util/style.py

   1 #! /usr/bin/env python
   2 # Copyright (c) 2007 The Regents of The University of Michigan
   3 # All rights reserved.
   4 #
   5 # Redistribution and use in source and binary forms, with or without
   6 # modification, are permitted provided that the following conditions are
   7 # met: redistributions of source code must retain the above copyright
   8 # notice, this list of conditions and the following disclaimer;
   9 # redistributions in binary form must reproduce the above copyright
  10 # notice, this list of conditions and the following disclaimer in the
  11 # documentation and/or other materials provided with the distribution;
  12 # neither the name of the copyright holders nor the names of its
  13 # contributors may be used to endorse or promote products derived from
  14 # this software without specific prior written permission.
  15 #
  16 # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
  17 # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
  18 # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
  19 # A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
  20 # OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
  21 # SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
  22 # LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
  23 # DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
  24 # THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
  25 # (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
  26 # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  27 #
  28 # Authors: Nathan Binkert
  29
  30 import re
  31 import os
  32 import sys
  33
  34 lead = re.compile(r'^([ \t]+)')
  35 trail = re.compile(r'([ \t]+)$')
  36 any_control = re.compile(r'\b(if|while|for)[ \t]*[(]')
  37 good_control = re.compile(r'\b(if|while|for) [(]')
  38
  39 lang_types = { 'c'   : "C",
  40                'h'   : "C",
  41                'cc'  : "C++",
  42                'hh'  : "C++",
  43                'cxx' : "C++",
  44                'hxx' : "C++",
  45                'cpp' : "C++",
  46                'hpp' : "C++",
  47                'C'   : "C++",
  48                'H'   : "C++",
  49                'i'   : "swig",
  50                'py'  : "python",
  51                's'   : "asm",
  52                'S'   : "asm",
  53                'isa' : "isa" }
  54 def file_type(filename):
  55     extension = filename.split('.')
  56     extension = len(extension) > 1 and extension[-1]
  57     return lang_types.get(extension, None)
  58
  59 whitespace_types = ('C', 'C++', 'swig', 'python', 'asm', 'isa')
  60 def whitespace_file(filename):
  61     if file_type(filename) in whitespace_types:
  62         return True
  63
  64     if filename.startswith("SCons"):
  65         return True
  66
  67     return False
  68
  69 format_types = ( 'C', 'C++' )
  70 def format_file(filename):
  71     if file_type(filename) in format_types:
  72         return True
  73
  74     return False
  75
  76 def checkwhite_line(line):
  77     match = lead.search(line)
  78     if match and match.group(1).find('\t') != -1:
  79         return False
  80
  81     match = trail.search(line)
  82     if match:
  83         return False
  84
  85     return True
  86
  87 def checkwhite(filename):
  88     if not whitespace_file(filename):
  89         return
  90
  91     try:
  92         f = file(filename, 'r+')
  93     except OSError, msg:
  94         print 'could not open file %s: %s' % (filename, msg)
  95         return
  96
  97     for num,line in enumerate(f):
  98         if not checkwhite_line(line):
  99             yield line,num + 1
 100
 101 def fixwhite_line(line, tabsize):
 102     if lead.search(line):
 103         newline = ''
 104         for i,c in enumerate(line):
 105             if c == ' ':
 106                 newline += ' '
 107             elif c == '\t':
 108                 newline += ' ' * (tabsize - len(newline) % tabsize)
 109             else:
 110                 newline += line[i:]
 111                 break
 112
 113         line = newline
 114
 115     return line.rstrip() + '\n'
 116
 117 def fixwhite(filename, tabsize, fixonly=None):
 118     if not whitespace_file(filename):
 119         return
 120
 121     try:
 122         f = file(filename, 'r+')
 123     except OSError, msg:
 124         print 'could not open file %s: %s' % (filename, msg)
 125         return
 126
 127     lines = list(f)
 128
 129     f.seek(0)
 130     f.truncate()
 131
 132     for i,line in enumerate(lines):
 133         if fixonly is None or i in fixonly:
 134             line = fixwhite_line(line, tabsize)
 135
 136         print >>f, line,
 137
 138 def linelen(line):
 139     tabs = line.count('\t')
 140     if not tabs:
 141         return len(line)
 142
 143     count = 0
 144     for c in line:
 145         if c == '\t':
 146             count += tabsize - count % tabsize
 147         else:
 148             count += 1
 149
 150     return count
 151
 152 class ValidationStats(object):
 153     def __init__(self):
 154         self.toolong = 0
 155         self.toolong80 = 0
 156         self.leadtabs = 0
 157         self.trailwhite = 0
 158         self.badcontrol = 0
 159         self.cret = 0
 160
 161     def dump(self):
 162         print '''\
 163 %d violations of lines over 79 chars. %d of which are 80 chars exactly.
