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
24
from bzrlib.delta import compare_trees
25
from bzrlib.errors import BzrError
26
import bzrlib.errors as errors
27
from bzrlib.patiencediff import unified_diff
28
import bzrlib.patiencediff
29
from bzrlib.symbol_versioning import *
30
from bzrlib.textfile import check_text_lines
31
from bzrlib.trace import mutter, warning
34
# TODO: Rather than building a changeset object, we should probably
35
# invoke callbacks on an object. That object can either accumulate a
36
# list, write them out directly, etc etc.
38
def internal_diff(old_filename, oldlines, new_filename, newlines, to_file,
39
allow_binary=False, sequence_matcher=None,
40
path_encoding='utf8'):
41
# FIXME: difflib is wrong if there is no trailing newline.
42
# The syntax used by patch seems to be "\ No newline at
43
# end of file" following the last diff line from that
44
# file. This is not trivial to insert into the
45
# unified_diff output and it might be better to just fix
46
# or replace that function.
48
# In the meantime we at least make sure the patch isn't
52
# Special workaround for Python2.3, where difflib fails if
53
# both sequences are empty.
54
if not oldlines and not newlines:
57
if allow_binary is False:
58
check_text_lines(oldlines)
59
check_text_lines(newlines)
61
if sequence_matcher is None:
62
sequence_matcher = bzrlib.patiencediff.PatienceSequenceMatcher
63
ud = unified_diff(oldlines, newlines,
64
fromfile=old_filename.encode(path_encoding),
65
tofile=new_filename.encode(path_encoding),
66
sequencematcher=sequence_matcher)
69
# work-around for difflib being too smart for its own good
70
# if /dev/null is "1,0", patch won't recognize it as /dev/null
72
ud[2] = ud[2].replace('-1,0', '-0,0')
74
ud[2] = ud[2].replace('+1,0', '+0,0')
75
# work around for difflib emitting random spaces after the label
76
ud[0] = ud[0][:-2] + '\n'
77
ud[1] = ud[1][:-2] + '\n'
81
if not line.endswith('\n'):
82
to_file.write("\n\\ No newline at end of file\n")
86
def external_diff(old_filename, oldlines, new_filename, newlines, to_file,
88
"""Display a diff by calling out to the external diff program."""
89
if not hasattr(to_file, 'fileno'):
90
raise NotImplementedError("sorry, can't send external diff other "
91
"than to a file descriptor", to_file)
93
# make sure our own output is properly ordered before the diff
96
oldtmp_fd, old_abspath = tempfile.mkstemp(prefix='bzr-diff-old-')
97
newtmp_fd, new_abspath = tempfile.mkstemp(prefix='bzr-diff-new-')
98
oldtmpf = os.fdopen(oldtmp_fd, 'wb')
99
newtmpf = os.fdopen(newtmp_fd, 'wb')
102
# TODO: perhaps a special case for comparing to or from the empty
103
# sequence; can just use /dev/null on Unix
105
# TODO: if either of the files being compared already exists as a
106
# regular named file (e.g. in the working directory) then we can
107
# compare directly to that, rather than copying it.
109
oldtmpf.writelines(oldlines)
110
newtmpf.writelines(newlines)
