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
if sys.platform == 'win32':
114
# diffutils+gettext from http://gnuwin32.sf.net use only LANGUAGE
115
env['LANGUAGE'] = 'C'
119
stderr = subprocess.PIPE
125
pipe = subprocess.Popen(diffcmd,
126
stdin=subprocess.PIPE,
127
stdout=subprocess.PIPE,
131
if e.errno == errno.ENOENT:
132
raise errors.NoDiff(str(e))
138
def external_diff(old_filename, oldlines, new_filename, newlines, to_file,
140
"""Display a diff by calling out to the external diff program."""
141
# make sure our own output is properly ordered before the diff
144
oldtmp_fd, old_abspath = tempfile.mkstemp(prefix='bzr-diff-old-')
145
newtmp_fd, new_abspath = tempfile.mkstemp(prefix='bzr-diff-new-')
146
oldtmpf = os.fdopen(oldtmp_fd, 'wb')
147
newtmpf = os.fdopen(newtmp_fd, 'wb')
150
# TODO: perhaps a special case for comparing to or from the empty
151
# sequence; can just use /dev/null on Unix
153
# TODO: if either of the files being compared already exists as a
154
# regular named file (e.g. in the working directory) then we can
155
# compare directly to that, rather than copying it.
157
oldtmpf.writelines(oldlines)
158
newtmpf.writelines(newlines)
166
'--label', old_filename,
168
'--label', new_filename,
173
# diff only allows one style to be specified; they don't override.
174
# note that some of these take optargs, and the optargs can be
175
# directly appended to the options.
176
# this is only an approximate parser; it doesn't properly understand
178
for s in ['-c', '-u', '-C', '-U',
183
'-y', '--side-by-side',
195
diffcmd.extend(diff_opts)
197
pipe = _spawn_external_diff(diffcmd, capture_errors=True)
198
out,err = pipe.communicate()
201
# internal_diff() adds a trailing newline, add one here for consistency
204
# 'diff' gives retcode == 2 for all sorts of errors
205
# one of those is 'Binary files differ'.
206
# Bad options could also be the problem.
207
# 'Binary files' is not a real error, so we suppress that error.
210
# Since we got here, we want to make sure to give an i18n error
211
pipe = _spawn_external_diff(diffcmd, capture_errors=False)
212
out, err = pipe.communicate()
214
# Write out the new i18n diff response
215
to_file.write(out+'\n')
216
if pipe.returncode != 2:
217
raise errors.BzrError(
218
'external diff failed with exit code 2'
219
' when run with LANG=C and LC_ALL=C,'
220
' but not when run natively: %r' % (diffcmd,))
222
first_line = lang_c_out.split('\n', 1)[0]
223
# Starting with diffutils 2.8.4 the word "binary" was dropped.
224
m = re.match('^(binary )?files.*differ$', first_line, re.I)
226
raise errors.BzrError('external diff failed with exit code 2;'
227
' command: %r' % (diffcmd,))
229
# Binary files differ, just return
232
# If we got to here, we haven't written out the output of diff
236
# returns 1 if files differ; that's OK
238
msg = 'signal %d' % (-rc)
240
msg = 'exit code %d' % rc
242
raise errors.BzrError('external diff failed with %s; command: %r'
247
oldtmpf.close() # and delete
249
# Clean up. Warn in case the files couldn't be deleted
250
# (in case windows still holds the file open, but not
251
# if the files have already been deleted)
253
os.remove(old_abspath)
255
if e.errno not in (errno.ENOENT,):
256
warning('Failed to delete temporary file: %s %s',
259
os.remove(new_abspath)
261
if e.errno not in (errno.ENOENT,):
262
warning('Failed to delete temporary file: %s %s',
266
@deprecated_function(zero_eight)
267
def show_diff(b, from_spec, specific_files, external_diff_options=None,
268
revision2=None, output=None, b2=None):
269
"""Shortcut for showing the diff to the working tree.
