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
28
from bzrlib.patiencediff import unified_diff
29
import bzrlib.patiencediff
30
from bzrlib.symbol_versioning import *
31
from bzrlib.textfile import check_text_lines
32
from bzrlib.trace import mutter, warning
35
# TODO: Rather than building a changeset object, we should probably
36
# invoke callbacks on an object. That object can either accumulate a
37
# list, write them out directly, etc etc.
39
def internal_diff(old_filename, oldlines, new_filename, newlines, to_file,
40
allow_binary=False, sequence_matcher=None,
41
path_encoding='utf8'):
42
# FIXME: difflib is wrong if there is no trailing newline.
43
# The syntax used by patch seems to be "\ No newline at
44
# end of file" following the last diff line from that
45
# file. This is not trivial to insert into the
46
# unified_diff output and it might be better to just fix
47
# or replace that function.
49
# In the meantime we at least make sure the patch isn't
53
# Special workaround for Python2.3, where difflib fails if
54
# both sequences are empty.
55
if not oldlines and not newlines:
58
if allow_binary is False:
59
check_text_lines(oldlines)
60
check_text_lines(newlines)
62
if sequence_matcher is None:
63
sequence_matcher = bzrlib.patiencediff.PatienceSequenceMatcher
64
ud = unified_diff(oldlines, newlines,
65
fromfile=old_filename.encode(path_encoding),
66
tofile=new_filename.encode(path_encoding),
67
sequencematcher=sequence_matcher)
70
# work-around for difflib being too smart for its own good
71
# if /dev/null is "1,0", patch won't recognize it as /dev/null
73
ud[2] = ud[2].replace('-1,0', '-0,0')
75
ud[2] = ud[2].replace('+1,0', '+0,0')
76
# work around for difflib emitting random spaces after the label
77
ud[0] = ud[0][:-2] + '\n'
78
ud[1] = ud[1][:-2] + '\n'
82
if not line.endswith('\n'):
83
to_file.write("\n\\ No newline at end of file\n")
87
def external_diff(old_filename, oldlines, new_filename, newlines, to_file,
89
"""Display a diff by calling out to the external diff program."""
90
if hasattr(to_file, 'fileno'):
94
out_file = subprocess.PIPE
97
# make sure our own output is properly ordered before the diff
100
oldtmp_fd, old_abspath = tempfile.mkstemp(prefix='bzr-diff-old-')
101
newtmp_fd, new_abspath = tempfile.mkstemp(prefix='bzr-diff-new-')
102
oldtmpf = os.fdopen(oldtmp_fd, 'wb')
103
newtmpf = os.fdopen(newtmp_fd, 'wb')
106
# TODO: perhaps a special case for comparing to or from the empty
107
# sequence; can just use /dev/null on Unix
109
# TODO: if either of the files being compared already exists as a
110
# regular named file (e.g. in the working directory) then we can
111
# compare directly to that, rather than copying it.
113
oldtmpf.writelines(oldlines)
114
newtmpf.writelines(newlines)
