13
13
# You should have received a copy of the GNU General Public License
14
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17
from io import BytesIO
20
from .lazy_import import lazy_import
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
from __future__ import absolute_import
21
from brzlib.lazy_import import lazy_import
21
22
lazy_import(globals(), """
22
23
from fnmatch import fnmatch
24
from breezy._termcolor import color_string, FG
25
from cStringIO import StringIO
27
from brzlib._termcolor import color_string, re_color_string, FG
29
from brzlib.revisionspec import (
34
40
revision as _mod_revision,
37
from .revisionspec import (
43
45
_user_encoding = osutils.get_user_encoding()
47
49
"""Raised when a revision is not on left-hand history."""
50
class GrepOptions(object):
51
"""Container to pass around grep options.
53
This class is used as a container to pass around user option and
54
some other params (like outf) to processing functions. This makes
55
it easier to add more options as grep evolves.
70
files_with_matches = False
71
files_without_match = False
86
52
def _rev_on_mainline(rev_tuple):
87
53
"""returns True is rev tuple is on mainline"""
88
54
if len(rev_tuple) == 1:
93
59
# NOTE: _linear_view_revisions is basided on
94
# breezy.log._linear_view_revisions.
60
# brzlib.log._linear_view_revisions.
95
61
# This should probably be a common public API
96
62
def _linear_view_revisions(branch, start_rev_id, end_rev_id):
97
63
# requires that start is older than end
110
76
# NOTE: _graph_view_revisions is copied from
111
# breezy.log._graph_view_revisions.
77
# brzlib.log._graph_view_revisions.
112
78
# This should probably be a common public API
113
79
def _graph_view_revisions(branch, start_rev_id, end_rev_id,
114
80
rebase_initial_depths=True):
153
119
def compile_pattern(pattern, flags=0):
155
return re.compile(pattern, flags)
156
except re.error as e:
122
# use python's re.compile as we need to catch re.error in case of bad pattern
123
lazy_regex.reset_compile()
124
patternc = re.compile(pattern, flags)
157
126
raise errors.BzrError("Invalid pattern: '%s'" % pattern)
161
130
def is_fixed_string(s):
162
if re.match("^([A-Za-z0-9_]|\\s)*$", s):
131
if re.match("^([A-Za-z0-9_]|\s)*$", s):
173
142
self.outf = opts.outf
174
143
if opts.show_color:
144
pat = opts.pattern.encode(_user_encoding, 'replace')
175
145
if opts.fixed_string:
176
self._old = opts.pattern
177
self._new = color_string(opts.pattern, FG.BOLD_RED)
147
self._new = color_string(pat, FG.BOLD_RED)
178
148
self.get_writer = self._get_writer_fixed_highlighted
180
150
flags = opts.patternc.flags
181
self._sub = re.compile(
182
opts.pattern.join(("((?:", ")+)")), flags).sub
151
self._sub = re.compile(pat.join(("((?:",")+)")), flags).sub
183
152
self._highlight = color_string("\\1", FG.BOLD_RED)
184
153
self.get_writer = self._get_writer_regexp_highlighted
230
194
"""Get function for writing output with regexp match highlighted"""
231
195
_line_writer = self._get_writer_plain()
232
196
sub, highlight = self._sub, self._highlight
234
197
def _line_writer_regexp_highlighted(line):
235
198
"""Write formatted line with matched pattern highlighted"""
236
199
return _line_writer(line=sub(highlight, line))
240
203
"""Get function for writing output with search string highlighted"""
241
204
_line_writer = self._get_writer_plain()
242
205
old, new = self._old, self._new
244
206
def _line_writer_fixed_highlighted(line):
245
207
"""Write formatted line with string searched for highlighted"""
246
208
return _line_writer(line=line.replace(old, new))
250
212
def grep_diff(opts):
251
213
wt, branch, relpath = \
252
controldir.ControlDir.open_containing_tree_or_branch('.')
253
with branch.lock_read():
214
bzrdir.BzrDir.open_containing_tree_or_branch('.')
