]>
git.saurik.com Git - bison.git/blob - lib/ansi2knr.c
1 /* Copyright (C) 1989, 1997, 1998 Aladdin Enterprises. All rights reserved. */
4 /* Convert ANSI C function definitions to K&R ("traditional C") syntax */
7 ansi2knr is distributed in the hope that it will be useful, but WITHOUT ANY
8 WARRANTY. No author or distributor accepts responsibility to anyone for the
9 consequences of using it or for whether it serves any particular purpose or
10 works at all, unless he says so in writing. Refer to the GNU General Public
11 License (the "GPL") for full details.
13 Everyone is granted permission to copy, modify and redistribute ansi2knr,
14 but only under the conditions described in the GPL. A copy of this license
15 is supposed to have been given to you along with ansi2knr so you can know
16 your rights and responsibilities. It should be in a file named COPYLEFT,
17 or, if there is no file named COPYLEFT, a file named COPYING. Among other
18 things, the copyright notice and this notice must be preserved on all
21 We explicitly state here what we believe is already implied by the GPL: if
22 the ansi2knr program is distributed as a separate set of sources and a
23 separate executable file which are aggregated on a storage medium together
24 with another program, this in itself does not bring the other program under
25 the GPL, nor does the mere fact that such a program or the procedures for
26 constructing it invoke the ansi2knr executable bring any other part of the
27 program under the GPL.
32 ansi2knr [--filename FILENAME] [INPUT_FILE [OUTPUT_FILE]]
33 * --filename provides the file name for the #line directive in the output,
34 * overriding input_file (if present).
35 * If no input_file is supplied, input is read from stdin.
36 * If no output_file is supplied, output goes to stdout.
37 * There are no error messages.
39 * ansi2knr recognizes function definitions by seeing a non-keyword
40 * identifier at the left margin, followed by a left parenthesis,
41 * with a right parenthesis as the last character on the line,
42 * and with a left brace as the first token on the following line
43 * (ignoring possible intervening comments), except that a line
45 * identifier1(identifier2)
46 * will not be considered a function definition unless identifier2 is
47 * the word "void". ansi2knr will recognize a multi-line header provided
48 * that no intervening line ends with a left or right brace or a semicolon.
49 * These algorithms ignore whitespace and comments, except that
50 * the function name must be the first thing on the line.
51 * The following constructs will confuse it:
52 * - Any other construct that starts at the left margin and
53 * follows the above syntax (such as a macro or function call).
54 * - Some macros that tinker with the syntax of the function header.
58 * The original and principal author of ansi2knr is L. Peter Deutsch
59 * <ghost@aladdin.com>. Other authors are noted in the change history
60 * that follows (in reverse chronological order):
61 lpd 1998-11-09 added further hack to recognize identifier(void)
63 lpd 1998-10-23 added hack to recognize lines consisting of
64 identifier1(identifier2) as *not* being procedures
65 lpd 1997-12-08 made input_file optional; only closes input and/or
66 output file if not stdin or stdout respectively; prints
67 usage message on stderr rather than stdout; adds
68 --filename switch (changes suggested by
69 <ceder@lysator.liu.se>)
70 lpd 1996-01-21 added code to cope with not HAVE_CONFIG_H and with
71 compilers that don't understand void, as suggested by
73 lpd 1996-01-15 changed to require that the first non-comment token
74 on the line following a function header be a left brace,
75 to reduce sensitivity to macros, as suggested by Tom Lane
77 lpd 1995-06-22 removed #ifndefs whose sole purpose was to define
78 undefined preprocessor symbols as 0; changed all #ifdefs
79 for configuration symbols to #ifs
80 lpd 1995-04-05 changed copyright notice to make it clear that
81 including ansi2knr in a program does not bring the entire
83 lpd 1994-12-18 added conditionals for systems where ctype macros
84 don't handle 8-bit characters properly, suggested by
85 Francois Pinard <pinard@iro.umontreal.ca>;
86 removed --varargs switch (this is now the default)
87 lpd 1994-10-10 removed CONFIG_BROKETS conditional
88 lpd 1994-07-16 added some conditionals to help GNU `configure',
89 suggested by Francois Pinard <pinard@iro.umontreal.ca>;
90 properly erase prototype args in function parameters,
91 contributed by Jim Avera <jima@netcom.com>;
92 correct error in writeblanks (it shouldn't erase EOLs)
93 lpd 1989-xx-xx original version
96 /* Most of the conditionals here are to make ansi2knr work with */
97 /* or without the GNU configure machinery. */
109 For properly autoconfiguring ansi2knr, use AC_CONFIG_HEADER(config.h).
110 This will define HAVE_CONFIG_H and so, activate the following lines.
113 # if STDC_HEADERS || HAVE_STRING_H
116 # include <strings.h>
119 #else /* not HAVE_CONFIG_H */
121 /* Otherwise do it the hard way */
124 # include <strings.h>
127 extern int strlen(), strncmp();
133 #endif /* not HAVE_CONFIG_H */
139 malloc and free should be declared in stdlib.h,
140 but if you've got a K&R compiler, they probably aren't.
