| [2] | 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 re
|
|---|
| 39 | import os
|
|---|
| 40 | from stat import *
|
|---|
| 41 | import getopt
|
|---|
| 42 |
|
|---|
| 43 | err = sys.stderr.write
|
|---|
| 44 | dbg = err
|
|---|
| 45 | rep = sys.stdout.write
|
|---|
| 46 |
|
|---|
| 47 | def usage():
|
|---|
| 48 | progname = sys.argv[0]
|
|---|
| 49 | err('Usage: ' + progname +
|
|---|
| 50 | ' [-c] [-r] [-s file] ... file-or-directory ...\n')
|
|---|
| 51 | err('\n')
|
|---|
| 52 | err('-c : substitute inside comments\n')
|
|---|
| 53 | err('-r : reverse direction for following -s options\n')
|
|---|
| 54 | err('-s substfile : add a file of substitutions\n')
|
|---|
| 55 | err('\n')
|
|---|
| 56 | err('Each non-empty non-comment line in a substitution file must\n')
|
|---|
| 57 | err('contain exactly two words: an identifier and its replacement.\n')
|
|---|
| 58 | err('Comments start with a # character and end at end of line.\n')
|
|---|
| 59 | err('If an identifier is preceded with a *, it is not substituted\n')
|
|---|
| 60 | err('inside a comment even when -c is specified.\n')
|
|---|
| 61 |
|
|---|
| 62 | def main():
|
|---|
| 63 | try:
|
|---|
| 64 | opts, args = getopt.getopt(sys.argv[1:], 'crs:')
|
|---|
| 65 | except getopt.error, msg:
|
|---|
| 66 | err('Options error: ' + str(msg) + '\n')
|
|---|
| 67 | usage()
|
|---|
| 68 | sys.exit(2)
|
|---|
| 69 | bad = 0
|
|---|
| 70 | if not args: # No arguments
|
|---|
| 71 | usage()
|
|---|
| 72 | sys.exit(2)
|
|---|
| 73 | for opt, arg in opts:
|
|---|
| 74 | if opt == '-c':
|
|---|
| 75 | setdocomments()
|
|---|
| 76 | if opt == '-r':
|
|---|
| 77 | setreverse()
|
|---|
| 78 | if opt == '-s':
|
|---|
| 79 | addsubst(arg)
|
|---|
| 80 | for arg in args:
|
|---|
| 81 | if os.path.isdir(arg):
|
|---|
| 82 | if recursedown(arg): bad = 1
|
|---|
| 83 | elif os.path.islink(arg):
|
|---|
| 84 | err(arg + ': will not process symbolic links\n')
|
|---|
| 85 | bad = 1
|
|---|
| 86 | else:
|
|---|
| 87 | if fix(arg): bad = 1
|
|---|
| 88 | sys.exit(bad)
|
|---|
| 89 |
|
|---|
| 90 | # Change this regular expression to select a different set of files
|
|---|
| 91 | Wanted = '^[a-zA-Z0-9_]+\.[ch]$'
|
|---|
| 92 | def wanted(name):
|
|---|
| 93 | return re.match(Wanted, name) >= 0
|
|---|
| 94 |
|
|---|
| 95 | def recursedown(dirname):
|
|---|
| 96 | dbg('recursedown(%r)\n' % (dirname,))
|
|---|
| 97 | bad = 0
|
|---|
| 98 | try:
|
|---|
| 99 | names = os.listdir(dirname)
|
|---|
| 100 | except os.error, msg:
|
|---|
| 101 | err(dirname + ': cannot list directory: ' + str(msg) + '\n')
|
|---|
| 102 | return 1
|
|---|
| 103 | names.sort()
|
|---|
| 104 | subdirs = []
|
|---|
| 105 | for name in names:
|
|---|
| 106 | if name in (os.curdir, os.pardir): continue
|
|---|
