1# Copyright (c) 2011 The Chromium OS Authors. 2# 3# See file CREDITS for list of people who contributed to this 4# project. 5# 6# This program is free software; you can redistribute it and/or 7# modify it under the terms of the GNU General Public License as 8# published by the Free Software Foundation; either version 2 of 9# the License, or (at your option) any later version. 10# 11# This program is distributed in the hope that it will be useful, 12# but WITHOUT ANY WARRANTY; without even the implied warranty of 13# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 14# GNU General Public License for more details. 15# 16# You should have received a copy of the GNU General Public License 17# along with this program; if not, write to the Free Software 18# Foundation, Inc., 59 Temple Place, Suite 330, Boston, 19# MA 02111-1307 USA 20# 21 22import os 23import re 24import shutil 25import tempfile 26 27import command 28import commit 29import gitutil 30from series import Series 31 32# Tags that we detect and remove 33re_remove = re.compile('^BUG=|^TEST=|^Change-Id:|^Review URL:' 34 '|Reviewed-on:|Reviewed-by:') 35 36# Lines which are allowed after a TEST= line 37re_allowed_after_test = re.compile('^Signed-off-by:') 38 39# Signoffs 40re_signoff = re.compile('^Signed-off-by:') 41 42# The start of the cover letter 43re_cover = re.compile('^Cover-letter:') 44 45# Patch series tag 46re_series = re.compile('^Series-(\w*): *(.*)') 47 48# Commit tags that we want to collect and keep 49re_tag = re.compile('^(Tested-by|Acked-by|Signed-off-by|Cc): (.*)') 50 51# The start of a new commit in the git log 52re_commit = re.compile('^commit (.*)') 53 54# We detect these since checkpatch doesn't always do it 55re_space_before_tab = re.compile('^[+].* \t') 56 57# States we can be in - can we use range() and still have comments? 58STATE_MSG_HEADER = 0 # Still in the message header 59STATE_PATCH_SUBJECT = 1 # In patch subject (first line of log for a commit) 60STATE_PATCH_HEADER = 2 # In patch header (after the subject) 61STATE_DIFFS = 3 # In the diff part (past --- line) 62 63class PatchStream: 64 """Class for detecting/injecting tags in a patch or series of patches 65 66 We support processing the output of 'git log' to read out the tags we 67 are interested in. We can also process a patch file in order to remove 68 unwanted tags or inject additional ones. These correspond to the two 69 phases of processing. 70 """ 71 def __init__(self, series, name=None, is_log=False): 72 self.skip_blank = False # True to skip a single blank line 73 self.found_test = False # Found a TEST= line 74 self.lines_after_test = 0 # MNumber of lines found after TEST= 75 self.warn = [] # List of warnings we have collected 76 self.linenum = 1 # Output line number we are up to 77 self.in_section = None # Name of start...END section we are in 78 self.notes = [] # Series notes 79 self.section = [] # The current section...END section 80 self.series = series # Info about the patch series 81 self.is_log = is_log # True if indent like git log 82 self.in_change = 0 # Non-zero if we are in a change list 83 self.blank_count = 0 # Number of blank lines stored up 84 self.state = STATE_MSG_HEADER # What state are we in? 85 self.tags = [] # Tags collected, like Tested-by... 86 self.signoff = [] # Contents of signoff line 87 self.commit = None # Current commit 88 89 def AddToSeries(self, line, name, value): 90 """Add a new Series-xxx tag. 91 92 When a Series-xxx tag is detected, we come here to record it, if we 93 are scanning a 'git log'. 94 95 Args: 96 line: Source line containing tag (useful for debug/error messages) 97 name: Tag name (part after 'Series-') 98 value: Tag value (part after 'Series-xxx: ') 99 """ 100 if name == 'notes': 101 self.in_section = name 102 self.skip_blank = False 103 if self.is_log: 104 self.series.AddTag(self.commit, line, name, value) 105 106 def CloseCommit(self): 107 """Save the current commit into our commit list, and reset our state""" 108 if self.commit and self.is_log: 109 self.series.AddCommit(self.commit) 110 self.commit = None 111 112 def FormatTags(self, tags): 113 out_list = [] 114 for tag in sorted(tags): 115 if tag.startswith('Cc:'): 116 tag_list = tag[4:].split(',') 117 out_list += gitutil.BuildEmailList(tag_list, 'Cc:') 118 else: 119 out_list.append(tag) 120 return out_list 121 122 def ProcessLine(self, line): 123 """Process a single line of a patch file or commit log 124 125 This process a line and returns a list of lines to output. The list 126 may be empty or may contain multiple output lines. 127 128 This is where all the complicated logic is located. The class's 129 state is used to move between different states and detect things 130 properly. 