xref: /netbsd-src/usr.bin/hexdump/parse.c (revision d0fed6c87ddc40a8bffa6f99e7433ddfc864dd83)
1 /*	$NetBSD: parse.c,v 1.4 1997/01/09 20:19:59 tls 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[] = "$NetBSD: parse.c,v 1.4 1997/01/09 20:19:59 tls 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 	int ch;
57 	char buf[2048 + 1];
58 
59 	if (!(fp = fopen(name, "r"))) {
60 		(void)fprintf(stderr, "hexdump: can't read %s.\n", name);
61 		exit(1);
62 	}
63 	while (fgets(buf, sizeof(buf), fp)) {
64 		if (!(p = index(buf, '\n'))) {
65 			(void)fprintf(stderr, "hexdump: line too long.\n");
66 			while ((ch = getchar()) != '\n' && ch != EOF);
67 			continue;
68 		}
69 		*p = '\0';
70 		for (p = buf; *p && isspace(*p); ++p);
71 		if (!*p || *p == '#')
72 			continue;
73 		add(p);
74 	}
75 	(void)fclose(fp);
76 }
77 
78 add(fmt)
79 	char *fmt;
80 {
81 	register char *p;
82 	static FS **nextfs;
83 	FS *tfs;
84 	FU *tfu, **nextfu;
85 	char *savep, *emalloc();
86 
87 	/* start new linked list of format units */
88 	/* NOSTRICT */
89 	tfs = (FS *)emalloc(sizeof(FS));
90 	if (!fshead)
91 		fshead = tfs;
92 	else
93 		*nextfs = tfs;
94 	nextfs = &tfs->nextfs;
95 	nextfu = &tfs->nextfu;
96 
97 	/* take the format string and break it up into format units */
98 	for (p = fmt;;) {
99 		/* skip leading white space */
100 		for (; isspace(*p); ++p);
101 		if (!*p)
102 			break;
103 
104 		/* allocate a new format unit and link it in */
105 		/* NOSTRICT */
106 		tfu = (FU *)emalloc(sizeof(FU));
107 		*nextfu = tfu;
108 		nextfu = &tfu->nextfu;
109 		tfu->reps = 1;
110 
111 		/* if leading digit, repetition count */
112 		if (isdigit(*p)) {
113 			for (savep = p; isdigit(*p); ++p);
114 			if (!isspace(*p) && *p != '/')
115 				badfmt(fmt);
116 			/* may overwrite either white space or slash */
117 			tfu->reps = atoi(savep);
118 			tfu->flags = F_SETREP;
119 			/* skip trailing white space */
120 			for (++p; isspace(*p); ++p);
121 		}
122 
123 		/* skip slash and trailing white space */
124 		if (*p == '/')
125 			while (isspace(*++p));
126 
127 		/* byte count */
128 		if (isdigit(*p)) {
129 			for (savep = p; isdigit(*p); ++p);
130 			if (!isspace(*p))
131 				badfmt(fmt);
132 			tfu->bcnt = atoi(savep);
133 			/* skip trailing white space */
134 			for (++p; isspace(*p); ++p);
135 		}
136 
137 		/* format */
138 		if (*p != '"')
139 			badfmt(fmt);
140 		for (savep = ++p; *p != '"';)
141 			if (*p++ == 0)
142 				badfmt(fmt);
143 		if (!(tfu->fmt = malloc(p - savep + 1)))
144 			nomem();
145 		(void) strncpy(tfu->fmt, savep, p - savep);
146 		tfu->fmt[p - savep] = '\0';
147 		escape(tfu->fmt);
148 		p++;
149 	}
150 }
151 
152 static char *spec = ".#-+ 0123456789";
153 size(fs)
154 	FS *fs;
155 {
156 	register FU *fu;
157 	register int bcnt, cursize;
158 	register char *fmt;
159 	int prec;
160 
161 	/* figure out the data block size needed for each format unit */
162 	for (cursize = 0, fu = fs->nextfu; fu; fu = fu->nextfu) {
163 		if (fu->bcnt) {
164 			cursize += fu->bcnt * fu->reps;
165 			continue;
166 		}
167 		for (bcnt = prec = 0, fmt = fu->fmt; *fmt; ++fmt) {
168 			if (*fmt != '%')
169 				continue;
170 			/*
171 			 * skip any special chars -- save precision in
172 			 * case it's a %s format.