146 extern char *malloc();
149 extern char *malloc();
157 * The ctype macros don't always handle 8-bit characters correctly.
158 * Compensate for this here.
161 # undef HAVE_ISASCII /* just in case */
162 # define HAVE_ISASCII 1
165 #if STDC_HEADERS || !HAVE_ISASCII
166 # define is_ascii(c) 1
168 # define is_ascii(c) isascii(c)
171 #define is_space(c) (is_ascii(c) && isspace(c))
172 #define is_alpha(c) (is_ascii(c) && isalpha(c))
173 #define is_alnum(c) (is_ascii(c) && isalnum(c))
175 /* Scanning macros */
176 #define isidchar(ch) (is_alnum(ch) || (ch) == '_')
177 #define isidfirstchar(ch) (is_alpha(ch) || (ch) == '_')
179 /* Forward references */
185 /* The main program */
193 #define bufsize 5000 /* arbitrary size */
198 "Usage: ansi2knr [--filename FILENAME] [INPUT_FILE [OUTPUT_FILE]]\n";
200 * In previous versions, ansi2knr recognized a --varargs switch.
201 * If this switch was supplied, ansi2knr would attempt to convert
202 * a ... argument to va_alist and va_dcl; if this switch was not
203 * supplied, ansi2knr would simply drop any such arguments.
204 * Now, ansi2knr always does this conversion, and we only
205 * check for this switch for backward compatibility.
207 int convert_varargs
= 1;
209 while ( argc
> 1 && argv
[1][0] == '-' ) {
210 if ( !strcmp(argv
[1], "--varargs") ) {
216 if ( !strcmp(argv
[1], "--filename") && argc
> 2 ) {
222 fprintf(stderr
, "Unrecognized switch: %s\n", argv
[1]);
223 fprintf(stderr
, usage
);
229 fprintf(stderr
, usage
);
232 out
= fopen(argv
[2], "w");
234 fprintf(stderr
, "Cannot open output file %s\n", argv
[2]);
239 in
= fopen(argv
[1], "r");
241 fprintf(stderr
, "Cannot open input file %s\n", argv
[1]);
251 fprintf(out
, "#line 1 \"%s\"\n", filename
);
252 buf
= malloc(bufsize
);
254 while ( fgets(line
, (unsigned)(buf
+ bufsize
- line
), in
) != NULL
)
256 test
: line
+= strlen(line
);
257 switch ( test1(buf
) )
259 case 2: /* a function header */
260 convert1(buf
, out
, 1, convert_varargs
);
262 case 1: /* a function */
263 /* Check for a { at the start of the next line. */
265 f
: if ( line
>= buf
+ (bufsize
- 1) ) /* overflow check */
267 if ( fgets(line
, (unsigned)(buf
+ bufsize
- line
), in
) == NULL
)
269 switch ( *skipspace(more
, 1) )
272 /* Definitely a function header. */
273 convert1(buf
, out
, 0, convert_varargs
);
277 /* The next line was blank or a comment: */
278 /* keep scanning for a non-comment. */
279 line
+= strlen(line
);
282 /* buf isn't a function header, but */
290 case -1: /* maybe the start of a function */
291 if ( line
!= buf
+ (bufsize
- 1) ) /* overflow check */
294 default: /* not a function */
310 /* Skip over space and comments, in either direction. */
314 register int dir
; /* 1 for forward, -1 for backward */
316 { while ( is_space(*p
) )
318 if ( !(*p
== '/' && p
[dir
] == '*') )
321 while ( !(*p
== '*' && p
[dir
] == '/') )
323 return p
; /* multi-line comment?? */
332 * Write blanks over part of a string.
333 * Don't overwrite end-of-line characters.
336 writeblanks(start
, end
)
340 for ( p
= start
; p
< end
; p
++ )
341 if ( *p
!= '\r' && *p
!= '\n' )
347 * Test whether the string in buf is a function definition.
348 * The string may contain and/or end with a newline.
350 * 0 - definitely not a function definition;
351 * 1 - definitely a function definition;
352 * 2 - definitely a function prototype (NOT USED);
353 * -1 - may be the beginning of a function definition,
354 * append another line and look again.
355 * The reason we don't attempt to convert function prototypes is that
356 * Ghostscript's declaration-generating macros look too much like
357 * prototypes, and confuse the algorithms.
