Home | History | Annotate | Line # | Download | only in hexdump
parse.c revision 1.6 
      1 /*	$NetBSD: parse.c,v 1.6 1997/10/18 13:54:39 mrg Exp $	*/
      2 
      3 /*
      4  * Copyright (c) 1989, 1993
      5  *	The Regents of the University of California.  All rights reserved.
      6  *
      7  * Redistribution and use in source and binary forms, with or without
      8  * modification, are permitted provided that the following conditions
      9  * are met:
     10  * 1. Redistributions of source code must retain the above copyright
     11  *    notice, this list of conditions and the following disclaimer.
     12  * 2. Redistributions in binary form must reproduce the above copyright
     13  *    notice, this list of conditions and the following disclaimer in the
     14  *    documentation and/or other materials provided with the distribution.
     15  * 3. All advertising materials mentioning features or use of this software
     16  *    must display the following acknowledgement:
     17  *	This product includes software developed by the University of
     18  *	California, Berkeley and its contributors.
     19  * 4. Neither the name of the University nor the names of its contributors
     20  *    may be used to endorse or promote products derived from this software
     21  *    without specific prior written permission.
     22  *
     23  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
     24  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     25  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     26  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
     27  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     28  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     29  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     30  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     31  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     32  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     33  * SUCH DAMAGE.
     34  */
     35 
     36 #ifndef lint
     37 #if 0
     38 static char sccsid[] = "@(#)parse.c	8.1 (Berkeley) 6/6/93";
     39 #else
     40 static char rcsid[] = "$NetBSD: parse.c,v 1.6 1997/10/18 13:54:39 mrg Exp $";
     41 #endif
     42 #endif /* not lint */
     43 
     44 #include <sys/types.h>
     45 #include <sys/file.h>
     46 
     47 #include <ctype.h>
     48 #include <errno.h>
     49 #include <fcntl.h>
     50 #include <stdio.h>
     51 #include <stdlib.h>
     52 #include <string.h>
     53 
     54 #include "hexdump.h"
     55 
     56 FU *endfu;					/* format at end-of-data */
     57 
     58 void
     59 addfile(name)
     60 	char *name;
     61 {
     62 	register char *p;
     63 	FILE *fp;
     64 	int ch;
     65 	char buf[2048 + 1];
     66 
     67 	if ((fp = fopen(name, "r")) == NULL)
     68 		err("%s: %s\n", name, strerror(errno));
     69 	while (fgets(buf, sizeof(buf), fp)) {
     70 		if (!(p = strchr(buf, '\n'))) {
     71 			(void)fprintf(stderr, "hexdump: line too long.\n");
     72 			while ((ch = getchar()) != '\n' && ch != EOF);
     73 			continue;
     74 		}
     75 		*p = '\0';
     76 		for (p = buf; *p && isspace(*p); ++p);
     77 		if (!*p || *p == '#')
     78 			continue;
     79 		add(p);
     80 	}
     81 	(void)fclose(fp);
     82 }
     83 
     84 void
     85 add(fmt)
     86 	char *fmt;
     87 {
     88 	register char *p;
     89 	static FS **nextfs;
     90 	FS *tfs;
     91 	FU *tfu, **nextfu;
     92 	char *savep;
     93 
     94 	/* start new linked list of format units */
     95 	tfs = emalloc(sizeof(FS));
     96 	if (!fshead)
     97 		fshead = tfs;
     98 	else
     99 		*nextfs = tfs;
    100 	nextfs = &tfs->nextfs;
    101 	nextfu = &tfs->nextfu;
    102 
    103 	/* take the format string and break it up into format units */
    104 	for (p = fmt;;) {
    105 		/* skip leading white space */
    106 		for (; isspace(*p); ++p);
    107 		if (!*p)
    108 			break;
    109 