173 			 */
174 			while (index(spec + 1, *++fmt));
175 			if (*fmt == '.' && isdigit(*++fmt)) {
176 				prec = atoi(fmt);
177 				while (isdigit(*++fmt));
178 			}
179 			switch(*fmt) {
180 			case 'c':
181 				bcnt += 1;
182 				break;
183 			case 'd': case 'i': case 'o': case 'u':
184 			case 'x': case 'X':
185 				bcnt += 4;
186 				break;
187 			case 'e': case 'E': case 'f': case 'g': case 'G':
188 				bcnt += 8;
189 				break;
190 			case 's':
191 				bcnt += prec;
192 				break;
193 			case '_':
194 				switch(*++fmt) {
195 				case 'c': case 'p': case 'u':
196 					bcnt += 1;
197 					break;
198 				}
199 			}
200 		}
201 		cursize += bcnt * fu->reps;
202 	}
203 	return(cursize);
204 }
205 
206 rewrite(fs)
207 	FS *fs;
208 {
209 	enum { NOTOKAY, USEBCNT, USEPREC } sokay;
210 	register PR *pr, **nextpr;
211 	register FU *fu;
212 	register char *p1, *p2;
213 	char savech, *fmtp;
214 	int nconv, prec;
215 
216 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
217 		/*
218 		 * break each format unit into print units; each
219 		 * conversion character gets its own.
220 		 */
221 		for (nconv = 0, fmtp = fu->fmt; *fmtp; nextpr = &pr->nextpr) {
222 			/* NOSTRICT */
223 			pr = (PR *)emalloc(sizeof(PR));
224 			if (!fu->nextpr)
225 				fu->nextpr = pr;
226 			else
227 				*nextpr = pr;
228 
229 			/* skip preceding text and up to the next % sign */
230 			for (p1 = fmtp; *p1 && *p1 != '%'; ++p1);
231 
232 			/* only text in the string */
233 			if (!*p1) {
234 				pr->fmt = fmtp;
235 				pr->flags = F_TEXT;
236 				break;
237 			}
238 
239 			/*
240 			 * get precision for %s -- if have a byte count, don't
241 			 * need it.
242 			 */
243 			if (fu->bcnt) {
244 				sokay = USEBCNT;
245 				/* skip to conversion character */
246 				for (++p1; index(spec, *p1); ++p1);
247 			} else {
248 				/* skip any special chars, field width */
249 				while (index(spec + 1, *++p1));
250 				if (*p1 == '.' && isdigit(*++p1)) {
251 					sokay = USEPREC;
252 					prec = atoi(p1);
253 					while (isdigit(*++p1));
254 				}
255 				else
256 					sokay = NOTOKAY;
257 			}
258 
259 			p2 = p1 + 1;		/* set end pointer */
260 
261 			/*
262 			 * figure out the byte count for each conversion;
263 			 * rewrite the format as necessary, set up blank-
264 			 * padding for end of data.
265 			 */
266 			switch(*p1) {
267 			case 'c':
268 				pr->flags = F_CHAR;
269 				switch(fu->bcnt) {
270 				case 0: case 1:
271 					pr->bcnt = 1;
272 					break;
273 				default:
274 					p1[1] = '\0';
275 					badcnt(p1);
276 				}
277 				break;
278 			case 'd': case 'i':
279 				pr->flags = F_INT;
280 				goto sw1;
281 			case 'l':
282 				++p2;
283 				switch(p1[1]) {
284 				case 'd': case 'i':
285 					++p1;
286 					pr->flags = F_INT;
287 					goto sw1;
288 				case 'o': case 'u': case 'x': case 'X':
289 					++p1;
290 					pr->flags = F_UINT;
291 					goto sw1;
292 				default:
293 					p1[2] = '\0';
294 					badconv(p1);
295 				}
296 				/* NOTREACHED */
297 			case 'o': case 'u': case 'x': case 'X':
298 				pr->flags = F_UINT;
299 sw1:				switch(fu->bcnt) {
300 				case 0: case 4:
301 					pr->bcnt = 4;
302 					break;
303 				case 1:
304 					pr->bcnt = 1;
305 					break;
306 				case 2:
307 					pr->bcnt = 2;
308 					break;
309 				default:
310 					p1[1] = '\0';