362 { register char *p
= buf
;
367 if ( !isidfirstchar(*p
) )
368 return 0; /* no name at left margin */
369 bend
= skipspace(buf
+ strlen(buf
) - 1, -1);
372 case ';': contin
= 0 /*2*/; break;
373 case ')': contin
= 1; break;
374 case '{': return 0; /* not a function */
375 case '}': return 0; /* not a function */
376 default: contin
= -1;
378 while ( isidchar(*p
) )
383 return 0; /* not a function */
386 return 0; /* no parameters */
387 /* Check that the apparent function name isn't a keyword. */
388 /* We only need to check for keywords that could be followed */
389 /* by a left parenthesis (which, unfortunately, is most of them). */
390 { static char *words
[] =
391 { "asm", "auto", "case", "char", "const", "double",
392 "extern", "float", "for", "if", "int", "long",
393 "register", "return", "short", "signed", "sizeof",
394 "static", "switch", "typedef", "unsigned",
395 "void", "volatile", "while", 0
399 int len
= endfn
- buf
;
401 while ( (kp
= *key
) != 0 )
402 { if ( strlen(kp
) == len
&& !strncmp(kp
, buf
, len
) )
403 return 0; /* name is a keyword */
411 * Check for identifier1(identifier2) and not
415 while ( isidchar(*p
) )
419 if ( *p
== ')' && (len
!= 4 || strncmp(id
, "void", 4)) )
420 return 0; /* not a function */
423 * If the last significant character was a ), we need to count
424 * parentheses, because it might be part of a formal parameter
425 * that is a procedure.
430 for (p
= skipspace(buf
, 1); *p
; p
= skipspace(p
+ 1, 1))
431 level
+= (*p
== '(' ? 1 : *p
== ')' ? -1 : 0);
438 /* Convert a recognized function definition or header to K&R syntax. */
440 convert1(buf
, out
, header
, convert_varargs
)
443 int header
; /* Boolean */
444 int convert_varargs
; /* Boolean */
448 * The breaks table contains pointers to the beginning and end
452 unsigned num_breaks
= 2; /* for testing */
458 /* Pre-ANSI implementations don't agree on whether strchr */
459 /* is called strchr or index, so we open-code it here. */
460 for ( endfn
= buf
; *(endfn
++) != '('; )
463 breaks
= (char **)malloc(sizeof(char *) * num_breaks
* 2);
465 { /* Couldn't allocate break table, give up */
466 fprintf(stderr
, "Unable to allocate break table!\n");
470 btop
= breaks
+ num_breaks
* 2 - 2;
472 /* Parse the argument list */
480 { /* Filled up break table. */
481 /* Allocate a bigger one and start over. */
482 free((char *)breaks
);
487 /* Find the end of the argument */
488 for ( ; end
== NULL
; p
++ )
492 if ( !level
) end
= p
;
495 if ( !level
) lp
= p
;
499 if ( --level
< 0 ) end
= p
;
503 p
= skipspace(p
, 1) - 1;
509 /* Erase any embedded prototype parameters. */
511 writeblanks(lp
+ 1, rp
);
512 p
--; /* back up over terminator */
513 /* Find the name being declared. */
514 /* This is complicated because of procedure and */
515 /* array modifiers. */
517 { p
= skipspace(p
- 1, -1);
520 case ']': /* skip array dimension(s) */
521 case ')': /* skip procedure args OR name */
526 case ']': case ')': level
++; break;
527 case '[': case '(': level
--; break;
528 case '/': p
= skipspace(p
, -1) + 1; break;
532 if ( *p
== '(' && *skipspace(p
+ 1, 1) == '*' )
533 { /* We found the name being declared */
534 while ( !isidfirstchar(*p
) )
535 p
= skipspace(p
, 1) + 1;
543 found
: if ( *p
== '.' && p
[-1] == '.' && p
[-2] == '.' )
544 { if ( convert_varargs
)
545 { *bp
++ = "va_alist";
550 if ( bp
== breaks
+ 1 ) /* sole argument */
551 writeblanks(breaks
[0], p
);
553 writeblanks(bp
[-1] - 1, p
);
558 { while ( isidchar(*p
) ) p
--;
563 while ( *p
++ == ',' );
565 /* Make a special check for 'void' arglist */
566 if ( bp
== breaks
+2 )
567 { p
= skipspace(breaks
[0], 1);
568 if ( !strncmp(p
, "void", 4) )
569 { p
= skipspace(p
+4, 1);
570 if ( p
== breaks
[2] - 1 )
571 { bp
= breaks
; /* yup, pretend arglist is empty */
572 writeblanks(breaks
[0], p
+ 1);
576 /* Put out the function name and left parenthesis. */
578 while ( p
!= endfn
) putc(*p
, out
), p
++;
579 /* Put out the declaration. */
582 for ( p
= breaks
[0]; *p
; p
++ )
583 if ( *p
== '\r' || *p
== '\n' )
587 { for ( ap
= breaks
+1; ap
< bp
; ap
+= 2 )
589 while ( isidchar(*p
) )
595 /* Put out the argument declarations */
596 for ( ap
= breaks
+2; ap
<= bp
; ap
+= 2 )
600 fputs(breaks
[0], out
); /* any prior args */
601 fputs("va_dcl", out
); /* the final arg */
605 fputs(breaks
[0], out
);
607 free((char *)breaks
);