118
'--label', old_filename,
120
'--label', new_filename,
125
# diff only allows one style to be specified; they don't override.
126
# note that some of these take optargs, and the optargs can be
127
# directly appended to the options.
128
# this is only an approximate parser; it doesn't properly understand
130
for s in ['-c', '-u', '-C', '-U',
135
'-y', '--side-by-side',
147
diffcmd.extend(diff_opts)
150
pipe = subprocess.Popen(diffcmd,
151
stdin=subprocess.PIPE,
154
if e.errno == errno.ENOENT:
155
raise errors.NoDiff(str(e))
160
if rc != 0 and rc != 1:
161
# returns 1 if files differ; that's OK
163
msg = 'signal %d' % (-rc)
165
msg = 'exit code %d' % rc
167
raise BzrError('external diff failed with %s; command: %r' % (rc, diffcmd))
169
oldtmpf.close() # and delete
171
# Clean up. Warn in case the files couldn't be deleted
172
# (in case windows still holds the file open, but not
173
# if the files have already been deleted)
175
os.remove(old_abspath)
177
if e.errno not in (errno.ENOENT,):
178
warning('Failed to delete temporary file: %s %s',
181
os.remove(new_abspath)
183
if e.errno not in (errno.ENOENT,):
184
warning('Failed to delete temporary file: %s %s',
188
@deprecated_function(zero_eight)
189
def show_diff(b, from_spec, specific_files, external_diff_options=None,
190
revision2=None, output=None, b2=None):
191
"""Shortcut for showing the diff to the working tree.
193
Please use show_diff_trees instead.
199
None for 'basis tree', or otherwise the old revision to compare against.
201
The more general form is show_diff_trees(), where the caller
202
supplies any two trees.
207
if from_spec is None:
208
old_tree = b.bzrdir.open_workingtree()
210
old_tree = old_tree = old_tree.basis_tree()
212
old_tree = b.repository.revision_tree(from_spec.in_history(b).rev_id)
214
if revision2 is None:
216
new_tree = b.bzrdir.open_workingtree()
218
new_tree = b2.bzrdir.open_workingtree()
220
new_tree = b.repository.revision_tree(revision2.in_history(b).rev_id)
222
return show_diff_trees(old_tree, new_tree, output, specific_files,
223
external_diff_options)
226
def diff_cmd_helper(tree, specific_files, external_diff_options,
227
old_revision_spec=None, new_revision_spec=None,
228
old_label='a/', new_label='b/'):
229
"""Helper for cmd_diff.
235
The specific files to compare, or None
237
external_diff_options
238
If non-None, run an external diff, and pass it these options
241
If None, use basis tree as old revision, otherwise use the tree for
242
the specified revision.
245
If None, use working tree as new revision, otherwise use the tree for
246
the specified revision.
248
The more general form is show_diff_trees(), where the caller
249
supplies any two trees.
253
revision_id = spec.in_store(tree.branch).rev_id
254
return tree.branch.repository.revision_tree(revision_id)
255
if old_revision_spec is None:
256
old_tree = tree.basis_tree()
258
old_tree = spec_tree(old_revision_spec)
260
if new_revision_spec is None:
263
new_tree = spec_tree(new_revision_spec)
265
return show_diff_trees(old_tree, new_tree, sys.stdout, specific_files,
266
external_diff_options,
267
old_label=old_label, new_label=new_label)
270
def show_diff_trees(old_tree, new_tree, to_file, specific_files=None,
271
external_diff_options=None,
272
old_label='a/', new_label='b/'):
273
"""Show in text form the changes from one tree to another.
276
If set, include only changes to these files.
278
external_diff_options
279
If set, use an external GNU diff and pass these options.
285
return _show_diff_trees(old_tree, new_tree, to_file,
286
specific_files, external_diff_options,
287
old_label=old_label, new_label=new_label)
294
def _show_diff_trees(old_tree, new_tree, to_file,
295
specific_files, external_diff_options,
296
old_label='a/', new_label='b/' ):
298
# GNU Patch uses the epoch date to detect files that are being added
299
# or removed in a diff.
300
EPOCH_DATE = '1970-01-01 00:00:00 +0000'