| 107 | fullname = os.path.join(dirname, name)
|
|---|
| 108 | if os.path.islink(fullname): pass
|
|---|
| 109 | elif os.path.isdir(fullname):
|
|---|
| 110 | subdirs.append(fullname)
|
|---|
| 111 | elif wanted(name):
|
|---|
| 112 | if fix(fullname): bad = 1
|
|---|
| 113 | for fullname in subdirs:
|
|---|
| 114 | if recursedown(fullname): bad = 1
|
|---|
| 115 | return bad
|
|---|
| 116 |
|
|---|
| 117 | def fix(filename):
|
|---|
| 118 | ## dbg('fix(%r)\n' % (filename,))
|
|---|
| 119 | if filename == '-':
|
|---|
| 120 | # Filter mode
|
|---|
| 121 | f = sys.stdin
|
|---|
| 122 | g = sys.stdout
|
|---|
| 123 | else:
|
|---|
| 124 | # File replacement mode
|
|---|
| 125 | try:
|
|---|
| 126 | f = open(filename, 'r')
|
|---|
| 127 | except IOError, msg:
|
|---|
| 128 | err(filename + ': cannot open: ' + str(msg) + '\n')
|
|---|
| 129 | return 1
|
|---|
| 130 | head, tail = os.path.split(filename)
|
|---|
| 131 | tempname = os.path.join(head, '@' + tail)
|
|---|
| 132 | g = None
|
|---|
| 133 | # If we find a match, we rewind the file and start over but
|
|---|
| 134 | # now copy everything to a temp file.
|
|---|
| 135 | lineno = 0
|
|---|
| 136 | initfixline()
|
|---|
| 137 | while 1:
|
|---|
| 138 | line = f.readline()
|
|---|
| 139 | if not line: break
|
|---|
| 140 | lineno = lineno + 1
|
|---|
| 141 | while line[-2:] == '\\\n':
|
|---|
| 142 | nextline = f.readline()
|
|---|
| 143 | if not nextline: break
|
|---|
| 144 | line = line + nextline
|
|---|
| 145 | lineno = lineno + 1
|
|---|
| 146 | newline = fixline(line)
|
|---|
| 147 | if newline != line:
|
|---|
| 148 | if g is None:
|
|---|
| 149 | try:
|
|---|
| 150 | g = open(tempname, 'w')
|
|---|
| 151 | except IOError, msg:
|
|---|
| 152 | f.close()
|
|---|
| 153 | err(tempname+': cannot create: '+
|
|---|
| 154 | str(msg)+'\n')
|
|---|
| 155 | return 1
|
|---|
| 156 | f.seek(0)
|
|---|
| 157 | lineno = 0
|
|---|
| 158 | initfixline()
|
|---|
| 159 | rep(filename + ':\n')
|
|---|
| 160 | continue # restart from the beginning
|
|---|
| 161 | rep(repr(lineno) + '\n')
|
|---|
| 162 | rep('< ' + line)
|
|---|
| 163 | rep('> ' + newline)
|
|---|
| 164 | if g is not None:
|
|---|
| 165 | g.write(newline)
|
|---|
| 166 |
|
|---|
| 167 | # End of file
|
|---|
| 168 | if filename == '-': return 0 # Done in filter mode
|
|---|
| 169 | f.close()
|
|---|
| 170 | if not g: return 0 # No changes
|
|---|
| 171 |
|
|---|
| 172 | # Finishing touch -- move files
|
|---|
| 173 |
|
|---|
| 174 | # First copy the file's mode to the temp file
|
|---|
| 175 | try:
|
|---|
| 176 | statbuf = os.stat(filename)
|
|---|
| 177 | os.chmod(tempname, statbuf[ST_MODE] & 07777)
|
|---|
| 178 | except os.error, msg:
|
|---|
| 179 | err(tempname + ': warning: chmod failed (' + str(msg) + ')\n')
|
