xref: /openbsd-src/usr.bin/hexdump/parse.c (revision b2ea75c1b17e1a9a339660e7ed45cd24946b230e)
1 /*	$OpenBSD: parse.c,v 1.5 1999/02/07 20:54:09 aaron Exp $	*/
2 
3 /*
4  * Copyright (c) 1989 The Regents of the University of California.
5  * 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 /*static char sccsid[] = "from: @(#)parse.c	5.6 (Berkeley) 3/9/91";*/
38 static char rcsid[] = "$OpenBSD: parse.c,v 1.5 1999/02/07 20:54:09 aaron Exp $";
39 #endif /* not lint */
40 
41 #include <sys/types.h>
42 #include <sys/file.h>
43 #include <stdio.h>
44 #include <stdlib.h>
45 #include <ctype.h>
46 #include <string.h>
47 #include "hexdump.h"
48 
49 FU *endfu;					/* format at end-of-data */
50 
51 addfile(name)
52 	char *name;
53 {
54 	register char *p;
55 	FILE *fp;
56 	size_t len;
57 
58 	if (!(fp = fopen(name, "r"))) {
59 		(void)fprintf(stderr, "hexdump: can't read %s.\n", name);
60 		exit(1);
61 	}
62 	while ((p = fgetln(fp, &len))) {
63 		if (*(p + len - 1) == '\n')
64 			*(p + len - 1) = '\0';
65 		else {
66 			(void)fprintf(stderr, "hexdump: incomplete line.\n");
67 			continue;
68 		}
69 		for (; *p && isspace(*p); ++p);
70 		if (!*p || *p == '#')
71 			continue;
72 		add(p);
73 	}
74 	(void)fclose(fp);
75 }
76 
77 add(fmt)
78 	char *fmt;
79 {
80 	register char *p;
81 	static FS **nextfs;
82 	FS *tfs;
83 	FU *tfu, **nextfu;
84 	char *savep, *emalloc();
85 
86 	/* start new linked list of format units */
87 	/* NOSTRICT */
88 	tfs = (FS *)emalloc(sizeof(FS));
89 	if (!fshead)
90 		fshead = tfs;
91 	else
92 		*nextfs = tfs;
93 	nextfs = &tfs->nextfs;
94 	nextfu = &tfs->nextfu;
95 
96 	/* take the format string and break it up into format units */
97 	for (p = fmt;;) {
98 		/* skip leading white space */
99 		for (; isspace(*p); ++p);
100 		if (!*p)
101 			break;
102 
103 		/* allocate a new format unit and link it in */
104 		/* NOSTRICT */
105 		tfu = (FU *)emalloc(sizeof(FU));
106 		*nextfu = tfu;
107 		nextfu = &tfu->nextfu;
108 		tfu->reps = 1;
109 
110 		/* if leading digit, repetition count */
111 		if (isdigit(*p)) {
112 			for (savep = p; isdigit(*p); ++p);
113 			if (!isspace(*p) && *p != '/')
114 				badfmt(fmt);
115 			/* may overwrite either white space or slash */
116 			tfu->reps = atoi(savep);
117 			tfu->flags = F_SETREP;
118 			/* skip trailing white space */
119 			for (++p; isspace(*p); ++p);
120 		}
121 
122 		/* skip slash and trailing white space */
123 		if (*p == '/')
124 			while (isspace(*++p));
125 
126 		/* byte count */
127 		if (isdigit(*p)) {
128 			for (savep = p; isdigit(*p); ++p);
129 			if (!isspace(*p))
130 				badfmt(fmt);
131 			tfu->bcnt = atoi(savep);
132 			/* skip trailing white space */
133 			for (++p; isspace(*p); ++p);
134 		}
135 
136 		/* format */
137 		if (*p != '"')
138 			badfmt(fmt);
139 		for (savep = ++p; *p != '"';)
140 			if (*p++ == 0)
141 				badfmt(fmt);
142 		if (!(tfu->fmt = malloc(p - savep + 1)))
143 			nomem();
144 		(void) strncpy(tfu->fmt, savep, p - savep);
145 		tfu->fmt[p - savep] = '\0';
146 		escape(tfu->fmt);
147 		p++;
148 	}
149 	/* no single fu in fmt */
150 	if (tfs->nextfu == NULL)
151 		badfmt(fmt);
152 }
153 
154 static char *spec = ".#-+ 0123456789";
155 size(fs)
156 	FS *fs;
157 {
158 	register FU *fu;
159 	register int bcnt, cursize;
160 	register char *fmt;
161 	int prec;
162 
163 	/* figure out the data block size needed for each format unit */
164 	for (cursize = 0, fu = fs->nextfu; fu; fu = fu->nextfu) {
165 		if (fu->bcnt) {
166 			cursize += fu->bcnt * fu->reps;
167 			continue;
168 		}
169 		for (bcnt = prec = 0, fmt = fu->fmt; *fmt; ++fmt) {
170 			if (*fmt != '%')
171 				continue;
172 			/*
173 			 * skip any special chars -- save precision in
174 			 * case it's a %s format.