254
217
if opts.revision:
255
218
start_rev = opts.revision[0]
257
220
# if no revision is sepcified for diff grep we grep all changesets.
258
221
opts.revision = [RevisionSpec.from_string('revno:1'),
259
RevisionSpec.from_string('last:1')]
222
RevisionSpec.from_string('last:1')]
260
223
start_rev = opts.revision[0]
261
224
start_revid = start_rev.as_revision_id(branch)
262
if start_revid == b'null:':
225
if start_revid == 'null:':
264
227
srevno_tuple = branch.revision_id_to_dotted_revno(start_revid)
265
228
if len(opts.revision) == 2:
269
232
end_revno, end_revid = branch.last_revision_info()
270
233
erevno_tuple = branch.revision_id_to_dotted_revno(end_revid)
272
grep_mainline = (_rev_on_mainline(srevno_tuple)
273
and _rev_on_mainline(erevno_tuple))
235
grep_mainline = (_rev_on_mainline(srevno_tuple) and
236
_rev_on_mainline(erevno_tuple))
275
238
# ensure that we go in reverse order
276
239
if srevno_tuple > erevno_tuple:
281
244
# faster when we don't want to look at merged revs. We try this
282
245
# with _linear_view_revisions. If all revs are to be grepped we
283
246
# use the slower _graph_view_revisions
284
if opts.levels == 1 and grep_mainline:
285
given_revs = _linear_view_revisions(
286
branch, start_revid, end_revid)
247
if opts.levels==1 and grep_mainline:
248
given_revs = _linear_view_revisions(branch, start_revid, end_revid)
288
given_revs = _graph_view_revisions(
289
branch, start_revid, end_revid)
250
given_revs = _graph_view_revisions(branch, start_revid, end_revid)
291
252
# We do an optimization below. For grepping a specific revison
292
253
# We don't need to call _graph_view_revisions which is slow.
296
257
start_rev_tuple = (start_revid, start_revno, 0)
297
258
given_revs = [start_rev_tuple]
298
259
repo = branch.repository
299
diff_pattern = re.compile(
300
b"^[+\\-].*(" + opts.pattern.encode(_user_encoding) + b")")
301
file_pattern = re.compile(b"=== (modified|added|removed) file '.*'")
260
diff_pattern = re.compile("^[+\-].*(" + opts.pattern + ")")
261
file_pattern = re.compile("=== (modified|added|removed) file '.*'", re.UNICODE)
302
262
outputter = _GrepDiffOutputter(opts)
303
263
writeline = outputter.get_writer()
304
264
writerevno = outputter.get_revision_header_writer()
309
269
# with level=1 show only top level
312
rev_spec = RevisionSpec_revid.from_string(
313
"revid:" + revid.decode('utf-8'))
272
rev_spec = RevisionSpec_revid.from_string("revid:"+revid)
314
273
new_rev = repo.get_revision(revid)
315
274
new_tree = rev_spec.as_tree(branch)
316
275
if len(new_rev.parent_ids) == 0:
319
278
ancestor_id = new_rev.parent_ids[0]
320
279
old_tree = repo.revision_tree(ancestor_id)
322
281
diff.show_diff_trees(old_tree, new_tree, s,
323
old_label='', new_label='')
282
old_label='', new_label='')
324
283
display_revno = True
325
284
display_file = False
326
285
file_header = None
334
293
writerevno("=== revno:%s ===" % (revno,))
335
294
display_revno = False
338
" %s" % (file_header.decode(file_encoding, 'replace'),))
296
writefileheader(" %s" % (file_header,))
339
297
display_file = False
340
298
line = line.decode(file_encoding, 'replace')
341
299
writeline(" %s" % (line,))
344
304
def versioned_grep(opts):
345
305
wt, branch, relpath = \
346
controldir.ControlDir.open_containing_tree_or_branch('.')
347
with branch.lock_read():
306
bzrdir.BzrDir.open_containing_tree_or_branch('.')
348
309
start_rev = opts.revision[0]
349
310
start_revid = start_rev.as_revision_id(branch)
350
311
if start_revid is None:
359
320
end_revno, end_revid = branch.last_revision_info()
360
321
erevno_tuple = branch.revision_id_to_dotted_revno(end_revid)
362
grep_mainline = (_rev_on_mainline(srevno_tuple)
363
and _rev_on_mainline(erevno_tuple))
323
grep_mainline = (_rev_on_mainline(srevno_tuple) and
324
_rev_on_mainline(erevno_tuple))
365
326
# ensure that we go in reverse order
366
327
if srevno_tuple > erevno_tuple:
372
333
# with _linear_view_revisions. If all revs are to be grepped we
373
334
# use the slower _graph_view_revisions
374
335
if opts.levels == 1 and grep_mainline:
375
given_revs = _linear_view_revisions(
376
branch, start_revid, end_revid)
336
given_revs = _linear_view_revisions(branch, start_revid, end_revid)
378
given_revs = _graph_view_revisions(
379
branch, start_revid, end_revid)
338
given_revs = _graph_view_revisions(branch, start_revid, end_revid)