 164 %d cases of whitespace at the end of a line.
 165 %d cases of tabs to indent.
 166 %d bad parens after if/while/for.
 167 %d carriage returns found.
 168 ''' % (self.toolong, self.toolong80, self.trailwhite, self.leadtabs,
 169        self.badcontrol, self.cret)
 170
 171     def __nonzero__(self):
 172         return self.toolong or self.toolong80 or self.leadtabs or \
 173                self.trailwhite or self.badcontrol or self.cret
 174
 175 def validate(filename, stats, verbose, exit_code):
 176     if not format_file(filename):
 177         return
 178
 179     def msg(lineno, line, message):
 180         print '%s:%d>' % (filename, lineno + 1), message
 181         if verbose > 2:
 182             print line
 183
 184     def bad():
 185         if exit_code is not None:
 186             sys.exit(exit_code)
 187
 188     cpp = filename.endswith('.cc') or filename.endswith('.hh')
 189     py = filename.endswith('.py')
 190
 191     if py + cpp != 1:
 192         raise AttributeError, \
 193               "I don't know how to deal with the file %s" % filename
 194
 195     try:
 196         f = file(filename, 'r')
 197     except OSError:
 198         if verbose > 0:
 199             print 'could not open file %s' % filename
 200         bad()
 201         return
 202
 203     for i,line in enumerate(f):
 204         line = line.rstrip('\n')
 205
 206         # no carriage returns
 207         if line.find('\r') != -1:
 208             self.cret += 1
 209             if verbose > 1:
 210                 msg(i, line, 'carriage return found')
 211             bad()
 212
 213         # lines max out at 79 chars
 214         llen = linelen(line)
 215         if llen > 79:
 216             stats.toolong += 1
 217             if llen == 80:
 218                 stats.toolong80 += 1
 219             if verbose > 1:
 220                 msg(i, line, 'line too long (%d chars)' % llen)
 221             bad()
 222
 223         # no tabs used to indent
 224         match = lead.search(line)
 225         if match and match.group(1).find('\t') != -1:
 226             stats.leadtabs += 1
 227             if verbose > 1:
 228                 msg(i, line, 'using tabs to indent')
 229             bad()
 230
 231         # no trailing whitespace
 232         if trail.search(line):
 233             stats.trailwhite +=1
 234             if verbose > 1:
 235                 msg(i, line, 'trailing whitespace')
 236             bad()
 237
 238         # for c++, exactly one space betwen if/while/for and (
 239         if cpp:
 240             match = any_control.search(line)
 241             if match and not good_control.search(line):
 242                 stats.badcontrol += 1
 243                 if verbose > 1:
 244                     msg(i, line, 'improper spacing after %s' % match.group(1))
 245                 bad()
 246
 247 def modified_lines(old_data, new_data):
 248     from itertools import count
 249     from mercurial import bdiff, mdiff
 250
 251     modified = set()
 252     counter = count()
 253     for pbeg, pend, fbeg, fend in bdiff.blocks(old_data, new_data):
 254         for i in counter:
 255             if i < fbeg:
 256                 modified.add(i)
 257             elif i + 1 >= fend:
 258                 break
 259     return modified
 260
 261 def check_whitespace(ui, repo, hooktype, node, parent1, parent2):
 262     from mercurial import mdiff
 263
 264     if hooktype != 'pretxncommit':
 265         raise AttributeError, \
 266               "This hook is only meant for pretxncommit, not %s" % hooktype
 267
 268     tabsize = 8
 269     verbose = ui.configbool('style', 'verbose', False)
 270     def prompt(name, fixonly=None):
 271         result = ui.prompt("(a)bort, (i)gnore, or (f)ix?", "^[aif]$", "a")
 272         if result == 'a':
 273             return True
 274         elif result == 'i':
 275             pass
 276         elif result == 'f':
 277             fixwhite(repo.wjoin(name), tabsize, fixonly)
 278         else:
 279             raise RepoError, "Invalid response: '%s'" % result
 280
 281         return False
 282
 283     modified, added, removed, deleted, unknown, ignore, clean = repo.status()
