410
# GNU Patch uses the epoch date to detect files that are being added
411
# or removed in a diff.
412
EPOCH_DATE = '1970-01-01 00:00:00 +0000'
414
# TODO: Generation of pseudo-diffs for added/deleted files could
415
# be usefully made into a much faster special case.
417
if external_diff_options:
418
assert isinstance(external_diff_options, basestring)
419
opts = external_diff_options.split()
420
def diff_file(olab, olines, nlab, nlines, to_file):
421
external_diff(olab, olines, nlab, nlines, to_file, opts)
423
diff_file = internal_diff
425
delta = new_tree.changes_from(old_tree,
426
specific_files=specific_files,
427
extra_trees=extra_trees, require_versioned=True)
430
for path, file_id, kind in delta.removed:
432
print >>to_file, '=== removed %s %r' % (kind, path.encode('utf8'))
433
old_name = '%s%s\t%s' % (old_label, path,
434
_patch_header_date(old_tree, file_id, path))
435
new_name = '%s%s\t%s' % (new_label, path, EPOCH_DATE)
436
old_tree.inventory[file_id].diff(diff_file, old_name, old_tree,
437
new_name, None, None, to_file)
438
for path, file_id, kind in delta.added:
440
print >>to_file, '=== added %s %r' % (kind, path.encode('utf8'))
441
old_name = '%s%s\t%s' % (old_label, path, EPOCH_DATE)
442
new_name = '%s%s\t%s' % (new_label, path,
443
_patch_header_date(new_tree, file_id, path))
444
new_tree.inventory[file_id].diff(diff_file, new_name, new_tree,
445
old_name, None, None, to_file,
447
for (old_path, new_path, file_id, kind,
448
text_modified, meta_modified) in delta.renamed:
450
prop_str = get_prop_change(meta_modified)
451
print >>to_file, '=== renamed %s %r => %r%s' % (
452
kind, old_path.encode('utf8'),
453
new_path.encode('utf8'), prop_str)
454
old_name = '%s%s\t%s' % (old_label, old_path,
455
_patch_header_date(old_tree, file_id,
457
new_name = '%s%s\t%s' % (new_label, new_path,
458
_patch_header_date(new_tree, file_id,
460
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
462
text_modified, kind, to_file, diff_file)
463
for path, file_id, kind, text_modified, meta_modified in delta.modified:
465
prop_str = get_prop_change(meta_modified)
466
print >>to_file, '=== modified %s %r%s' % (kind, path.encode('utf8'), prop_str)
467
old_name = '%s%s\t%s' % (old_label, path,
468
_patch_header_date(old_tree, file_id, path))
469
new_name = '%s%s\t%s' % (new_label, path,
470
_patch_header_date(new_tree, file_id, path))
472
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
474
True, kind, to_file, diff_file)
479
def _patch_header_date(tree, file_id, path):
480
"""Returns a timestamp suitable for use in a patch header."""
481
tm = time.gmtime(tree.get_file_mtime(file_id, path))
482
return time.strftime('%Y-%m-%d %H:%M:%S +0000', tm)
485
def _raise_if_nonexistent(paths, old_tree, new_tree):
486
"""Complain if paths are not in either inventory or tree.
488
It's OK with the files exist in either tree's inventory, or
489
if they exist in the tree but are not versioned.
491
This can be used by operations such as bzr status that can accept
492
unknown or ignored files.
494
mutter("check paths: %r", paths)
497
s = old_tree.filter_unversioned_files(paths)
498
s = new_tree.filter_unversioned_files(s)
499
s = [path for path in s if not new_tree.has_filename(path)]
501
raise errors.PathsDoNotExist(sorted(s))
504
def get_prop_change(meta_modified):
506
return " (properties changed)"
511
def _maybe_diff_file_or_symlink(old_path, old_tree, file_id,
512
new_path, new_tree, text_modified,
513
kind, to_file, diff_file):
515
new_entry = new_tree.inventory[file_id]
516
old_tree.inventory[file_id].diff(diff_file,
added
removed
bzrlib/diff.py
