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
25
# compatability - plugins import compare_trees from diff!!!
26
# deprecated as of 0.10
27
from bzrlib.delta import compare_trees
28
from bzrlib.errors import BzrError
29
import bzrlib.errors as errors
31
from bzrlib.patiencediff import unified_diff
32
import bzrlib.patiencediff
33
from bzrlib.symbol_versioning import (deprecated_function,
35
from bzrlib.textfile import check_text_lines
36
from bzrlib.trace import mutter, warning
39
# TODO: Rather than building a changeset object, we should probably
40
# invoke callbacks on an object. That object can either accumulate a
41
# list, write them out directly, etc etc.
43
def internal_diff(old_filename, oldlines, new_filename, newlines, to_file,
44
allow_binary=False, sequence_matcher=None,
45
path_encoding='utf8'):
46
# FIXME: difflib is wrong if there is no trailing newline.
47
# The syntax used by patch seems to be "\ No newline at
48
# end of file" following the last diff line from that
49
# file. This is not trivial to insert into the
50
# unified_diff output and it might be better to just fix
51
# or replace that function.
53
# In the meantime we at least make sure the patch isn't
57
# Special workaround for Python2.3, where difflib fails if
58
# both sequences are empty.
59
if not oldlines and not newlines:
62
if allow_binary is False:
63
check_text_lines(oldlines)
64
check_text_lines(newlines)
66
if sequence_matcher is None:
67
sequence_matcher = bzrlib.patiencediff.PatienceSequenceMatcher
68
ud = unified_diff(oldlines, newlines,
69
fromfile=old_filename.encode(path_encoding),
70
tofile=new_filename.encode(path_encoding),
71
sequencematcher=sequence_matcher)
74
# work-around for difflib being too smart for its own good
75
# if /dev/null is "1,0", patch won't recognize it as /dev/null
77
ud[2] = ud[2].replace('-1,0', '-0,0')
79
ud[2] = ud[2].replace('+1,0', '+0,0')
80
# work around for difflib emitting random spaces after the label
81
ud[0] = ud[0][:-2] + '\n'
82
ud[1] = ud[1][:-2] + '\n'
86
if not line.endswith('\n'):
87
to_file.write("\n\\ No newline at end of file\n")
92
"""Set the env var LANG=C"""
93
os.environ['LANG'] = 'C'
96
def _spawn_external_diff(diffcmd, capture_errors=True):
97
"""Spawn the externall diff process, and return the child handle.
99
:param diffcmd: The command list to spawn
100
:param capture_errors: Capture stderr as well as setting LANG=C.
101
This lets us read and understand the output of diff, and respond
103
:return: A Popen object.
106
if sys.platform == 'win32':
107
# Win32 doesn't support preexec_fn, but that is
108
# okay, because it doesn't support LANG either.
111
preexec_fn = _set_lang_C
112
stderr = subprocess.PIPE
118
pipe = subprocess.Popen(diffcmd,
119
stdin=subprocess.PIPE,
120
stdout=subprocess.PIPE,
122
preexec_fn=preexec_fn)
124
if e.errno == errno.ENOENT:
125
raise errors.NoDiff(str(e))
131
def external_diff(old_filename, oldlines, new_filename, newlines, to_file,
133
"""Display a diff by calling out to the external diff program."""
134
# make sure our own output is properly ordered before the diff
137
oldtmp_fd, old_abspath = tempfile.mkstemp(prefix='bzr-diff-old-')
138
newtmp_fd, new_abspath = tempfile.mkstemp(prefix='bzr-diff-new-')
139
oldtmpf = os.fdopen(oldtmp_fd, 'wb')
140
newtmpf = os.fdopen(newtmp_fd, 'wb')
143
# TODO: perhaps a special case for comparing to or from the empty
144
# sequence; can just use /dev/null on Unix
146
# TODO: if either of the files being compared already exists as a
147
# regular named file (e.g. in the working directory) then we can
148
# compare directly to that, rather than copying it.
150
oldtmpf.writelines(oldlines)
151
newtmpf.writelines(newlines)