381
340
# We do an optimization below. For grepping a specific revison
382
341
# We don't need to call _graph_view_revisions which is slow.
394
353
# with level=1 show only top level
397
rev = RevisionSpec_revid.from_string(
398
"revid:" + revid.decode('utf-8'))
356
rev = RevisionSpec_revid.from_string("revid:"+revid)
399
357
tree = rev.as_tree(branch)
400
358
for path in opts.path_list:
401
tree_path = osutils.pathjoin(relpath, path)
402
if not tree.has_filename(tree_path):
403
trace.warning("Skipped unknown file '%s'.", path)
359
path_for_id = osutils.pathjoin(relpath, path)
360
id = tree.path2id(path_for_id)
362
trace.warning("Skipped unknown file '%s'." % path)
406
365
if osutils.isdir(path):
407
366
path_prefix = path
408
367
dir_grep(tree, path, relpath, opts, revno, path_prefix)
411
tree, tree_path, '.', path, opts, revno)
369
versioned_file_grep(tree, id, '.', path, opts, revno)
414
374
def workingtree_grep(opts):
415
revno = opts.print_revno = None # for working tree set revno to None
375
revno = opts.print_revno = None # for working tree set revno to None
417
377
tree, branch, relpath = \
418
controldir.ControlDir.open_containing_tree_or_branch('.')
378
bzrdir.BzrDir.open_containing_tree_or_branch('.')
420
380
msg = ('Cannot search working tree. Working tree not found.\n'
421
'To search for specific revision in history use the -r option.')
381
'To search for specific revision in history use the -r option.')
422
382
raise errors.BzrCommandError(msg)
424
384
# GZ 2010-06-02: Shouldn't be smuggling this on opts, but easy for now
425
385
opts.outputter = _Outputter(opts)
427
with tree.lock_read():
428
389
for path in opts.path_list:
429
390
if osutils.isdir(path):
430
391
path_prefix = path
431
392
dir_grep(tree, path, relpath, opts, revno, path_prefix)
433
with open(path, 'rb') as f:
434
_file_grep(f.read(), path, opts, revno)
394
_file_grep(open(path).read(), path, opts, revno)
437
399
def _skip_file(include, exclude, path):
446
408
# setup relpath to open files relative to cwd
449
rpath = osutils.pathjoin('..', relpath)
411
rpath = osutils.pathjoin('..',relpath)
451
413
from_dir = osutils.pathjoin(relpath, path)
452
414
if opts.from_root:
453
415
# start searching recursively from root
458
420
to_grep_append = to_grep.append
460
422
# and hits manually refilled. Could do this again if it was
461
423
# for a good reason, otherwise cache might want purging.
462
424
outputter = opts.outputter
463
for fp, fc, fkind, entry in tree.list_files(
464
include_root=False, from_dir=from_dir, recursive=opts.recursive):
425
for fp, fc, fkind, fid, entry in tree.list_files(include_root=False,
426
from_dir=from_dir, recursive=opts.recursive):
466
428
if _skip_file(opts.include, opts.exclude, fp):
469
431
if fc == 'V' and fkind == 'file':
470
tree_path = osutils.pathjoin(from_dir if from_dir else '', fp)
471
if revno is not None:
472
433
# If old result is valid, print results immediately.
473
434
# Otherwise, add file info to to_grep so that the
474
435
# loop later will get chunks and grep them
475
cache_id = tree.get_file_revision(tree_path)
436
cache_id = tree.get_file_revision(fid)
476
437
if cache_id in outputter.cache:
477
438
# GZ 2010-06-05: Not really sure caching and re-outputting
478
439
# the old path is really the right thing,
479
440
# but it's what the old code seemed to do
480
441
outputter.write_cached_lines(cache_id, revno)
482
to_grep_append((tree_path, (fp, tree_path)))
443
to_grep_append((fid, (fp, fid)))
