Tools/scripts/fixcid.py

   1 #! /usr/bin/env python
   2
   3 # Perform massive identifier substitution on C source files.
   4 # This actually tokenizes the files (to some extent) so it can
   5 # avoid making substitutions inside strings or comments.
   6 # Inside strings, substitutions are never made; inside comments,
   7 # it is a user option (off by default).
   8 #
   9 # The substitutions are read from one or more files whose lines,
  10 # when not empty, after stripping comments starting with #,
  11 # must contain exactly two words separated by whitespace: the
  12 # old identifier and its replacement.
  13 #
  14 # The option -r reverses the sense of the substitutions (this may be
  15 # useful to undo a particular substitution).
  16 #
  17 # If the old identifier is prefixed with a '*' (with no intervening
  18 # whitespace), then it will not be substituted inside comments.
  19 #
  20 # Command line arguments are files or directories to be processed.
  21 # Directories are searched recursively for files whose name looks
  22 # like a C file (ends in .h or .c).  The special filename '-' means
  23 # operate in filter mode: read stdin, write stdout.
  24 #
  25 # Symbolic links are always ignored (except as explicit directory
  26 # arguments).
  27 #
  28 # The original files are kept as back-up with a "~" suffix.
  29 #
  30 # Changes made are reported to stdout in a diff-like format.
  31 #
  32 # NB: by changing only the function fixline() you can turn this
  33 # into a program for different changes to C source files; by
  34 # changing the function wanted() you can make a different selection of
  35 # files.
  36
  37 import sys
  38 import regex
  39 import string
  40 import os
  41 from stat import *
  42 import getopt
  43
  44 err = sys.stderr.write
  45 dbg = err
  46 rep = sys.stdout.write
  47
  48 def usage():
  49     progname = sys.argv[0]
  50     err('Usage: ' + progname +
  51               ' [-c] [-r] [-s file] ... file-or-directory ...\n')
  52     err('\n')
  53     err('-c           : substitute inside comments\n')
  54     err('-r           : reverse direction for following -s options\n')
  55     err('-s substfile : add a file of substitutions\n')
  56     err('\n')
  57     err('Each non-empty non-comment line in a substitution file must\n')
  58     err('contain exactly two words: an identifier and its replacement.\n')
  59     err('Comments start with a # character and end at end of line.\n')
  60     err('If an identifier is preceded with a *, it is not substituted\n')
  61     err('inside a comment even when -c is specified.\n')
  62
  63 def main():
  64     try:
  65         opts, args = getopt.getopt(sys.argv[1:], 'crs:')
  66     except getopt.error, msg:
  67         err('Options error: ' + str(msg) + '\n')
  68         usage()
  69         sys.exit(2)
  70     bad = 0
  71     if not args: # No arguments
  72         usage()
  73         sys.exit(2)
  74     for opt, arg in opts:
  75         if opt == '-c':
  76             setdocomments()
  77         if opt == '-r':
  78             setreverse()
  79         if opt == '-s':
  80             addsubst(arg)
  81     for arg in args:
  82         if os.path.isdir(arg):
  83             if recursedown(arg): bad = 1
  84         elif os.path.islink(arg):
  85             err(arg + ': will not process symbolic links\n')
  86             bad = 1
  87         else:
  88             if fix(arg): bad = 1
  89     sys.exit(bad)
  90
  91 # Change this regular expression to select a different set of files
  92 Wanted = '^[a-zA-Z0-9_]+\.[ch]$'
  93 def wanted(name):
  94     return regex.match(Wanted, name) >= 0
  95
  96 def recursedown(dirname):
  97     dbg('recursedown(' + `dirname` + ')\n')
  98     bad = 0
  99     try:
 100         names = os.listdir(dirname)
 101     except os.error, msg:
 102         err(dirname + ': cannot list directory: ' + str(msg) + '\n')
 103         return 1
 104     names.sort()
 105     subdirs = []
 106     for name in names:
 107         if name in (os.curdir, os.pardir): continue
 108         fullname = os.path.join(dirname, name)
 109         if os.path.islink(fullname): pass
 110         elif os.path.isdir(fullname):
 111             subdirs.append(fullname)
 112         elif wanted(name):
 113             if fix(fullname): bad = 1
 114     for fullname in subdirs:
 115         if recursedown(fullname): bad = 1
 116     return bad
 117
 118 def fix(filename):
 119 ##  dbg('fix(' + `filename` + ')\n')
 120     if filename == '-':
 121         # Filter mode
 122         f = sys.stdin
 123         g = sys.stdout
 124     else:
 125         # File replacement mode
 126         try:
 127             f = open(filename, 'r')
 128         except IOError, msg:
 129             err(filename + ': cannot open: ' + str(msg) + '\n')
 130             return 1
 131         head, tail = os.path.split(filename)
 132         tempname = os.path.join(head, '@' + tail)
 133         g = None
 134     # If we find a match, we rewind the file and start over but
 135     # now copy everything to a temp file.