175 			 */
176 			while (strchr(spec + 1, *++fmt));
177 			if (*fmt == '.' && isdigit(*++fmt)) {
178 				prec = atoi(fmt);
179 				while (isdigit(*++fmt));
180 			}
181 			switch(*fmt) {
182 			case 'c':
183 				bcnt += 1;
184 				break;
185 			case 'd': case 'i': case 'o': case 'u':
186 			case 'x': case 'X':
187 				bcnt += 4;
188 				break;
189 			case 'e': case 'E': case 'f': case 'g': case 'G':
190 				bcnt += 8;
191 				break;
192 			case 's':
193 				bcnt += prec;
194 				break;
195 			case '_':
196 				switch(*++fmt) {
197 				case 'c': case 'p': case 'u':
198 					bcnt += 1;
199 					break;
200 				}
201 			}
202 		}
203 		cursize += bcnt * fu->reps;
204 	}
205 	return(cursize);
206 }
207 
208 rewrite(fs)
209 	FS *fs;
210 {
211 	enum { NOTOKAY, USEBCNT, USEPREC } sokay;
212 	register PR *pr, **nextpr;
213 	register FU *fu;
214 	register char *p1, *p2;
215 	char savech, *fmtp;
216 	int nconv, prec;
217 
218 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
219 		/*
220 		 * break each format unit into print units; each
221 		 * conversion character gets its own.
222 		 */
223 		for (nconv = 0, fmtp = fu->fmt; *fmtp; nextpr = &pr->nextpr) {
224 			/* NOSTRICT */
225 			pr = (PR *)emalloc(sizeof(PR));
226 			if (!fu->nextpr)
227 				fu->nextpr = pr;
228 			else
229 				*nextpr = pr;
230 
231 			/* skip preceding text and up to the next % sign */
232 			for (p1 = fmtp; *p1 && *p1 != '%'; ++p1);
233 
234 			/* only text in the string */
235 			if (!*p1) {
236 				pr->fmt = fmtp;
237 				pr->flags = F_TEXT;
238 				break;
239 			}
240 
241 			/*
242 			 * get precision for %s -- if have a byte count, don't
243 			 * need it.
244 			 */
245 			if (fu->bcnt) {
246 				sokay = USEBCNT;
247 				/* skip to conversion character */
248 				for (++p1; strchr(spec, *p1); ++p1);
249 			} else {
250 				/* skip any special chars, field width */
251 				while (strchr(spec + 1, *++p1));
252 				if (*p1 == '.' && isdigit(*++p1)) {
253 					sokay = USEPREC;
254 					prec = atoi(p1);
255 					while (isdigit(*++p1));
256 				}
257 				else
258 					sokay = NOTOKAY;
259 			}
260 
261 			p2 = p1 + 1;		/* set end pointer */
262 
263 			/*
264 			 * figure out the byte count for each conversion;
265 			 * rewrite the format as necessary, set up blank-
266 			 * padding for end of data.
267 			 */
268 			switch(*p1) {
269 			case 'c':
270 				pr->flags = F_CHAR;
271 				switch(fu->bcnt) {
272 				case 0: case 1:
273 					pr->bcnt = 1;
274 					break;
275 				default:
276 					p1[1] = '\0';
277 					badcnt(p1);
278 				}
279 				break;
280 			case 'd': case 'i':
281 				pr->flags = F_INT;
282 				goto sw1;
283 			case 'l':
284 				++p2;
285 				switch(p1[1]) {
286 				case 'd': case 'i':
287 					++p1;
288 					pr->flags = F_INT;
289 					goto sw1;
290 				case 'o': case 'u': case 'x': case 'X':
291 					++p1;
292 					pr->flags = F_UINT;
293 					goto sw1;
294 				default:
295 					p1[2] = '\0';
296 					badconv(p1);
297 				}
298 				/* NOTREACHED */
299 			case 'o': case 'u': case 'x': case 'X':
300 				pr->flags = F_UINT;
301 sw1:				switch(fu->bcnt) {
302 				case 0: case 4:
303 					pr->bcnt = 4;
304 					break;
305 				case 1:
306 					pr->bcnt = 1;
307 					break;
308 				case 2:
309 					pr->bcnt = 2;
310 					break;
311 				default:
312 					p1[1] = '\0';
313 					badcnt(p1);
314 				}
315 				break;
316 			case 'e': case 'E': case 'f': case 'g': case 'G':
317 				pr->flags = F_DBL;
318 				switch(fu->bcnt) {
319 				case 0: case 8:
320 					pr->bcnt = 8;
321 					break;
322 				case 4:
323 					pr->bcnt = 4;
324 					break;
325 				default:
326 					p1[1] = '\0';
327 					badcnt(p1);