302
# TODO: Generation of pseudo-diffs for added/deleted files could
303
# be usefully made into a much faster special case.
305
_raise_if_doubly_unversioned(specific_files, old_tree, new_tree)
307
if external_diff_options:
308
assert isinstance(external_diff_options, basestring)
309
opts = external_diff_options.split()
310
def diff_file(olab, olines, nlab, nlines, to_file):
311
external_diff(olab, olines, nlab, nlines, to_file, opts)
313
diff_file = internal_diff
315
delta = compare_trees(old_tree, new_tree, want_unchanged=False,
316
specific_files=specific_files)
319
for path, file_id, kind in delta.removed:
321
print >>to_file, '=== removed %s %r' % (kind, path.encode('utf8'))
322
old_name = '%s%s\t%s' % (old_label, path,
323
_patch_header_date(old_tree, file_id, path))
324
new_name = '%s%s\t%s' % (new_label, path, EPOCH_DATE)
325
old_tree.inventory[file_id].diff(diff_file, old_name, old_tree,
326
new_name, None, None, to_file)
327
for path, file_id, kind in delta.added:
329
print >>to_file, '=== added %s %r' % (kind, path.encode('utf8'))
330
old_name = '%s%s\t%s' % (old_label, path, EPOCH_DATE)
331
new_name = '%s%s\t%s' % (new_label, path,
332
_patch_header_date(new_tree, file_id, path))
333
new_tree.inventory[file_id].diff(diff_file, new_name, new_tree,
334
old_name, None, None, to_file,
336
for (old_path, new_path, file_id, kind,
337
text_modified, meta_modified) in delta.renamed:
339
prop_str = get_prop_change(meta_modified)
340
print >>to_file, '=== renamed %s %r => %r%s' % (
341
kind, old_path.encode('utf8'),
342
new_path.encode('utf8'), prop_str)
343
old_name = '%s%s\t%s' % (old_label, old_path,
344
_patch_header_date(old_tree, file_id,
346
new_name = '%s%s\t%s' % (new_label, new_path,
347
_patch_header_date(new_tree, file_id,
349
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
351
text_modified, kind, to_file, diff_file)
352
for path, file_id, kind, text_modified, meta_modified in delta.modified:
354
prop_str = get_prop_change(meta_modified)
355
print >>to_file, '=== modified %s %r%s' % (kind, path.encode('utf8'), prop_str)
356
old_name = '%s%s\t%s' % (old_label, path,
357
_patch_header_date(old_tree, file_id, path))
358
new_name = '%s%s\t%s' % (new_label, path,
359
_patch_header_date(new_tree, file_id, path))
361
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
363
True, kind, to_file, diff_file)
368
def _patch_header_date(tree, file_id, path):
369
"""Returns a timestamp suitable for use in a patch header."""
370
tm = time.gmtime(tree.get_file_mtime(file_id, path))
371
return time.strftime('%Y-%m-%d %H:%M:%S +0000', tm)
374
def _raise_if_doubly_unversioned(specific_files, old_tree, new_tree):
375
"""Complain if paths are not versioned in either tree."""
376
if not specific_files:
378
old_unversioned = old_tree.filter_unversioned_files(specific_files)
379
new_unversioned = new_tree.filter_unversioned_files(specific_files)
380
unversioned = old_unversioned.intersection(new_unversioned)
382
raise errors.PathsNotVersionedError(sorted(unversioned))
385
def _raise_if_nonexistent(paths, old_tree, new_tree):
386
"""Complain if paths are not in either inventory or tree.
388
It's OK with the files exist in either tree's inventory, or
389
if they exist in the tree but are not versioned.
391
This can be used by operations such as bzr status that can accept
392
unknown or ignored files.
394
mutter("check paths: %r", paths)
397
s = old_tree.filter_unversioned_files(paths)
398
s = new_tree.filter_unversioned_files(s)
399
s = [path for path in s if not new_tree.has_filename(path)]
401
raise errors.PathsDoNotExist(sorted(s))
404
def get_prop_change(meta_modified):
406
return " (properties changed)"
411
def _maybe_diff_file_or_symlink(old_path, old_tree, file_id,
412
new_path, new_tree, text_modified,
413
kind, to_file, diff_file):
415
new_entry = new_tree.inventory[file_id]
416
old_tree.inventory[file_id].diff(diff_file,
added
removed
bzrlib/diff.py