 136     lineno = 0
 137     initfixline()
 138     while 1:
 139         line = f.readline()
 140         if not line: break
 141         lineno = lineno + 1
 142         while line[-2:] == '\\\n':
 143             nextline = f.readline()
 144             if not nextline: break
 145             line = line + nextline
 146             lineno = lineno + 1
 147         newline = fixline(line)
 148         if newline != line:
 149             if g is None:
 150                 try:
 151                     g = open(tempname, 'w')
 152                 except IOError, msg:
 153                     f.close()
 154                     err(tempname+': cannot create: '+
 155                         str(msg)+'\n')
 156                     return 1
 157                 f.seek(0)
 158                 lineno = 0
 159                 initfixline()
 160                 rep(filename + ':\n')
 161                 continue # restart from the beginning
 162             rep(`lineno` + '\n')
 163             rep('< ' + line)
 164             rep('> ' + newline)
 165         if g is not None:
 166             g.write(newline)
 167
 168     # End of file
 169     if filename == '-': return 0 # Done in filter mode
 170     f.close()
 171     if not g: return 0 # No changes
 172
 173     # Finishing touch -- move files
 174
 175     # First copy the file's mode to the temp file
 176     try:
 177         statbuf = os.stat(filename)
 178         os.chmod(tempname, statbuf[ST_MODE] & 07777)
 179     except os.error, msg:
 180         err(tempname + ': warning: chmod failed (' + str(msg) + ')\n')
 181     # Then make a backup of the original file as filename~
 182     try:
 183         os.rename(filename, filename + '~')
 184     except os.error, msg:
 185         err(filename + ': warning: backup failed (' + str(msg) + ')\n')
 186     # Now move the temp file to the original file
 187     try:
 188         os.rename(tempname, filename)
 189     except os.error, msg:
 190         err(filename + ': rename failed (' + str(msg) + ')\n')
 191         return 1
 192     # Return succes
 193     return 0
 194
 195 # Tokenizing ANSI C (partly)