271
Please use show_diff_trees instead.
277
None for 'basis tree', or otherwise the old revision to compare against.
279
The more general form is show_diff_trees(), where the caller
280
supplies any two trees.
285
if from_spec is None:
286
old_tree = b.bzrdir.open_workingtree()
288
old_tree = old_tree = old_tree.basis_tree()
290
old_tree = b.repository.revision_tree(from_spec.in_history(b).rev_id)
292
if revision2 is None:
294
new_tree = b.bzrdir.open_workingtree()
296
new_tree = b2.bzrdir.open_workingtree()
298
new_tree = b.repository.revision_tree(revision2.in_history(b).rev_id)
300
return show_diff_trees(old_tree, new_tree, output, specific_files,
301
external_diff_options)
304
def diff_cmd_helper(tree, specific_files, external_diff_options,
305
old_revision_spec=None, new_revision_spec=None,
307
old_label='a/', new_label='b/'):
308
"""Helper for cmd_diff.
313
:param specific_files:
314
The specific files to compare, or None
316
:param external_diff_options:
317
If non-None, run an external diff, and pass it these options
319
:param old_revision_spec:
320
If None, use basis tree as old revision, otherwise use the tree for
321
the specified revision.
323
:param new_revision_spec:
324
If None, use working tree as new revision, otherwise use the tree for
325
the specified revision.
327
:param revision_specs:
328
Zero, one or two RevisionSpecs from the command line, saying what revisions
329
to compare. This can be passed as an alternative to the old_revision_spec
330
and new_revision_spec parameters.
332
The more general form is show_diff_trees(), where the caller
333
supplies any two trees.
336
# TODO: perhaps remove the old parameters old_revision_spec and
337
# new_revision_spec, since this is only really for use from cmd_diff and
338
# it now always passes through a sequence of revision_specs -- mbp
343
revision = spec.in_store(tree.branch)
345
revision = spec.in_store(None)
346
revision_id = revision.rev_id
347
branch = revision.branch
348
return branch.repository.revision_tree(revision_id)
350
if revision_specs is not None:
351
assert (old_revision_spec is None
352
and new_revision_spec is None)
353
if len(revision_specs) > 0:
354
old_revision_spec = revision_specs[0]
355
if len(revision_specs) > 1:
356
new_revision_spec = revision_specs[1]
358
if old_revision_spec is None:
359
old_tree = tree.basis_tree()
361
old_tree = spec_tree(old_revision_spec)
363
if (new_revision_spec is None
364
or new_revision_spec.spec is None):
367
new_tree = spec_tree(new_revision_spec)
369
if new_tree is not tree:
370
extra_trees = (tree,)
374
return show_diff_trees(old_tree, new_tree, sys.stdout, specific_files,
375
external_diff_options,
376
old_label=old_label, new_label=new_label,
377
extra_trees=extra_trees)
380
def show_diff_trees(old_tree, new_tree, to_file, specific_files=None,
381
external_diff_options=None,
382
old_label='a/', new_label='b/',
384
"""Show in text form the changes from one tree to another.
387
If set, include only changes to these files.
389
external_diff_options
390
If set, use an external GNU diff and pass these options.
393
If set, more Trees to use for looking up file ids
397
if extra_trees is not None:
398
for tree in extra_trees:
402
return _show_diff_trees(old_tree, new_tree, to_file,
403
specific_files, external_diff_options,
404
old_label=old_label, new_label=new_label,
405
extra_trees=extra_trees)
408
if extra_trees is not None:
409
for tree in extra_trees:
415
def _show_diff_trees(old_tree, new_tree, to_file,
416
specific_files, external_diff_options,
417
old_label='a/', new_label='b/', extra_trees=None):