159
'--label', old_filename,
161
'--label', new_filename,
166
# diff only allows one style to be specified; they don't override.
167
# note that some of these take optargs, and the optargs can be
168
# directly appended to the options.
169
# this is only an approximate parser; it doesn't properly understand
171
for s in ['-c', '-u', '-C', '-U',
176
'-y', '--side-by-side',
188
diffcmd.extend(diff_opts)
190
pipe = _spawn_external_diff(diffcmd, capture_errors=True)
191
out,err = pipe.communicate()
194
# internal_diff() adds a trailing newline, add one here for consistency
197
# 'diff' gives retcode == 2 for all sorts of errors
198
# one of those is 'Binary files differ'.
199
# Bad options could also be the problem.
200
# 'Binary files' is not a real error, so we suppress that error.
203
# Since we got here, we want to make sure to give an i18n error
204
pipe = _spawn_external_diff(diffcmd, capture_errors=False)
205
out, err = pipe.communicate()
207
# Write out the new i18n diff response
208
to_file.write(out+'\n')
209
if pipe.returncode != 2:
210
raise BzrError('external diff failed with exit code 2'
211
' when run with LANG=C, but not when run'
212
' natively: %r' % (diffcmd,))
214
first_line = lang_c_out.split('\n', 1)[0]
215
# Starting with diffutils 2.8.4 the word "binary" was dropped.
216
m = re.match('^(binary )?files.*differ$', first_line, re.I)
218
raise BzrError('external diff failed with exit code 2;'
219
' command: %r' % (diffcmd,))
221
# Binary files differ, just return
224
# If we got to here, we haven't written out the output of diff
228
# returns 1 if files differ; that's OK
230
msg = 'signal %d' % (-rc)
232
msg = 'exit code %d' % rc
234
raise BzrError('external diff failed with %s; command: %r'
239
oldtmpf.close() # and delete
241
# Clean up. Warn in case the files couldn't be deleted
242
# (in case windows still holds the file open, but not
243
# if the files have already been deleted)
245
os.remove(old_abspath)
247
if e.errno not in (errno.ENOENT,):
248
warning('Failed to delete temporary file: %s %s',
251
os.remove(new_abspath)
253
if e.errno not in (errno.ENOENT,):
254
warning('Failed to delete temporary file: %s %s',
258
@deprecated_function(zero_eight)
259
def show_diff(b, from_spec, specific_files, external_diff_options=None,
260
revision2=None, output=None, b2=None):
261
"""Shortcut for showing the diff to the working tree.
263
Please use show_diff_trees instead.
269
None for 'basis tree', or otherwise the old revision to compare against.
271
The more general form is show_diff_trees(), where the caller
272
supplies any two trees.
277
if from_spec is None:
278
old_tree = b.bzrdir.open_workingtree()
280
old_tree = old_tree = old_tree.basis_tree()
282
old_tree = b.repository.revision_tree(from_spec.in_history(b).rev_id)
284
if revision2 is None:
286
new_tree = b.bzrdir.open_workingtree()
288
new_tree = b2.bzrdir.open_workingtree()
290
new_tree = b.repository.revision_tree(revision2.in_history(b).rev_id)
292
return show_diff_trees(old_tree, new_tree, output, specific_files,
293
external_diff_options)
296
def diff_cmd_helper(tree, specific_files, external_diff_options,
297
old_revision_spec=None, new_revision_spec=None,
298
old_label='a/', new_label='b/'):
299
"""Helper for cmd_diff.
305
The specific files to compare, or None
307
external_diff_options
308
If non-None, run an external diff, and pass it these options
311
If None, use basis tree as old revision, otherwise use the tree for
312
the specified revision.
315
If None, use working tree as new revision, otherwise use the tree for
316
the specified revision.
318
The more general form is show_diff_trees(), where the caller
319
supplies any two trees.
323
revision = spec.in_store(tree.branch)
325
revision = spec.in_store(None)
326
revision_id = revision.rev_id
327
branch = revision.branch
328
return branch.repository.revision_tree(revision_id)
329
if old_revision_spec is None:
330
old_tree = tree.basis_tree()
332
old_tree = spec_tree(old_revision_spec)
334
if new_revision_spec is None:
337
new_tree = spec_tree(new_revision_spec)
338
if new_tree is not tree:
339
extra_trees = (tree,)
343
return show_diff_trees(old_tree, new_tree, sys.stdout, specific_files,
344
external_diff_options,
345
old_label=old_label, new_label=new_label,
346
extra_trees=extra_trees)
349
def show_diff_trees(old_tree, new_tree, to_file, specific_files=None,
350
external_diff_options=None,
351
old_label='a/', new_label='b/',
353
"""Show in text form the changes from one tree to another.
356
If set, include only changes to these files.
358
external_diff_options
359
If set, use an external GNU diff and pass these options.