131 132 We can be in one of two modes: 133 self.is_log == True: This is 'git log' mode, where most output is 134 indented by 4 characters and we are scanning for tags 135 136 self.is_log == False: This is 'patch' mode, where we already have 137 all the tags, and are processing patches to remove junk we 138 don't want, and add things we think are required. 139 140 Args: 141 line: text line to process 142 143 Returns: 144 list of output lines, or [] if nothing should be output 145 """ 146 # Initially we have no output. Prepare the input line string 147 out = [] 148 line = line.rstrip('\n') 149 if self.is_log: 150 if line[:4] == ' ': 151 line = line[4:] 152 153 # Handle state transition and skipping blank lines 154 series_match = re_series.match(line) 155 commit_match = re_commit.match(line) if self.is_log else None 156 tag_match = None 157 if self.state == STATE_PATCH_HEADER: 158 tag_match = re_tag.match(line) 159 is_blank = not line.strip() 160 if is_blank: 161 if (self.state == STATE_MSG_HEADER 162 or self.state == STATE_PATCH_SUBJECT): 163 self.state += 1 164 165 # We don't have a subject in the text stream of patch files 166 # It has its own line with a Subject: tag 167 if not self.is_log and self.state == STATE_PATCH_SUBJECT: 168 self.state += 1 169 elif commit_match: 170 self.state = STATE_MSG_HEADER 171 172 # If we are in a section, keep collecting lines until we see END 173 if self.in_section: 174 if line == 'END': 175 if self.in_section == 'cover': 176 self.series.cover = self.section 177 elif self.in_section == 'notes': 178 if self.is_log: 179 self.series.notes += self.section 180 else: 181 self.warn.append("Unknown section '%s'" % self.in_section) 182 self.in_section = None 183 self.skip_blank = True 184 self.section = [] 185 else: 186 self.section.append(line) 187 188 # Detect the commit subject 189 elif not is_blank and self.state == STATE_PATCH_SUBJECT: 190 self.commit.subject = line 191 192 # Detect the tags we want to remove, and skip blank lines 193 elif re_remove.match(line): 194 self.skip_blank = True 195 196 # TEST= should be the last thing in the commit, so remove 197 # everything after it 198 if line.startswith('TEST='): 199 self.found_test = True 200 elif self.skip_blank and is_blank: 201 self.skip_blank = False 202 203 # Detect the start of a cover letter section 204 elif re_cover.match(line): 205 self.in_section = 'cover' 206 self.skip_blank = False 207 208 # If we are in a change list, key collected lines until a blank one 209 elif self.in_change: 210 if is_blank: 211 # Blank line ends this change list 212 self.in_change = 0 213 elif line == '---' or re_signoff.match(line): 214 self.in_change = 0 215 out = self.ProcessLine(line) 216 else: 217 self.series.AddChange(self.in_change, self.commit, line) 218 self.skip_blank = False 219 220 # Detect Series-xxx tags 221 elif series_match: 222 name = series_match.group(1) 223 value = series_match.group(2) 224 if name == 'changes': 225 # value is the version number: e.g. 1, or 2 226 try: 227 value = int(value) 228 except ValueError as str: 229 raise ValueError("%s: Cannot decode version info '%s'" % 230 (self.commit.hash, line)) 231 self.in_change = int(value) 232 else: 233 self.AddToSeries(line, name, value) 234 self.skip_blank = True 235 236 # Detect the start of a new commit 237 elif commit_match: 238 self.CloseCommit() 239 self.commit = commit.Commit(commit_match.group(1)[:7]) 240 241 # Detect tags in the commit message 242 elif tag_match: 243 # Onlly allow a single signoff tag 244 if tag_match.group(1) == 'Signed-off-by': 245 if self.signoff: 246 self.warn.append('Patch has more than one Signed-off-by ' 247 'tag') 248 self.signoff += [line] 249 250 # Remove Tested-by self, since few will take much notice 251 elif (tag_match.group(1) == 'Tested-by' and 252 tag_match.group(2).find(os.getenv('USER') + '@') != -1): 253 self.warn.append("Ignoring %s" % line) 254 elif tag_match.group(1) == 'Cc': 255 self.commit.AddCc(tag_match.group(2).split(',')) 256 else: 257 self.tags.append(line); 258 259 # Well that means this is an ordinary line 260 else: 261 pos = 1 262 # Look for ugly ASCII characters 263 for ch in line: 264 # TODO: Would be nicer to report source filename and line 265 if ord(ch) > 0x80: 266 self.warn.append("Line %d/%d ('%s') has funny ascii char" % 267 (self.linenum, pos, line)) 268 pos += 1 269 270 # Look for space before tab 271 m = re_space_before_tab.match(line) 272 if m: 273 self.warn.append('Line %d/%d has space before tab' % 274 (self.linenum, m.start())) 275 276 # OK, we have a valid non-blank line 277 out = [line] 278 self.linenum += 1 279 self.skip_blank = False 280 if self.state == STATE_DIFFS: 281 pass 282 283 # If this is the start of the diffs section, emit our tags and 284 # change log 285 elif line == '---': 286 self.state = STATE_DIFFS 287 288 # Output the tags (signeoff first), then change list 289 out = [] 290 if self.signoff: 291 out += self.signoff 292 log = self.series.MakeChangeLog(self.commit) 293 out += self.FormatTags(self.tags) 294 out += [line] + log 295 elif self.found_test: 296 if not re_allowed_after_test.match(line): 297 self.lines_after_test += 1 298 299 return out 300 301 def Finalize(self): 302 """Close out processing of this patch stream""" 303 self.CloseCommit() 304 if self.lines_after_test: 305 self.warn.append('Found %d lines after TEST=' % 306 self.lines_after_test) 307 308 def ProcessStream(self, infd, outfd): 309 """Copy a stream from infd to outfd, filtering out unwanting things. 