    110 		/* allocate a new format unit and link it in */
    111 		tfu = emalloc(sizeof(FU));
    112 		*nextfu = tfu;
    113 		nextfu = &tfu->nextfu;
    114 		tfu->reps = 1;
    115 
    116 		/* if leading digit, repetition count */
    117 		if (isdigit(*p)) {
    118 			for (savep = p; isdigit(*p); ++p);
    119 			if (!isspace(*p) && *p != '/')
    120 				badfmt(fmt);
    121 			/* may overwrite either white space or slash */
    122 			tfu->reps = atoi(savep);
    123 			tfu->flags = F_SETREP;
    124 			/* skip trailing white space */
    125 			for (++p; isspace(*p); ++p);
    126 		}
    127 
    128 		/* skip slash and trailing white space */
    129 		if (*p == '/')
    130 			while (isspace(*++p));
    131 
    132 		/* byte count */
    133 		if (isdigit(*p)) {
    134 			for (savep = p; isdigit(*p); ++p);
    135 			if (!isspace(*p))
    136 				badfmt(fmt);
    137 			tfu->bcnt = atoi(savep);
    138 			/* skip trailing white space */
    139 			for (++p; isspace(*p); ++p);
    140 		}
    141 
    142 		/* format */
    143 		if (*p != '"')
    144 			badfmt(fmt);
    145 		for (savep = ++p; *p != '"';)
    146 			if (*p++ == 0)
    147 				badfmt(fmt);
    148 		if (!(tfu->fmt = malloc(p - savep + 1)))
    149 			nomem();
    150 		(void) strncpy(tfu->fmt, savep, p - savep);
    151 		tfu->fmt[p - savep] = '\0';
    152 		escape(tfu->fmt);
    153 		p++;
    154 	}
    155 }
    156 
    157 static const char *spec = ".#-+ 0123456789";
    158 
    159 int
    160 size(fs)
    161 	FS *fs;
    162 {
    163 	register FU *fu;
    164 	register int bcnt, cursize;
    165 	register char *fmt;
    166 	int prec;
    167 
    168 	/* figure out the data block size needed for each format unit */
    169 	for (cursize = 0, fu = fs->nextfu; fu; fu = fu->nextfu) {
    170 		if (fu->bcnt) {
    171 			cursize += fu->bcnt * fu->reps;
    172 			continue;
    173 		}
    174 		for (bcnt = prec = 0, fmt = fu->fmt; *fmt; ++fmt) {
    175 			if (*fmt != '%')
    176 				continue;
    177 			/*
    178 			 * skip any special chars -- save precision in
    179 			 * case it's a %s format.
    180 			 */
    181 			while (strchr(spec + 1, *++fmt));
    182 			if (*fmt == '.' && isdigit(*++fmt)) {
    183 				prec = atoi(fmt);
    184 				while (isdigit(*++fmt));
    185 			}
    186 			switch(*fmt) {
    187 			case 'c':
    188 				bcnt += 1;
    189 				break;
    190 			case 'd': case 'i': case 'o': case 'u':
    191 			case 'x': case 'X':
    192 				bcnt += 4;
    193 				break;
    194 			case 'e': case 'E': case 'f': case 'g': case 'G':
    195 				bcnt += 8;
    196 				break;
    197 			case 's':
    198 				bcnt += prec;
    199 				break;
    200 			case '_':
    201 				switch(*++fmt) {
    202 				case 'c': case 'p': case 'u':
    203 					bcnt += 1;
    204 					break;
    205 				}
    206 			}
    207 		}
    208 		cursize += bcnt * fu->reps;
    209 	}
    210 	return (cursize);
    211 }
    212 
    213 void
    214 rewrite(fs)
    215 	FS *fs;
    216 {
    217 	enum { NOTOKAY, USEBCNT, USEPREC } sokay;
    218 	register PR *pr, **nextpr;
    219 	register FU *fu;
    220 	register char *p1, *p2;
    221 	char savech, *fmtp, cs[3];
    222 	int nconv, prec;
    223 
    224 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
    225 		/*
    226 		 * Break each format unit into print units; each conversion
    227 		 * character gets its own.
    228 		 */
    229 		for (nconv = 0, fmtp = fu->fmt; *fmtp; nextpr = &pr->nextpr) {
    230 			pr = emalloc(sizeof(PR));
    231 			if (!fu->nextpr)
    232 				fu->nextpr = pr;
    233 			else
    234 				*nextpr = pr;
    235 
    236 			/* Skip preceding text and up to the next % sign. */
    237 			for (p1 = fmtp; *p1 && *p1 != '%'; ++p1);
    238 
    239 			/* Only text in the string. */
    240 			if (!*p1) {
    241 				pr->fmt = fmtp;
    242 				pr->flags = F_TEXT;
    243 				break;
    244 			}
    245 
    246 			/*
    247 			 * Get precision for %s -- if have a byte count, don't
    248 			 * need it.