122
'--label', old_filename,
124
'--label', new_filename,
129
# diff only allows one style to be specified; they don't override.
130
# note that some of these take optargs, and the optargs can be
131
# directly appended to the options.
132
# this is only an approximate parser; it doesn't properly understand
134
for s in ['-c', '-u', '-C', '-U',
139
'-y', '--side-by-side',
151
diffcmd.extend(diff_opts)
154
pipe = subprocess.Popen(diffcmd,
155
stdin=subprocess.PIPE,
158
if e.errno == errno.ENOENT:
159
raise errors.NoDiff(str(e))
164
bzrlib.osutils.pumpfile(pipe.stdout, to_file)
167
if rc != 0 and rc != 1:
168
# returns 1 if files differ; that's OK
170
msg = 'signal %d' % (-rc)
172
msg = 'exit code %d' % rc
174
raise BzrError('external diff failed with %s; command: %r' % (rc, diffcmd))
176
oldtmpf.close() # and delete
178
# Clean up. Warn in case the files couldn't be deleted
179
# (in case windows still holds the file open, but not
180
# if the files have already been deleted)
182
os.remove(old_abspath)
184
if e.errno not in (errno.ENOENT,):
185
warning('Failed to delete temporary file: %s %s',
188
os.remove(new_abspath)
190
if e.errno not in (errno.ENOENT,):
191
warning('Failed to delete temporary file: %s %s',
195
@deprecated_function(zero_eight)
196
def show_diff(b, from_spec, specific_files, external_diff_options=None,
197
revision2=None, output=None, b2=None):
198
"""Shortcut for showing the diff to the working tree.
200
Please use show_diff_trees instead.
206
None for 'basis tree', or otherwise the old revision to compare against.
208
The more general form is show_diff_trees(), where the caller
209
supplies any two trees.
214
if from_spec is None:
215
old_tree = b.bzrdir.open_workingtree()
217
old_tree = old_tree = old_tree.basis_tree()
219
old_tree = b.repository.revision_tree(from_spec.in_history(b).rev_id)
221
if revision2 is None:
223
new_tree = b.bzrdir.open_workingtree()
225
new_tree = b2.bzrdir.open_workingtree()
227
new_tree = b.repository.revision_tree(revision2.in_history(b).rev_id)
229
return show_diff_trees(old_tree, new_tree, output, specific_files,
230
external_diff_options)
233
def diff_cmd_helper(tree, specific_files, external_diff_options,
234
old_revision_spec=None, new_revision_spec=None,
235
old_label='a/', new_label='b/'):
236
"""Helper for cmd_diff.
242
The specific files to compare, or None
244
external_diff_options
245
If non-None, run an external diff, and pass it these options
248
If None, use basis tree as old revision, otherwise use the tree for
249
the specified revision.
252
If None, use working tree as new revision, otherwise use the tree for
253
the specified revision.
255
The more general form is show_diff_trees(), where the caller
256
supplies any two trees.
260
revision_id = spec.in_store(tree.branch).rev_id
261
return tree.branch.repository.revision_tree(revision_id)
262
if old_revision_spec is None:
263
old_tree = tree.basis_tree()
265
old_tree = spec_tree(old_revision_spec)
267
if new_revision_spec is None:
270
new_tree = spec_tree(new_revision_spec)
272
return show_diff_trees(old_tree, new_tree, sys.stdout, specific_files,
273
external_diff_options,
274
old_label=old_label, new_label=new_label)
277
def show_diff_trees(old_tree, new_tree, to_file, specific_files=None,
278
external_diff_options=None,
279
old_label='a/', new_label='b/'):
280
"""Show in text form the changes from one tree to another.
283
If set, include only changes to these files.
285
external_diff_options
286
If set, use an external GNU diff and pass these options.
292
return _show_diff_trees(old_tree, new_tree, to_file,
293
specific_files, external_diff_options,
294
old_label=old_label, new_label=new_label)
301
def _show_diff_trees(old_tree, new_tree, to_file,
302
specific_files, external_diff_options,
303
old_label='a/', new_label='b/' ):
305
# GNU Patch uses the epoch date to detect files that are being added
306
# or removed in a diff.
307
EPOCH_DATE = '1970-01-01 00:00:00 +0000'