310 311 This is used to process patch files one at a time. 312 313 Args: 314 infd: Input stream file object 315 outfd: Output stream file object 316 """ 317 # Extract the filename from each diff, for nice warnings 318 fname = None 319 last_fname = None 320 re_fname = re.compile('diff --git a/(.*) b/.*') 321 while True: 322 line = infd.readline() 323 if not line: 324 break 325 out = self.ProcessLine(line) 326 327 # Try to detect blank lines at EOF 328 for line in out: 329 match = re_fname.match(line) 330 if match: 331 last_fname = fname 332 fname = match.group(1) 333 if line == '+': 334 self.blank_count += 1 335 else: 336 if self.blank_count and (line == '-- ' or match): 337 self.warn.append("Found possible blank line(s) at " 338 "end of file '%s'" % last_fname) 339 outfd.write('+\n' * self.blank_count) 340 outfd.write(line + '\n') 341 self.blank_count = 0 342 self.Finalize() 343 344 345def GetMetaData(start, count): 346 """Reads out patch series metadata from the commits 347 348 This does a 'git log' on the relevant commits and pulls out the tags we 349 are interested in. 350 351 Args: 352 start: Commit to start from: 0=HEAD, 1=next one, etc. 353 count: Number of commits to list 354 """ 355 pipe = [['git', 'log', '--reverse', 'HEAD~%d' % start, '-n%d' % count]] 356 stdout = command.RunPipe(pipe, capture=True) 357 series = Series() 358 ps = PatchStream(series, is_log=True) 359 for line in stdout.splitlines(): 360 ps.ProcessLine(line) 361 ps.Finalize() 362 return series 363 364def FixPatch(backup_dir, fname, series, commit): 365 """Fix up a patch file, by adding/removing as required. 366 367 We remove our tags from the patch file, insert changes lists, etc. 368 The patch file is processed in place, and overwritten. 369 370 A backup file is put into backup_dir (if not None). 371 372 Args: 373 fname: Filename to patch file to process 374 series: Series information about this patch set 375 commit: Commit object for this patch file 376 Return: 377 A list of errors, or [] if all ok. 378 """ 379 handle, tmpname = tempfile.mkstemp() 380 outfd = os.fdopen(handle, 'w') 381 infd = open(fname, 'r') 382 ps = PatchStream(series) 383 ps.commit = commit 384 ps.ProcessStream(infd, outfd) 385 infd.close() 386 outfd.close() 387 388 # Create a backup file if required 389 if backup_dir: 390 shutil.copy(fname, os.path.join(backup_dir, os.path.basename(fname))) 391 shutil.move(tmpname, fname) 392 return ps.warn 393 394def FixPatches(series, fnames): 395 """Fix up a list of patches identified by filenames 396 397 The patch files are processed in place, and overwritten. 398 399 Args: 400 series: The series object 401 fnames: List of patch files to process 402 """ 403 # Current workflow creates patches, so we shouldn't need a backup 404 backup_dir = None #tempfile.mkdtemp('clean-patch') 405 count = 0 406 for fname in fnames: 407 commit = series.commits[count] 408 commit.patch = fname 409 result = FixPatch(backup_dir, fname, series, commit) 410 if result: 411 print '%d warnings for %s:' % (len(result), fname) 412 for warn in result: 413 print '\t', warn 414 print 415 count += 1 416 print 'Cleaned %d patches' % count 417 return series 418 419def InsertCoverLetter(fname, series, count): 420 """Inserts a cover letter with the required info into patch 0 421 422 Args: 423 fname: Input / output filename of the cover letter file 424 series: Series object 425 count: Number of patches in the series 426 """ 427 fd = open(fname, 'r') 428 lines = fd.readlines() 429 fd.close() 430 431 fd = open(fname, 'w') 432 text = series.cover 433 prefix = series.GetPatchPrefix() 434 for line in lines: 435 if line.startswith('Subject:'): 436 # TODO: if more than 10 patches this should save 00/xx, not 0/xx 437 line = 'Subject: [%s 0/%d] %s\n' % (prefix, count, text[0]) 438 439 # Insert our cover letter 440 elif line.startswith('*** BLURB HERE ***'): 441 # First the blurb test 442 line = '\n'.join(text[1:]) + '\n' 443 if series.get('notes'): 444 line += '\n'.join(series.notes) + '\n' 445 446 # Now the change list 447 out = series.MakeChangeLog(None) 448 line += '\n' + '\n'.join(out) 449 fd.write(line) 450 fd.close() 451