    249 			 */
    250 			if (fu->bcnt) {
    251 				sokay = USEBCNT;
    252 				/* Skip to conversion character. */
    253 				for (++p1; strchr(spec, *p1); ++p1);
    254 			} else {
    255 				/* Skip any special chars, field width. */
    256 				while (strchr(spec + 1, *++p1));
    257 				if (*p1 == '.' && isdigit(*++p1)) {
    258 					sokay = USEPREC;
    259 					prec = atoi(p1);
    260 					while (isdigit(*++p1));
    261 				} else
    262 					sokay = NOTOKAY;
    263 			}
    264 
    265 			p2 = p1 + 1;		/* Set end pointer. */
    266 			cs[0] = *p1;		/* Set conversion string. */
    267 			cs[1] = '\0';
    268 
    269 			/*
    270 			 * Figure out the byte count for each conversion;
    271 			 * rewrite the format as necessary, set up blank-
    272 			 * padding for end of data.
    273 			 */
    274 			switch(cs[0]) {
    275 			case 'c':
    276 				pr->flags = F_CHAR;
    277 				switch(fu->bcnt) {
    278 				case 0: case 1:
    279 					pr->bcnt = 1;
    280 					break;
    281 				default:
    282 					p1[1] = '\0';
    283 					badcnt(p1);
    284 				}
    285 				break;
    286 			case 'd': case 'i':
    287 				pr->flags = F_INT;
    288 				goto isint;
    289 			case 'o': case 'u': case 'x': case 'X':
    290 				pr->flags = F_UINT;
    291 isint:				cs[2] = '\0';
    292 				cs[1] = cs[0];
    293 				cs[0] = 'q';
    294 				switch(fu->bcnt) {
    295 				case 0: case 4:
    296 					pr->bcnt = 4;
    297 					break;
    298 				case 1:
    299 					pr->bcnt = 1;
    300 					break;
    301 				case 2:
    302 					pr->bcnt = 2;
    303 					break;
    304 				default:
    305 					p1[1] = '\0';
    306 					badcnt(p1);
    307 				}
    308 				break;
    309 			case 'e': case 'E': case 'f': case 'g': case 'G':
    310 				pr->flags = F_DBL;
    311 				switch(fu->bcnt) {
    312 				case 0: case 8:
    313 					pr->bcnt = 8;
    314 					break;
    315 				case 4:
    316 					pr->bcnt = 4;
    317 					break;
    318 				default:
    319 					p1[1] = '\0';
    320 					badcnt(p1);
    321 				}
    322 				break;
    323 			case 's':
    324 				pr->flags = F_STR;
    325 				switch(sokay) {
    326 				case NOTOKAY:
    327 					badsfmt();
    328 				case USEBCNT:
    329 					pr->bcnt = fu->bcnt;
    330 					break;
    331 				case USEPREC:
    332 					pr->bcnt = prec;
    333 					break;
    334 				}
    335 				break;
    336 			case '_':
    337 				++p2;
    338 				switch(p1[1]) {
    339 				case 'A':
    340 					endfu = fu;
    341 					fu->flags |= F_IGNORE;
    342 					/* FALLTHROUGH */
    343 				case 'a':
    344 					pr->flags = F_ADDRESS;
    345 					++p2;
    346 					switch(p1[2]) {
    347 					case 'd': case 'o': case'x':
    348 						cs[0] = 'q';
    349 						cs[1] = p1[2];
    350 						cs[2] = '\0';
    351 						break;
    352 					default:
    353 						p1[3] = '\0';
    354 						badconv(p1);
    355 					}
    356 					break;
    357 				case 'c':
    358 					pr->flags = F_C;
    359 					/* cs[0] = 'c';	set in conv_c */
    360 					goto isint2;
    361 				case 'p':
    362 					pr->flags = F_P;
    363 					cs[0] = 'c';
    364 					goto isint2;
    365 				case 'u':
    366 					pr->flags = F_U;
    367 					/* cs[0] = 'c';	set in conv_u */
    368 isint2:					switch(fu->bcnt) {
    369 					case 0: case 1:
    370 						pr->bcnt = 1;
    371 						break;
    372 					default:
    373 						p1[2] = '\0';
    374 						badcnt(p1);
    375 					}
    376 					break;
    377 				default:
    378 					p1[2] = '\0';
    379 					badconv(p1);
    380 				}
    381 				break;
    382 			default:
    383 				p1[1] = '\0';
    384 				badconv(p1);
    385 			}
    386 
    387 			/*
    388 			 * Copy to PR format string, set conversion character
    389 			 * pointer, update original.