328 				}
329 				break;
330 			case 's':
331 				pr->flags = F_STR;
332 				switch(sokay) {
333 				case NOTOKAY:
334 					badsfmt();
335 				case USEBCNT:
336 					pr->bcnt = fu->bcnt;
337 					break;
338 				case USEPREC:
339 					pr->bcnt = prec;
340 					break;
341 				}
342 				break;
343 			case '_':
344 				++p2;
345 				switch(p1[1]) {
346 				case 'A':
347 					endfu = fu;
348 					fu->flags |= F_IGNORE;
349 					/* FALLTHROUGH */
350 				case 'a':
351 					pr->flags = F_ADDRESS;
352 					++p2;
353 					switch(p1[2]) {
354 					case 'd': case 'o': case'x':
355 						*p1 = 'q';
356 						p1[1] = p1[2];
357 						break;
358 					default:
359 						p1[3] = '\0';
360 						badconv(p1);
361 					}
362 					break;
363 				case 'c':
364 					pr->flags = F_C;
365 					/* *p1 = 'c';	set in conv_c */
366 					goto sw2;
367 				case 'p':
368 					pr->flags = F_P;
369 					*p1 = 'c';
370 					goto sw2;
371 				case 'u':
372 					pr->flags = F_U;
373 					/* *p1 = 'c';	set in conv_u */
374 sw2:					switch(fu->bcnt) {
375 					case 0: case 1:
376 						pr->bcnt = 1;
377 						break;
378 					default:
379 						p1[2] = '\0';
380 						badcnt(p1);
381 					}
382 					break;
383 				default:
384 					p1[2] = '\0';
385 					badconv(p1);
386 				}
387 				break;
388 			default:
389 				p1[1] = '\0';
390 				badconv(p1);
391 			}
392 
393 			/*
394 			 * copy to PR format string, set conversion character
395 			 * pointer, update original.
396 			 */
397 			savech = *p2;
398 			p1[(pr->flags&F_ADDRESS)?2:1] = '\0';
399 			if (!(pr->fmt = strdup(fmtp)))
400 				nomem();
401 			*p2 = savech;
402 			pr->cchar = pr->fmt + (p1 - fmtp);
403 			fmtp = p2;
404 
405 			/* only one conversion character if byte count */
406 			if (!(pr->flags&F_ADDRESS) && fu->bcnt && nconv++) {
407 				(void)fprintf(stderr,
408 				    "hexdump: byte count with multiple conversion characters.\n");
409 				exit(1);
410 			}
411 		}
412 		/*
413 		 * if format unit byte count not specified, figure it out
414 		 * so can adjust rep count later.
415 		 */
416 		if (!fu->bcnt)
417 			for (pr = fu->nextpr; pr; pr = pr->nextpr)
418 				fu->bcnt += pr->bcnt;
419 	}
420 	/*
421 	 * if the format string interprets any data at all, and it's
422 	 * not the same as the blocksize, and its last format unit
423 	 * interprets any data at all, and has no iteration count,
424 	 * repeat it as necessary.
425 	 *
426 	 * if, rep count is greater than 1, no trailing whitespace
427 	 * gets output from the last iteration of the format unit.
428 	 */
429 	for (fu = fs->nextfu;; fu = fu->nextfu) {
430 		if (!fu->nextfu && fs->bcnt < blocksize &&
431 		    !(fu->flags&F_SETREP) && fu->bcnt)
432 			fu->reps += (blocksize - fs->bcnt) / fu->bcnt;
433 		if (fu->reps > 1) {
434 			for (pr = fu->nextpr;; pr = pr->nextpr)
435 				if (!pr->nextpr)
436 					break;
437 			for (p1 = pr->fmt, p2 = NULL; *p1; ++p1)
438 				p2 = isspace(*p1) ? p1 : NULL;
439 			if (p2)
440 				pr->nospace = p2;
441 		}
442 		if (!fu->nextfu)
443 			break;
444 	}
445 }
446 
447 
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 badcnt(s)
491 	char *s;
492 {
493 	(void)fprintf(stderr,
494 	    "hexdump: bad byte count for conversion character %s.\n", s);
495 	exit(1);
496 }
497 
498 badsfmt()
499 {
500 	(void)fprintf(stderr,
501 	    "hexdump: %%s requires a precision or a byte count.\n");
502 	exit(1);
503 }
504 
505 badfmt(fmt)
506 	char *fmt;
507 {
508 	(void)fprintf(stderr, "hexdump: bad format {%s}\n", fmt);
509 	exit(1);
510 }
511 
512 badconv(ch)
513 	char *ch;
514 {
515 	(void)fprintf(stderr, "hexdump: bad conversion character %%%s.\n", ch);
516 	exit(1);
517 }
518