|---|
| 180 | # Then make a backup of the original file as filename~
|
|---|
| 181 | try:
|
|---|
| 182 | os.rename(filename, filename + '~')
|
|---|
| 183 | except os.error, msg:
|
|---|
| 184 | err(filename + ': warning: backup failed (' + str(msg) + ')\n')
|
|---|
| 185 | # Now move the temp file to the original file
|
|---|
| 186 | try:
|
|---|
| 187 | os.rename(tempname, filename)
|
|---|
| 188 | except os.error, msg:
|
|---|
| 189 | err(filename + ': rename failed (' + str(msg) + ')\n')
|
|---|
| 190 | return 1
|
|---|
| [391] | 191 | # Return success
|
|---|
| [2] | 192 | return 0
|
|---|
| 193 |
|
|---|
| 194 | # Tokenizing ANSI C (partly)
|
|---|
| 195 |
|
|---|
| 196 | Identifier = '\(struct \)?[a-zA-Z_][a-zA-Z0-9_]+'
|
|---|
| 197 | String = '"\([^\n\\"]\|\\\\.\)*"'
|
|---|
| 198 | Char = '\'\([^\n\\\']\|\\\\.\)*\''
|
|---|
| 199 | CommentStart = '/\*'
|
|---|
| 200 | CommentEnd = '\*/'
|
|---|
| 201 |
|
|---|
| 202 | Hexnumber = '0[xX][0-9a-fA-F]*[uUlL]*'
|
|---|
| 203 | Octnumber = '0[0-7]*[uUlL]*'
|
|---|
| 204 | Decnumber = '[1-9][0-9]*[uUlL]*'
|
|---|
| 205 | Intnumber = Hexnumber + '\|' + Octnumber + '\|' + Decnumber
|
|---|
| 206 | Exponent = '[eE][-+]?[0-9]+'
|
|---|
| 207 | Pointfloat = '\([0-9]+\.[0-9]*\|\.[0-9]+\)\(' + Exponent + '\)?'
|
|---|
| 208 | Expfloat = '[0-9]+' + Exponent
|
|---|
| 209 | Floatnumber = Pointfloat + '\|' + Expfloat
|
|---|
| 210 | Number = Floatnumber + '\|' + Intnumber
|
|---|
| 211 |
|
|---|
| 212 | # Anything else is an operator -- don't list this explicitly because of '/*'
|
|---|
| 213 |
|
|---|
| 214 | OutsideComment = (Identifier, Number, String, Char, CommentStart)
|
|---|
| 215 | OutsideCommentPattern = '(' + '|'.join(OutsideComment) + ')'
|
|---|
| 216 | OutsideCommentProgram = re.compile(OutsideCommentPattern)
|
|---|
| 217 |
|
|---|
| 218 | InsideComment = (Identifier, Number, CommentEnd)
|
|---|
| 219 | InsideCommentPattern = '(' + '|'.join(InsideComment) + ')'
|
|---|
| 220 | InsideCommentProgram = re.compile(InsideCommentPattern)
|
|---|
| 221 |
|
|---|
| 222 | def initfixline():
|
|---|
| 223 | global Program
|
|---|
| 224 | Program = OutsideCommentProgram
|
|---|
| 225 |
|
|---|
| 226 | def fixline(line):
|
|---|
| 227 | global Program
|
|---|
| 228 | ## print '-->', repr(line)
|
|---|
| 229 | i = 0
|
|---|
| 230 | while i < len(line):
|
|---|
| 231 | i = Program.search(line, i)
|
|---|
| 232 | if i < 0: break
|
|---|
| 233 | found = Program.group(0)
|
|---|
| 234 | ## if Program is InsideCommentProgram: print '...',
|
|---|
| 235 | ## else: print ' ',
|
|---|
| 236 | ## print found
|
|---|
| 237 | if len(found) == 2:
|
|---|
| 238 | if found == '/*':
|
|---|
| 239 | Program = InsideCommentProgram
|
|---|
| 240 | elif found == '*/':
|
|---|
| 241 | Program = OutsideCommentProgram