 196
 197 Identifier = '\(struct \)?[a-zA-Z_][a-zA-Z0-9_]+'
 198 String = '"\([^\n\\"]\|\\\\.\)*"'
 199 Char = '\'\([^\n\\\']\|\\\\.\)*\''
 200 CommentStart = '/\*'
 201 CommentEnd = '\*/'
 202
 203 Hexnumber = '0[xX][0-9a-fA-F]*[uUlL]*'
 204 Octnumber = '0[0-7]*[uUlL]*'
 205 Decnumber = '[1-9][0-9]*[uUlL]*'
 206 Intnumber = Hexnumber + '\|' + Octnumber + '\|' + Decnumber
 207 Exponent = '[eE][-+]?[0-9]+'
 208 Pointfloat = '\([0-9]+\.[0-9]*\|\.[0-9]+\)\(' + Exponent + '\)?'
 209 Expfloat = '[0-9]+' + Exponent
 210 Floatnumber = Pointfloat + '\|' + Expfloat
 211 Number = Floatnumber + '\|' + Intnumber
 212
 213 # Anything else is an operator -- don't list this explicitly because of '/*'
 214
 215 OutsideComment = (Identifier, Number, String, Char, CommentStart)
 216 OutsideCommentPattern = '\(' + string.joinfields(OutsideComment, '\|') + '\)'
 217 OutsideCommentProgram = regex.compile(OutsideCommentPattern)
 218
 219 InsideComment = (Identifier, Number, CommentEnd)
 220 InsideCommentPattern = '\(' + string.joinfields(InsideComment, '\|') + '\)'
 221 InsideCommentProgram = regex.compile(InsideCommentPattern)
 222
 223 def initfixline():
 224     global Program
 225     Program = OutsideCommentProgram
 226
 227 def fixline(line):
 228     global Program
 229 ##  print '-->', `line`
 230     i = 0
 231     while i < len(line):
 232         i = Program.search(line, i)
 233         if i < 0: break
 234         found = Program.group(0)
 235 ##      if Program is InsideCommentProgram: print '...',
 236 ##      else: print '   ',
 237 ##      print found
 238         if len(found) == 2:
 239             if found == '/*':
 240                 Program = InsideCommentProgram
 241             elif found == '*/':
 242                 Program = OutsideCommentProgram
 243         n = len(found)
 244         if Dict.has_key(found):
 245             subst = Dict[found]
 246             if Program is InsideCommentProgram:
 247                 if not Docomments:
 248                     print 'Found in comment:', found
 249                     i = i + n
 250                     continue
 251                 if NotInComment.has_key(found):
 252 ##                  print 'Ignored in comment:',
 253 ##                  print found, '-->', subst
 254 ##                  print 'Line:', line,
 255                     subst = found
 256 ##              else:
 257 ##                  print 'Substituting in comment:',
 258 ##                  print found, '-->', subst
 259 ##                  print 'Line:', line,
 260             line = line[:i] + subst + line[i+n:]
 261             n = len(subst)
 262         i = i + n
 263     return line
 264
 265 Docomments = 0
 266 def setdocomments():
 267     global Docomments
 268     Docomments = 1
 269
 270 Reverse = 0
 271 def setreverse():
 272     global Reverse
 273     Reverse = (not Reverse)
 274
 275 Dict = {}
 276 NotInComment = {}
 277 def addsubst(substfile):
 278     try:
 279         fp = open(substfile, 'r')
 280     except IOError, msg:
 281         err(substfile + ': cannot read substfile: ' + str(msg) + '\n')
 282         sys.exit(1)
 283     lineno = 0
 284     while 1:
 285         line = fp.readline()
 286         if not line: break
 287         lineno = lineno + 1
 288         try:
 289             i = string.index(line, '#')
 290         except string.index_error:
 291             i = -1          # Happens to delete trailing \n
 292         words = string.split(line[:i])
 293         if not words: continue
 294         if len(words) == 3 and words[0] == 'struct':
 295             words[:2] = [words[0] + ' ' + words[1]]
 296         elif len(words) <> 2:
 297             err(substfile + ':' + `lineno` +
 298                       ': warning: bad line: ' + line)
 299             continue
 300         if Reverse:
 301             [value, key] = words
 302         else:
 303             [key, value] = words
 304         if value[0] == '*':
 305             value = value[1:]
 306         if key[0] == '*':
 307             key = key[1:]
 308             NotInComment[key] = value
 309         if Dict.has_key(key):
 310             err(substfile + ':' + `lineno` +
 311                       ': warning: overriding: ' +
 312                       key + ' ' + value + '\n')
 313             err(substfile + ':' + `lineno` +
 314                       ': warning: previous: ' + Dict[key] + '\n')
 315         Dict[key] = value
 316     fp.close()
 317
 318 main()