 284
 285     for fname in added:
 286         ok = True
 287         for line,num in checkwhite(fname):
 288             ui.write("invalid whitespace in %s:%d\n" % (fname, num))
 289             if verbose:
 290                 ui.write(">>%s<<\n" % line[-1])
 291             ok = False
 292
 293         if not ok:
 294             if prompt(fname):
 295                 return True
 296
 297     wctx = repo.workingctx()
 298     for fname in modified:
 299         fctx = wctx.filectx(fname)
 300         pctx = fctx.parents()
 301         assert len(pctx) in (1, 2)
 302
 303         file_data = fctx.data()
 304         mod_lines = modified_lines(pctx[0].data(), file_data)
 305         if len(pctx) == 2:
 306             m2 = modified_lines(pctx[1].data(), file_data)
 307             mod_lines = mod_lines & m2 # only the lines that are new in both
 308
 309         fixonly = set()
 310         for i,line in enumerate(mdiff.splitnewlines(file_data)):
 311             if i not in mod_lines:
 312                 continue
 313
 314             if checkwhite_line(line):
 315                 continue
 316
 317             ui.write("invalid whitespace: %s:%d\n" % (fname, i+1))
 318             if verbose:
 319                 ui.write(">>%s<<\n" % line[:-1])
 320             fixonly.add(i)
 321
 322         if fixonly:
 323             if prompt(fname, fixonly):
 324                 return True
 325
 326 def check_format(ui, repo, hooktype, node, parent1, parent2):
 327     if hooktype != 'pretxncommit':
 328         raise AttributeError, \
 329               "This hook is only meant for pretxncommit, not %s" % hooktype
 330
 331     modified, added, removed, deleted, unknown, ignore, clean = repo.status()
 332
 333     verbose = 0
 334     stats = ValidationStats()
 335     for f in modified + added:
 336         validate(f, stats, verbose, None)
 337
 338     if stats:
 339         stats.dump()
 340         result = ui.prompt("invalid formatting\n(i)gnore or (a)bort?",
 341                            "^[ia]$", "a")
 342         if result.startswith('i'):
 343             pass
 344         elif result.startswith('a'):
 345             return True
 346         else:
 347             raise RepoError, "Invalid response: '%s'" % result
 348
 349     return False
 350
 351 if __name__ == '__main__':
 352     import getopt
 353
 354     progname = sys.argv[0]
 355     if len(sys.argv) < 2:
 356         sys.exit('usage: %s <command> [<command args>]' % progname)
 357
 358     fixwhite_usage = '%s fixwhite [-t <tabsize> ] <path> [...] \n' % progname
 359     chkformat_usage = '%s chkformat <path> [...] \n' % progname
 360     chkwhite_usage = '%s chkwhite <path> [...] \n' % progname
 361
 362     command = sys.argv[1]
 363     if command == 'fixwhite':
 364         flags = 't:'
 365         usage = fixwhite_usage
 366     elif command == 'chkwhite':
 367         flags = 'nv'
 368         usage = chkwhite_usage
 369     elif command == 'chkformat':
 370         flags = 'nv'
 371         usage = chkformat_usage
 372     else:
 373         sys.exit(fixwhite_usage + chkwhite_usage + chkformat_usage)
 374
 375     opts, args = getopt.getopt(sys.argv[2:], flags)
 376
 377     code = 1
 378     verbose = 1
 379     tabsize = 8
 380     for opt,arg in opts:
 381         if opt == '-n':
 382             code = None
 383         if opt == '-t':
 384             tabsize = int(arg)
 385         if opt == '-v':
 386             verbose += 1
 387
 388     if command == 'fixwhite':
 389         for filename in args:
 390             fixwhite(filename, tabsize)
 391     elif command == 'chkwhite':
 392         for filename in args:
 393             for line,num in checkwhite(filename):
 394                 print 'invalid whitespace: %s:%d' % (filename, num)
 395                 if verbose:
 396                     print '>>%s<<' % line[:-1]
 397     elif command == 'chkformat':
 398         stats = ValidationStats()
 399         for filename in files:
 400             validate(filename, stats=stats, verbose=verbose, exit_code=code)
 401
 402         if verbose > 0:
 403             stats.dump()
 404     else:
 405         sys.exit("command '%s' not found" % command)