484
445
# we are grepping working tree.
485
446
if from_dir is None:
489
450
if opts.files_with_matches or opts.files_without_match:
490
451
# Optimize for wtree list-only as we don't need to read the
492
with open(path_for_file, 'rb', buffering=4096) as file:
493
_file_grep_list_only_wtree(file, fp, opts, path_prefix)
453
file = open(path_for_file, 'r', buffering=4096)
454
_file_grep_list_only_wtree(file, fp, opts, path_prefix)
495
with open(path_for_file, 'rb') as f:
496
_file_grep(f.read(), fp, opts, revno, path_prefix)
456
file_text = open(path_for_file, 'r').read()
457
_file_grep(file_text, fp, opts, revno, path_prefix)
498
if revno is not None: # grep versioned files
499
for (path, tree_path), chunks in tree.iter_files_bytes(to_grep):
459
if revno != None: # grep versioned files
460
for (path, fid), chunks in tree.iter_files_bytes(to_grep):
500
461
path = _make_display_path(relpath, path)
501
_file_grep(b''.join(chunks), path, opts, revno, path_prefix,
502
tree.get_file_revision(tree_path))
462
_file_grep(chunks[0], path, opts, revno, path_prefix,
463
tree.get_file_revision(fid, path))
505
466
def _make_display_path(relpath, path):
520
def versioned_file_grep(tree, tree_path, relpath, path, opts, revno, path_prefix=None):
481
def versioned_file_grep(tree, id, relpath, path, opts, revno, path_prefix = None):
521
482
"""Create a file object for the specified id and pass it on to _file_grep.
524
485
path = _make_display_path(relpath, path)
525
file_text = tree.get_file_text(tree_path)
486
file_text = tree.get_file_text(id)
526
487
_file_grep(file_text, path, opts, revno, path_prefix)
536
497
def _file_grep_list_only_wtree(file, path, opts, path_prefix=None):
537
498
# test and skip binary files
538
if b'\x00' in file.read(1024):
499
if '\x00' in file.read(1024):
540
trace.warning("Binary file '%s' skipped.", path)
501
trace.warning("Binary file '%s' skipped." % path)
543
file.seek(0) # search from beginning
504
file.seek(0) # search from beginning
546
507
if opts.fixed_string:
549
510
if pattern in line:
552
else: # not fixed_string
513
else: # not fixed_string
553
514
for line in file:
554
515
if opts.patternc.search(line):
558
519
if (opts.files_with_matches and found) or \
559
(opts.files_without_match and not found):
520
(opts.files_without_match and not found):
560
521
if path_prefix and path_prefix != '.':
561
522
# user has passed a dir arg, show that as result prefix
562
523
path = osutils.pathjoin(path_prefix, path)
584
544
no_line = opts.files_with_matches or opts.files_without_match
586
546
if opts.show_color:
547
pat = opts.pattern.encode(_user_encoding, 'replace')
588
549
self.get_writer = self._get_writer_plain
589
550
elif opts.fixed_string:
590
self._old = opts.pattern
591
self._new = color_string(opts.pattern, FG.BOLD_RED)
552
self._new = color_string(pat, FG.BOLD_RED)
592
553
self.get_writer = self._get_writer_fixed_highlighted
594
555
flags = opts.patternc.flags
595
self._sub = re.compile(
596
opts.pattern.join(("((?:", ")+)")), flags).sub
556
self._sub = re.compile(pat.join(("((?:",")+)")), flags).sub
597
557
self._highlight = color_string("\\1", FG.BOLD_RED)
598
558
self.get_writer = self._get_writer_regexp_highlighted
599
559
path_start = FG.MAGENTA
624
584
def _get_writer_plain(self, path, revno, cache_id):
625
585
"""Get function for writing uncoloured output"""
626
586
per_line = self._format_perline
627
start = self._format_initial % {"path": path, "revno": revno}
587
start = self._format_initial % {"path":path, "revno":revno}
628
588
write = self.outf.write
629
589
if self.cache is not None and cache_id is not None:
631
591
self.cache[cache_id] = path, result_list
632
592
add_to_cache = result_list.append
634
593
def _line_cache_and_writer(**kwargs):
635
594
"""Write formatted line and cache arguments"""
636
595
end = per_line % kwargs
637
596
add_to_cache(end)
638
597
write(start + end)
639
598
return _line_cache_and_writer
641
599
def _line_writer(**kwargs):
642
600
"""Write formatted line from arguments given by underlying opts"""
643
601
write(start + per_line % kwargs)
