1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
|
# Copyright (c) 2011 The Chromium OS Authors.
#
# SPDX-License-Identifier: GPL-2.0+
#
import math
import os
import re
import shutil
import tempfile
import command
import commit
import gitutil
from series import Series
# Tags that we detect and remove
re_remove = re.compile('^BUG=|^TEST=|^BRANCH=|^Change-Id:|^Review URL:'
'|Reviewed-on:|Commit-\w*:')
# Lines which are allowed after a TEST= line
re_allowed_after_test = re.compile('^Signed-off-by:')
# Signoffs
re_signoff = re.compile('^Signed-off-by: *(.*)')
# The start of the cover letter
re_cover = re.compile('^Cover-letter:')
# A cover letter Cc
re_cover_cc = re.compile('^Cover-letter-cc: *(.*)')
# Patch series tag
re_series_tag = re.compile('^Series-([a-z-]*): *(.*)')
# Commit series tag
re_commit_tag = re.compile('^Commit-([a-z-]*): *(.*)')
# Commit tags that we want to collect and keep
re_tag = re.compile('^(Tested-by|Acked-by|Reviewed-by|Patch-cc): (.*)')
# The start of a new commit in the git log
re_commit = re.compile('^commit ([0-9a-f]*)$')
# We detect these since checkpatch doesn't always do it
re_space_before_tab = re.compile('^[+].* \t')
# States we can be in - can we use range() and still have comments?
STATE_MSG_HEADER = 0 # Still in the message header
STATE_PATCH_SUBJECT = 1 # In patch subject (first line of log for a commit)
STATE_PATCH_HEADER = 2 # In patch header (after the subject)
STATE_DIFFS = 3 # In the diff part (past --- line)
class PatchStream:
"""Class for detecting/injecting tags in a patch or series of patches
We support processing the output of 'git log' to read out the tags we
are interested in. We can also process a patch file in order to remove
unwanted tags or inject additional ones. These correspond to the two
phases of processing.
"""
def __init__(self, series, name=None, is_log=False):
self.skip_blank = False # True to skip a single blank line
self.found_test = False # Found a TEST= line
self.lines_after_test = 0 # MNumber of lines found after TEST=
self.warn = [] # List of warnings we have collected
self.linenum = 1 # Output line number we are up to
self.in_section = None # Name of start...END section we are in
self.notes = [] # Series notes
self.section = [] # The current section...END section
self.series = series # Info about the patch series
self.is_log = is_log # True if indent like git log
self.in_change = 0 # Non-zero if we are in a change list
self.blank_count = 0 # Number of blank lines stored up
self.state = STATE_MSG_HEADER # What state are we in?
self.signoff = [] # Contents of signoff line
self.commit = None # Current commit
def AddToSeries(self, line, name, value):
"""Add a new Series-xxx tag.
When a Series-xxx tag is detected, we come here to record it, if we
are scanning a 'git log'.
Args:
line: Source line containing tag (useful for debug/error messages)
name: Tag name (part after 'Series-')
value: Tag value (part after 'Series-xxx: ')
"""
if name == 'notes':
self.in_section = name
self.skip_blank = False
if self.is_log:
self.series.AddTag(self.commit, line, name, value)
def AddToCommit(self, line, name, value):
"""Add a new Commit-xxx tag.
When a Commit-xxx tag is detected, we come here to record it.
Args:
line: Source line containing tag (useful for debug/error messages)
name: Tag name (part after 'Commit-')
value: Tag value (part after 'Commit-xxx: ')
"""
if name == 'notes':
self.in_section = 'commit-' + name
self.skip_blank = False
def CloseCommit(self):
"""Save the current commit into our commit list, and reset our state"""
if self.commit and self.is_log:
self.series.AddCommit(self.commit)
self.commit = None
def ProcessLine(self, line):
"""Process a single line of a patch file or commit log
This process a line and returns a list of lines to output. The list
may be empty or may contain multiple output lines.
This is where all the complicated logic is located. The class's
state is used to move between different states and detect things
properly.
We can be in one of two modes:
self.is_log == True: This is 'git log' mode, where most output is
indented by 4 characters and we are scanning for tags
self.is_log == False: This is 'patch' mode, where we already have
all the tags, and are processing patches to remove junk we
don't want, and add things we think are required.