    390 			 */
    391 			savech = *p2;
    392 			p1[0] = '\0';
    393 			pr->fmt = emalloc(strlen(fmtp) + 2);
    394 			(void)strcpy(pr->fmt, fmtp);
    395 			(void)strcat(pr->fmt, cs);
    396 			*p2 = savech;
    397 			pr->cchar = pr->fmt + (p1 - fmtp);
    398 			fmtp = p2;
    399 
    400 			/* Only one conversion character if byte count. */
    401 			if (!(pr->flags&F_ADDRESS) && fu->bcnt && nconv++)
    402 	    err("byte count with multiple conversion characters");
    403 		}
    404 		/*
    405 		 * If format unit byte count not specified, figure it out
    406 		 * so can adjust rep count later.
    407 		 */
    408 		if (!fu->bcnt)
    409 			for (pr = fu->nextpr; pr; pr = pr->nextpr)
    410 				fu->bcnt += pr->bcnt;
    411 	}
    412 	/*
    413 	 * If the format string interprets any data at all, and it's
    414 	 * not the same as the blocksize, and its last format unit
    415 	 * interprets any data at all, and has no iteration count,
    416 	 * repeat it as necessary.
    417 	 *
    418 	 * If, rep count is greater than 1, no trailing whitespace
    419 	 * gets output from the last iteration of the format unit.
    420 	 */
    421 	for (fu = fs->nextfu;; fu = fu->nextfu) {
    422 		if (!fu->nextfu && fs->bcnt < blocksize &&
    423 		    !(fu->flags&F_SETREP) && fu->bcnt)
    424 			fu->reps += (blocksize - fs->bcnt) / fu->bcnt;
    425 		if (fu->reps > 1) {
    426 			for (pr = fu->nextpr;; pr = pr->nextpr)
    427 				if (!pr->nextpr)
    428 					break;
    429 			for (p1 = pr->fmt, p2 = NULL; *p1; ++p1)
    430 				p2 = isspace(*p1) ? p1 : NULL;
    431 			if (p2)
    432 				pr->nospace = p2;
    433 		}
    434 		if (!fu->nextfu)
    435 			break;
    436 	}
    437 #ifdef DEBUG
    438 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
    439 		(void)printf("fmt:");
    440 		for (pr = fu->nextpr; pr; pr = pr->nextpr)
    441 			(void)printf(" {%s}", pr->fmt);
    442 		(void)printf("\n");
    443 	}
    444 #endif
    445 }
    446 
    447 void
    448 escape(p1)
    449 	register char *p1;
    450 {
    451 	register char *p2;
    452 
    453 	/* alphabetic escape sequences have to be done in place */
    454 	for (p2 = p1;; ++p1, ++p2) {
    455 		if (!*p1) {
    456 			*p2 = *p1;
    457 			break;
    458 		}
    459 		if (*p1 == '\\')
    460 			switch(*++p1) {
    461 			case 'a':
    462 			     /* *p2 = '\a'; */
    463 				*p2 = '\007';
    464 				break;
    465 			case 'b':
    466 				*p2 = '\b';
    467 				break;
    468 			case 'f':
    469 				*p2 = '\f';
    470 				break;
    471 			case 'n':
    472 				*p2 = '\n';
    473 				break;
    474 			case 'r':
    475 				*p2 = '\r';
    476 				break;
    477 			case 't':
    478 				*p2 = '\t';
    479 				break;
    480 			case 'v':
    481 				*p2 = '\v';
    482 				break;
    483 			default:
    484 				*p2 = *p1;
    485 				break;
    486 			}
    487 	}
    488 }
    489 
    490 void
    491 badcnt(s)
    492 	char *s;
    493 {
    494 	err("%s: bad byte count", s);
    495 }
    496 
    497 void
    498 badsfmt()
    499 {
    500 	err("%%s: requires a precision or a byte count\n");
    501 }
    502 
    503 void
    504 badfmt(fmt)
    505 	char *fmt;
    506 {
    507 	err("\"%s\": bad format\n", fmt);
    508 }
    509 
    510 void
    511 badconv(ch)
    512 	char *ch;
    513 {
    514 	err("%%%s: bad conversion character\n", ch);
    515 }
    516