646
604
def write_cached_lines(self, cache_id, revno):
647
605
"""Write cached results out again for new revision"""
648
606
cached_path, cached_matches = self.cache[cache_id]
649
start = self._format_initial % {"path": cached_path, "revno": revno}
607
start = self._format_initial % {"path":cached_path, "revno":revno}
650
608
write = self.outf.write
651
609
for end in cached_matches:
652
610
write(start + end)
655
613
"""Get function for writing output with regexp match highlighted"""
656
614
_line_writer = self._get_writer_plain(path, revno, cache_id)
657
615
sub, highlight = self._sub, self._highlight
659
616
def _line_writer_regexp_highlighted(line, **kwargs):
660
617
"""Write formatted line with matched pattern highlighted"""
661
618
return _line_writer(line=sub(highlight, line), **kwargs)
665
622
"""Get function for writing output with search string highlighted"""
666
623
_line_writer = self._get_writer_plain(path, revno, cache_id)
667
624
old, new = self._old, self._new
669
625
def _line_writer_fixed_highlighted(line, **kwargs):
670
626
"""Write formatted line with string searched for highlighted"""
671
627
return _line_writer(line=line.replace(old, new), **kwargs)
675
631
def _file_grep(file_text, path, opts, revno, path_prefix=None, cache_id=None):
676
632
# test and skip binary files
677
if b'\x00' in file_text[:1024]:
633
if '\x00' in file_text[:1024]:
679
trace.warning("Binary file '%s' skipped.", path)
635
trace.warning("Binary file '%s' skipped." % path)
682
638
if path_prefix and path_prefix != '.':
694
650
if opts.files_with_matches or opts.files_without_match:
695
651
if opts.fixed_string:
696
found = pattern in file_text
652
if sys.platform > (2, 5):
653
found = pattern in file_text
655
for line in file_text.splitlines():
698
662
search = opts.patternc.search
699
if b"$" not in pattern:
663
if "$" not in pattern:
700
664
found = search(file_text) is not None
702
666
for line in file_text.splitlines():
711
675
elif opts.fixed_string:
712
676
# Fast path for no match, search through the entire file at once rather
713
# than a line at a time. <http://effbot.org/zone/stringlib.htm>
714
i = file_text.find(pattern)
717
b = file_text.rfind(b"\n", 0, i) + 1
719
start = file_text.count(b"\n", 0, b) + 1
720
file_text = file_text[b:]
677
# than a line at a time. However, we don't want this without Python 2.5
678
# as the quick string search algorithm wasn't implemented till then:
679
# <http://effbot.org/zone/stringlib.htm>
680
if sys.version_info > (2, 5):
681
i = file_text.find(pattern)
684
b = file_text.rfind("\n", 0, i) + 1
686
start = file_text.count("\n", 0, b) + 1
687
file_text = file_text[b:]
721
690
if opts.line_number:
722
691
for index, line in enumerate(file_text.splitlines()):
723
692
if pattern in line:
724
693
line = line.decode(file_encoding, 'replace')
725
writeline(lineno=index + start, line=line)
694
writeline(lineno=index+start, line=line)
727
696
for line in file_text.splitlines():
728
697
if pattern in line:
733
702
# standard cases, but perhaps could try and detect backtracking
734
703
# patterns here and avoid whole text search in those cases
735
704
search = opts.patternc.search
736
if b"$" not in pattern:
705
if "$" not in pattern:
737
706
# GZ 2010-06-05: Grr, re.MULTILINE can't save us when searching
738
707
# through revisions as bazaar returns binary mode
739
708
# and trailing \r breaks $ as line ending match
740
709
m = search(file_text)
743
b = file_text.rfind(b"\n", 0, m.start()) + 1
712
b = file_text.rfind("\n", 0, m.start()) + 1
744
713
if opts.line_number:
745
start = file_text.count(b"\n", 0, b) + 1
714
start = file_text.count("\n", 0, b) + 1
746
715
file_text = file_text[b:]
750
719
for index, line in enumerate(file_text.splitlines()):
752
721
line = line.decode(file_encoding, 'replace')
753
writeline(lineno=index + start, line=line)
722
writeline(lineno=index+start, line=line)
755
724
for line in file_text.splitlines():
757
726
line = line.decode(file_encoding, 'replace')
758
727
writeline(line=line)
added
removed
brzlib/plugins/grep/grep.py