Args:
line: text line to process
Returns:
list of output lines, or [] if nothing should be output
"""
# Initially we have no output. Prepare the input line string
out = []
line = line.rstrip('\n')
commit_match = re_commit.match(line) if self.is_log else None
if self.is_log:
if line[:4] == ' ':
line = line[4:]
# Handle state transition and skipping blank lines
series_tag_match = re_series_tag.match(line)
commit_tag_match = re_commit_tag.match(line)
cover_match = re_cover.match(line)
cover_cc_match = re_cover_cc.match(line)
signoff_match = re_signoff.match(line)
tag_match = None
if self.state == STATE_PATCH_HEADER:
tag_match = re_tag.match(line)
is_blank = not line.strip()
if is_blank:
if (self.state == STATE_MSG_HEADER
or self.state == STATE_PATCH_SUBJECT):
self.state += 1
# We don't have a subject in the text stream of patch files
# It has its own line with a Subject: tag
if not self.is_log and self.state == STATE_PATCH_SUBJECT:
self.state += 1
elif commit_match:
self.state = STATE_MSG_HEADER
# If a tag is detected, but we are already in a section,
# this means 'END' is missing for that section, fix it up.
if series_tag_match or commit_tag_match or \
cover_match or cover_cc_match or signoff_match:
if self.in_section:
self.warn.append("Missing 'END' in section '%s'" % self.in_section)
if self.in_section == 'cover':
self.series.cover = self.section
elif self.in_section == 'notes':
if self.is_log:
self.series.notes += self.section
elif self.in_section == 'commit-notes':
if self.is_log:
self.commit.notes += self.section
else:
self.warn.append("Unknown section '%s'" % self.in_section)
self.in_section = None
self.skip_blank = True
self.section = []
# If we are in a section, keep collecting lines until we see END
if self.in_section:
if line == 'END':
if self.in_section == 'cover':
self.series.cover = self.section
elif self.in_section == 'notes':
if self.is_log:
self.series.notes += self.section
elif self.in_section == 'commit-notes':
if self.is_log:
self.commit.notes += self.section
else:
self.warn.append("Unknown section '%s'" % self.in_section)
self.in_section = None
self.skip_blank = True
self.section = []
else:
self.section.append(line)
# Detect the commit subject
elif not is_blank and self.state == STATE_PATCH_SUBJECT:
self.commit.subject = line
# Detect the tags we want to remove, and skip blank lines
elif re_remove.match(line) and not commit_tag_match:
self.skip_blank = True
# TEST= should be the last thing in the commit, so remove
# everything after it
if line.startswith('TEST='):
self.found_test = True
elif self.skip_blank and is_blank:
self.skip_blank = False
# Detect the start of a cover letter section
elif cover_match:
self.in_section = 'cover'
self.skip_blank = False
elif cover_cc_match:
value = cover_cc_match.group(1)
self.AddToSeries(line, 'cover-cc', value)
# If we are in a change list, key collected lines until a blank one
elif self.in_change:
if is_blank:
# Blank line ends this change list
self.in_change = 0
elif line == '---':
self.in_change = 0
out = self.ProcessLine(line)
else:
if self.is_log:
self.series.AddChange(self.in_change, self.commit, line)
self.skip_blank = False
# Detect Series-xxx tags
elif series_tag_match:
name = series_tag_match.group(1)
value = series_tag_match.group(2)
if name == 'changes':
# value is the version number: e.g. 1, or 2
try:
value = int(value)
except ValueError as str:
raise ValueError("%s: Cannot decode version info '%s'" %
(self.commit.hash, line))
self.in_change = int(value)
else:
self.AddToSeries(line, name, value)
self.skip_blank = True
# Detect Commit-xxx tags
elif commit_tag_match:
name = commit_tag_match.group(1)
value = commit_tag_match.group(2)
if name == 'notes':
self.AddToCommit(line, name, value)
self.skip_blank = True
# Detect the start of a new commit
elif commit_match:
self.CloseCommit()
self.commit = commit.Commit(commit_match.group(1))
# Detect tags in the commit message
elif tag_match:
# Remove Tested-by self, since few will take much notice
if (tag_match.group(1) == 'Tested-by' and
tag_match.group(2).find(os.getenv('USER') + '@') != -1):
self.warn.append("Ignoring %s" % line)
elif tag_match.group(1) == 'Patch-cc':
self.commit.AddCc(tag_match.group(2).split(','))
else:
out = [line]
# Suppress duplicate signoffs
elif signoff_match:
if (self.is_log or not self.commit or
self.commit.CheckDuplicateSignoff(signoff_match.group(1))):
out = [line]
# Well that means this is an ordinary line
else:
pos = 1
# Look for ugly ASCII characters
for ch in line:
# TODO: Would be nicer to report source filename and line
if ord(ch) > 0x80:
self.warn.append("Line %d/%d ('%s') has funny ascii char" %
(self.linenum, pos, line))
pos += 1
# Look for space before tab
m = re_space_before_tab.match(line)
if m:
self.warn.append('Line %d/%d has space before tab' %
(self.linenum, m.start()))
# OK, we have a valid non-blank line
out = [line]
self.linenum += 1
self.skip_blank = False
if self.state == STATE_DIFFS:
pass
# If this is the start of the diffs section, emit our tags and
# change log
elif line == '---':
self.state = STATE_DIFFS
# Output the tags (signeoff first), then change list
out = []
log = self.series.MakeChangeLog(self.commit)
out += [line]
if self.commit:
out += self.commit.notes
out += [''] + log
elif self.found_test:
if not re_allowed_after_test.match(line):
self.lines_after_test += 1
return out
def Finalize(self):
"""Close out processing of this patch stream"""
self.CloseCommit()
if self.lines_after_test:
self.warn.append('Found %d lines after TEST=' %
self.lines_after_test)
def ProcessStream(self, infd, outfd):
"""Copy a stream from infd to outfd, filtering out unwanting things.