|
|---|
| 242 | n = len(found)
|
|---|
| 243 | if Dict.has_key(found):
|
|---|
| 244 | subst = Dict[found]
|
|---|
| 245 | if Program is InsideCommentProgram:
|
|---|
| 246 | if not Docomments:
|
|---|
| 247 | print 'Found in comment:', found
|
|---|
| 248 | i = i + n
|
|---|
| 249 | continue
|
|---|
| 250 | if NotInComment.has_key(found):
|
|---|
| 251 | ## print 'Ignored in comment:',
|
|---|
| 252 | ## print found, '-->', subst
|
|---|
| 253 | ## print 'Line:', line,
|
|---|
| 254 | subst = found
|
|---|
| 255 | ## else:
|
|---|
| 256 | ## print 'Substituting in comment:',
|
|---|
| 257 | ## print found, '-->', subst
|
|---|
| 258 | ## print 'Line:', line,
|
|---|
| 259 | line = line[:i] + subst + line[i+n:]
|
|---|
| 260 | n = len(subst)
|
|---|
| 261 | i = i + n
|
|---|
| 262 | return line
|
|---|
| 263 |
|
|---|
| 264 | Docomments = 0
|
|---|
| 265 | def setdocomments():
|
|---|
| 266 | global Docomments
|
|---|
| 267 | Docomments = 1
|
|---|
| 268 |
|
|---|
| 269 | Reverse = 0
|
|---|
| 270 | def setreverse():
|
|---|
| 271 | global Reverse
|
|---|
| 272 | Reverse = (not Reverse)
|
|---|
| 273 |
|
|---|
| 274 | Dict = {}
|
|---|
| 275 | NotInComment = {}
|
|---|
| 276 | def addsubst(substfile):
|
|---|
| 277 | try:
|
|---|
| 278 | fp = open(substfile, 'r')
|
|---|
| 279 | except IOError, msg:
|
|---|
| 280 | err(substfile + ': cannot read substfile: ' + str(msg) + '\n')
|
|---|
| 281 | sys.exit(1)
|
|---|
| 282 | lineno = 0
|
|---|
| 283 | while 1:
|
|---|
| 284 | line = fp.readline()
|
|---|
| 285 | if not line: break
|
|---|
| 286 | lineno = lineno + 1
|
|---|
| 287 | try:
|
|---|
| 288 | i = line.index('#')
|
|---|
| 289 | except ValueError:
|
|---|
| 290 | i = -1 # Happens to delete trailing \n
|
|---|
| 291 | words = line[:i].split()
|
|---|
| 292 | if not words: continue
|
|---|
| 293 | if len(words) == 3 and words[0] == 'struct':
|
|---|
| 294 | words[:2] = [words[0] + ' ' + words[1]]
|
|---|
| 295 | elif len(words) <> 2:
|
|---|
| 296 | err(substfile + '%s:%r: warning: bad line: %r' % (substfile, lineno, line))
|
|---|
| 297 | continue
|
|---|
| 298 | if Reverse:
|
|---|
| 299 | [value, key] = words
|
|---|
| 300 | else:
|
|---|
| 301 | [key, value] = words
|
|---|
| 302 | if value[0] == '*':
|
|---|
| 303 | value = value[1:]
|
|---|
| 304 | if key[0] == '*':
|
|---|
| 305 | key = key[1:]
|
|---|
| 306 | NotInComment[key] = value
|
|---|
| 307 | if Dict.has_key(key):
|
|---|
| 308 | err('%s:%r: warning: overriding: %r %r\n' % (substfile, lineno, key, value))
|
|---|
| 309 | err('%s:%r: warning: previous: %r\n' % (substfile, lineno, Dict[key]))
|
|---|
| 310 | Dict[key] = value
|
|---|
| 311 | fp.close()
|
|---|
| 312 |
|
|---|
| 313 | if __name__ == '__main__':
|
|---|
| 314 | main()
|
|---|