419
# GNU Patch uses the epoch date to detect files that are being added
420
# or removed in a diff.
421
EPOCH_DATE = '1970-01-01 00:00:00 +0000'
423
# TODO: Generation of pseudo-diffs for added/deleted files could
424
# be usefully made into a much faster special case.
426
if external_diff_options:
427
assert isinstance(external_diff_options, basestring)
428
opts = external_diff_options.split()
429
def diff_file(olab, olines, nlab, nlines, to_file):
430
external_diff(olab, olines, nlab, nlines, to_file, opts)
432
diff_file = internal_diff
434
delta = new_tree.changes_from(old_tree,
435
specific_files=specific_files,
436
extra_trees=extra_trees, require_versioned=True)
439
for path, file_id, kind in delta.removed:
441
print >>to_file, '=== removed %s %r' % (kind, path.encode('utf8'))
442
old_name = '%s%s\t%s' % (old_label, path,
443
_patch_header_date(old_tree, file_id, path))
444
new_name = '%s%s\t%s' % (new_label, path, EPOCH_DATE)
445
old_tree.inventory[file_id].diff(diff_file, old_name, old_tree,
446
new_name, None, None, to_file)
447
for path, file_id, kind in delta.added:
449
print >>to_file, '=== added %s %r' % (kind, path.encode('utf8'))
450
old_name = '%s%s\t%s' % (old_label, path, EPOCH_DATE)
451
new_name = '%s%s\t%s' % (new_label, path,
452
_patch_header_date(new_tree, file_id, path))
453
new_tree.inventory[file_id].diff(diff_file, new_name, new_tree,
454
old_name, None, None, to_file,
456
for (old_path, new_path, file_id, kind,
457
text_modified, meta_modified) in delta.renamed:
459
prop_str = get_prop_change(meta_modified)
460
print >>to_file, '=== renamed %s %r => %r%s' % (
461
kind, old_path.encode('utf8'),
462
new_path.encode('utf8'), prop_str)
463
old_name = '%s%s\t%s' % (old_label, old_path,
464
_patch_header_date(old_tree, file_id,
466
new_name = '%s%s\t%s' % (new_label, new_path,
467
_patch_header_date(new_tree, file_id,
469
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
471
text_modified, kind, to_file, diff_file)
472
for path, file_id, kind, text_modified, meta_modified in delta.modified:
474
prop_str = get_prop_change(meta_modified)
475
print >>to_file, '=== modified %s %r%s' % (kind, path.encode('utf8'), prop_str)
476
old_name = '%s%s\t%s' % (old_label, path,
477
_patch_header_date(old_tree, file_id, path))
478
new_name = '%s%s\t%s' % (new_label, path,
479
_patch_header_date(new_tree, file_id, path))
481
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
483
True, kind, to_file, diff_file)
488
def _patch_header_date(tree, file_id, path):
489
"""Returns a timestamp suitable for use in a patch header."""
490
tm = time.gmtime(tree.get_file_mtime(file_id, path))
491
return time.strftime('%Y-%m-%d %H:%M:%S +0000', tm)
494
def _raise_if_nonexistent(paths, old_tree, new_tree):
495
"""Complain if paths are not in either inventory or tree.
497
It's OK with the files exist in either tree's inventory, or
498
if they exist in the tree but are not versioned.
500
This can be used by operations such as bzr status that can accept
501
unknown or ignored files.
503
mutter("check paths: %r", paths)
506
s = old_tree.filter_unversioned_files(paths)
507
s = new_tree.filter_unversioned_files(s)
508
s = [path for path in s if not new_tree.has_filename(path)]
510
raise errors.PathsDoNotExist(sorted(s))
513
def get_prop_change(meta_modified):
515
return " (properties changed)"
520
def _maybe_diff_file_or_symlink(old_path, old_tree, file_id,
521
new_path, new_tree, text_modified,
522
kind, to_file, diff_file):
524
new_entry = new_tree.inventory[file_id]
525
old_tree.inventory[file_id].diff(diff_file,
added
removed
bzrlib/diff.py