This is used to process patch files one at a time.
Args:
infd: Input stream file object
outfd: Output stream file object
"""
# Extract the filename from each diff, for nice warnings
fname = None
last_fname = None
re_fname = re.compile('diff --git a/(.*) b/.*')
while True:
line = infd.readline()
if not line:
break
out = self.ProcessLine(line)
# Try to detect blank lines at EOF
for line in out:
match = re_fname.match(line)
if match:
last_fname = fname
fname = match.group(1)
if line == '+':
self.blank_count += 1
else:
if self.blank_count and (line == '-- ' or match):
self.warn.append("Found possible blank line(s) at "
"end of file '%s'" % last_fname)
outfd.write('+\n' * self.blank_count)
outfd.write(line + '\n')
self.blank_count = 0
self.Finalize()
def GetMetaDataForList(commit_range, git_dir=None, count=None,
series = None, allow_overwrite=False):
"""Reads out patch series metadata from the commits
This does a 'git log' on the relevant commits and pulls out the tags we
are interested in.
Args:
commit_range: Range of commits to count (e.g. 'HEAD..base')
git_dir: Path to git repositiory (None to use default)
count: Number of commits to list, or None for no limit
series: Series object to add information into. By default a new series
is started.
allow_overwrite: Allow tags to overwrite an existing tag
Returns:
A Series object containing information about the commits.
"""
if not series:
series = Series()
series.allow_overwrite = allow_overwrite
params = gitutil.LogCmd(commit_range, reverse=True, count=count,
git_dir=git_dir)
stdout = command.RunPipe([params], capture=True).stdout
ps = PatchStream(series, is_log=True)
for line in stdout.splitlines():
ps.ProcessLine(line)
ps.Finalize()
return series
def GetMetaData(start, count):
"""Reads out patch series metadata from the commits
This does a 'git log' on the relevant commits and pulls out the tags we
are interested in.
Args:
start: Commit to start from: 0=HEAD, 1=next one, etc.
count: Number of commits to list
"""
return GetMetaDataForList('HEAD~%d' % start, None, count)
def FixPatch(backup_dir, fname, series, commit):
"""Fix up a patch file, by adding/removing as required.
We remove our tags from the patch file, insert changes lists, etc.
The patch file is processed in place, and overwritten.
A backup file is put into backup_dir (if not None).
Args:
fname: Filename to patch file to process
series: Series information about this patch set
commit: Commit object for this patch file
Return:
A list of errors, or [] if all ok.
"""
handle, tmpname = tempfile.mkstemp()
outfd = os.fdopen(handle, 'w')
infd = open(fname, 'r')
ps = PatchStream(series)
ps.commit = commit
ps.ProcessStream(infd, outfd)
infd.close()
outfd.close()
# Create a backup file if required
if backup_dir:
shutil.copy(fname, os.path.join(backup_dir, os.path.basename(fname)))
shutil.move(tmpname, fname)
return ps.warn
def FixPatches(series, fnames):
"""Fix up a list of patches identified by filenames
The patch files are processed in place, and overwritten.
Args:
series: The series object
fnames: List of patch files to process
"""
# Current workflow creates patches, so we shouldn't need a backup
backup_dir = None #tempfile.mkdtemp('clean-patch')
count = 0
for fname in fnames:
commit = series.commits[count]
commit.patch = fname
result = FixPatch(backup_dir, fname, series, commit)
if result:
print '%d warnings for %s:' % (len(result), fname)
for warn in result:
print '\t', warn
print
count += 1
print 'Cleaned %d patches' % count
return series
def InsertCoverLetter(fname, series, count):
"""Inserts a cover letter with the required info into patch 0
Args:
fname: Input / output filename of the cover letter file
series: Series object
count: Number of patches in the series
"""
fd = open(fname, 'r')
lines = fd.readlines()
fd.close()
fd = open(fname, 'w')
text = series.cover
prefix = series.GetPatchPrefix()
for line in lines:
if line.startswith('Subject:'):
# if more than 10 or 100 patches, it should say 00/xx, 000/xxx, etc
zero_repeat = int(math.log10(count)) + 1
zero = '0' * zero_repeat
line = 'Subject: [%s %s/%d] %s\n' % (prefix, zero, count, text[0])
# Insert our cover letter
elif line.startswith('*** BLURB HERE ***'):
# First the blurb test
line = '\n'.join(text[1:]) + '\n'
if series.get('notes'):
line += '\n'.join(series.notes) + '\n'
# Now the change list
out = series.MakeChangeLog(None)
line += '\n' + '\n'.join(out)
fd.write(line)
fd.close()
|