362
If set, more Trees to use for looking up file ids
368
return _show_diff_trees(old_tree, new_tree, to_file,
369
specific_files, external_diff_options,
370
old_label=old_label, new_label=new_label,
371
extra_trees=extra_trees)
378
def _show_diff_trees(old_tree, new_tree, to_file,
379
specific_files, external_diff_options,
380
old_label='a/', new_label='b/', extra_trees=None):
382
# GNU Patch uses the epoch date to detect files that are being added
383
# or removed in a diff.
384
EPOCH_DATE = '1970-01-01 00:00:00 +0000'
386
# TODO: Generation of pseudo-diffs for added/deleted files could
387
# be usefully made into a much faster special case.
389
if external_diff_options:
390
assert isinstance(external_diff_options, basestring)
391
opts = external_diff_options.split()
392
def diff_file(olab, olines, nlab, nlines, to_file):
393
external_diff(olab, olines, nlab, nlines, to_file, opts)
395
diff_file = internal_diff
397
delta = new_tree.changes_from(old_tree,
398
specific_files=specific_files,
399
extra_trees=extra_trees, require_versioned=True)
402
for path, file_id, kind in delta.removed:
404
print >>to_file, '=== removed %s %r' % (kind, path.encode('utf8'))
405
old_name = '%s%s\t%s' % (old_label, path,
406
_patch_header_date(old_tree, file_id, path))
407
new_name = '%s%s\t%s' % (new_label, path, EPOCH_DATE)
408
old_tree.inventory[file_id].diff(diff_file, old_name, old_tree,
409
new_name, None, None, to_file)
410
for path, file_id, kind in delta.added:
412
print >>to_file, '=== added %s %r' % (kind, path.encode('utf8'))
413
old_name = '%s%s\t%s' % (old_label, path, EPOCH_DATE)
414
new_name = '%s%s\t%s' % (new_label, path,
415
_patch_header_date(new_tree, file_id, path))
416
new_tree.inventory[file_id].diff(diff_file, new_name, new_tree,
417
old_name, None, None, to_file,
419
for (old_path, new_path, file_id, kind,
420
text_modified, meta_modified) in delta.renamed:
422
prop_str = get_prop_change(meta_modified)
423
print >>to_file, '=== renamed %s %r => %r%s' % (
424
kind, old_path.encode('utf8'),
425
new_path.encode('utf8'), prop_str)
426
old_name = '%s%s\t%s' % (old_label, old_path,
427
_patch_header_date(old_tree, file_id,
429
new_name = '%s%s\t%s' % (new_label, new_path,
430
_patch_header_date(new_tree, file_id,
432
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
434
text_modified, kind, to_file, diff_file)
435
for path, file_id, kind, text_modified, meta_modified in delta.modified:
437
prop_str = get_prop_change(meta_modified)
438
print >>to_file, '=== modified %s %r%s' % (kind, path.encode('utf8'), prop_str)
439
old_name = '%s%s\t%s' % (old_label, path,
440
_patch_header_date(old_tree, file_id, path))
441
new_name = '%s%s\t%s' % (new_label, path,
442
_patch_header_date(new_tree, file_id, path))
444
_maybe_diff_file_or_symlink(old_name, old_tree, file_id,
446
True, kind, to_file, diff_file)
451
def _patch_header_date(tree, file_id, path):
452
"""Returns a timestamp suitable for use in a patch header."""
453
tm = time.gmtime(tree.get_file_mtime(file_id, path))
454
return time.strftime('%Y-%m-%d %H:%M:%S +0000', tm)
457
def _raise_if_nonexistent(paths, old_tree, new_tree):
458
"""Complain if paths are not in either inventory or tree.
460
It's OK with the files exist in either tree's inventory, or
461
if they exist in the tree but are not versioned.
463
This can be used by operations such as bzr status that can accept
464
unknown or ignored files.
466
mutter("check paths: %r", paths)
469
s = old_tree.filter_unversioned_files(paths)
470
s = new_tree.filter_unversioned_files(s)
471
s = [path for path in s if not new_tree.has_filename(path)]
473
raise errors.PathsDoNotExist(sorted(s))
476
def get_prop_change(meta_modified):
478
return " (properties changed)"
483
def _maybe_diff_file_or_symlink(old_path, old_tree, file_id,
484
new_path, new_tree, text_modified,
485
kind, to_file, diff_file):
487
new_entry = new_tree.inventory[file_id]
488
old_tree.inventory[file_id].diff(diff_file,
added
removed
bzrlib/diff.py