309
# TODO: Generation of pseudo-diffs for added/deleted files could
310
# be usefully made into a much faster special case.
312
_raise_if_doubly_unversioned(specific_files, old_tree, new_tree)
314
if external_diff_options:
315
assert isinstance(external_diff_options, basestring)
316
opts = external_diff_options.split()
317
def diff_file(olab, olines, nlab, nlines, to_file):
318
external_diff(olab, olines, nlab, nlines, to_file, opts)
320
diff_file = internal_diff
322
delta = compare_trees(old_tree, new_tree, want_unchanged=False,
323
specific_files=specific_files)
326
for path, file_id, kind in delta.removed:
328
print >>to_file, '=== removed %s %r' % (kind, path.encode('utf8'))
329
old_name = '%s%s\t%s' % (old_label, path,
330
_patch_header_date(old_tree, file_id, path))
331
new_name = '%s%s\t%s' % (new_label, path, EPOCH_DATE)
332
old_tree.inventory[file_id].diff(diff_file, old_name, old_tree,
333
new_name, None, None, to_file)
334
for path, file_id, kind in delta.added:
336
print >>to_file, '=== added %s %r' % (kind, path.encode('utf8'))
337
old_name = '%s%s\t%s' % (old_label, path, EPOCH_DATE)
338
new_name = '%s%s\t%s' % (new_label, path,
339
_patch_header_date(new_tree, file_id, path))
340
new_tree.inventory[file_id].diff(diff_file, new_name, new_tree,
341
old_name, None, None, to_file,
343
for (old_path, new_path, file_id, kind,
344
text_modified, meta_modified) in delta.renamed:
346
prop_str = get_prop_change(meta_modified)
347
print >>to_file, '=== renamed %s %r => %r%s' % (
348
kind, old_path.encode('utf8'),
349
new_path.encode('utf8'), prop_str)
350
old_name = '%s%s\t%s' % (old_label, old_path,
351
_patch_header_date(old_tree, file_id,
353
new_name = '%s%s\t%s' % (new_label, new_path,
354
_patch_header_date(new_tree, file_id,
356
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
358
text_modified, kind, to_file, diff_file)
359
for path, file_id, kind, text_modified, meta_modified in delta.modified:
361
prop_str = get_prop_change(meta_modified)
362
print >>to_file, '=== modified %s %r%s' % (kind, path.encode('utf8'), prop_str)
363
old_name = '%s%s\t%s' % (old_label, path,
364
_patch_header_date(old_tree, file_id, path))
365
new_name = '%s%s\t%s' % (new_label, path,
366
_patch_header_date(new_tree, file_id, path))
368
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
370
True, kind, to_file, diff_file)
375
def _patch_header_date(tree, file_id, path):
376
"""Returns a timestamp suitable for use in a patch header."""
377
tm = time.gmtime(tree.get_file_mtime(file_id, path))
378
return time.strftime('%Y-%m-%d %H:%M:%S +0000', tm)
381
def _raise_if_doubly_unversioned(specific_files, old_tree, new_tree):
382
"""Complain if paths are not versioned in either tree."""
383
if not specific_files:
385
old_unversioned = old_tree.filter_unversioned_files(specific_files)
386
new_unversioned = new_tree.filter_unversioned_files(specific_files)
387
unversioned = old_unversioned.intersection(new_unversioned)
389
raise errors.PathsNotVersionedError(sorted(unversioned))
392
def _raise_if_nonexistent(paths, old_tree, new_tree):
393
"""Complain if paths are not in either inventory or tree.
395
It's OK with the files exist in either tree's inventory, or
396
if they exist in the tree but are not versioned.
398
This can be used by operations such as bzr status that can accept
399
unknown or ignored files.
401
mutter("check paths: %r", paths)
404
s = old_tree.filter_unversioned_files(paths)
405
s = new_tree.filter_unversioned_files(s)
406
s = [path for path in s if not new_tree.has_filename(path)]
408
raise errors.PathsDoNotExist(sorted(s))
411
def get_prop_change(meta_modified):
413
return " (properties changed)"
418
def _maybe_diff_file_or_symlink(old_path, old_tree, file_id,
419
new_path, new_tree, text_modified,
420
kind, to_file, diff_file):
422
new_entry = new_tree.inventory[file_id]
423
old_tree.inventory[file_id].diff(diff_file,
added
removed
bzrlib/diff.py