311 					badcnt(p1);
312 				}
313 				break;
314 			case 'e': case 'E': case 'f': case 'g': case 'G':
315 				pr->flags = F_DBL;
316 				switch(fu->bcnt) {
317 				case 0: case 8:
318 					pr->bcnt = 8;
319 					break;
320 				case 4:
321 					pr->bcnt = 4;
322 					break;
323 				default:
324 					p1[1] = '\0';
325 					badcnt(p1);
326 				}
327 				break;
328 			case 's':
329 				pr->flags = F_STR;
330 				switch(sokay) {
331 				case NOTOKAY:
332 					badsfmt();
333 				case USEBCNT:
334 					pr->bcnt = fu->bcnt;
335 					break;
336 				case USEPREC:
337 					pr->bcnt = prec;
338 					break;
339 				}
340 				break;
341 			case '_':
342 				++p2;
343 				switch(p1[1]) {
344 				case 'A':
345 					endfu = fu;
346 					fu->flags |= F_IGNORE;
347 					/* FALLTHROUGH */
348 				case 'a':
349 					pr->flags = F_ADDRESS;
350 					++p2;
351 					switch(p1[2]) {
352 					case 'd': case 'o': case'x':
353 						*p1 = 'q';
354 						p1[1] = p1[2];
355 						break;
356 					default:
357 						p1[3] = '\0';
358 						badconv(p1);
359 					}
360 					break;
361 				case 'c':
362 					pr->flags = F_C;
363 					/* *p1 = 'c';	set in conv_c */
364 					goto sw2;
365 				case 'p':
366 					pr->flags = F_P;
367 					*p1 = 'c';
368 					goto sw2;
369 				case 'u':
370 					pr->flags = F_U;
371 					/* *p1 = 'c';	set in conv_u */
372 sw2:					switch(fu->bcnt) {
373 					case 0: case 1:
374 						pr->bcnt = 1;
375 						break;
376 					default:
377 						p1[2] = '\0';
378 						badcnt(p1);
379 					}
380 					break;
381 				default:
382 					p1[2] = '\0';
383 					badconv(p1);
384 				}
385 				break;
386 			default:
387 				p1[1] = '\0';
388 				badconv(p1);
389 			}
390 
391 			/*
392 			 * copy to PR format string, set conversion character
393 			 * pointer, update original.
394 			 */
395 			savech = *p2;
396 			p1[(pr->flags&F_ADDRESS)?2:1] = '\0';
397 			if (!(pr->fmt = strdup(fmtp)))
398 				nomem();
399 			*p2 = savech;
400 			pr->cchar = pr->fmt + (p1 - fmtp);
401 			fmtp = p2;
402 
403 			/* only one conversion character if byte count */
404 			if (!(pr->flags&F_ADDRESS) && fu->bcnt && nconv++) {
405 				(void)fprintf(stderr,
406 				    "hexdump: byte count with multiple conversion characters.\n");
407 				exit(1);
408 			}
409 		}
410 		/*
411 		 * if format unit byte count not specified, figure it out
412 		 * so can adjust rep count later.
413 		 */
414 		if (!fu->bcnt)
415 			for (pr = fu->nextpr; pr; pr = pr->nextpr)
416 				fu->bcnt += pr->bcnt;
417 	}
418 	/*
419 	 * if the format string interprets any data at all, and it's
420 	 * not the same as the blocksize, and its last format unit
421 	 * interprets any data at all, and has no iteration count,
422 	 * repeat it as necessary.
423 	 *
424 	 * if, rep count is greater than 1, no trailing whitespace
425 	 * gets output from the last iteration of the format unit.
426 	 */
427 	for (fu = fs->nextfu;; fu = fu->nextfu) {
428 		if (!fu->nextfu && fs->bcnt < blocksize &&
429 		    !(fu->flags&F_SETREP) && fu->bcnt)
430 			fu->reps += (blocksize - fs->bcnt) / fu->bcnt;
431 		if (fu->reps > 1) {
432 			for (pr = fu->nextpr;; pr = pr->nextpr)
433 				if (!pr->nextpr)
434 					break;
435 			for (p1 = pr->fmt, p2 = NULL; *p1; ++p1)
436 				p2 = isspace(*p1) ? p1 : NULL;
437 			if (p2)
438 				pr->nospace = p2;
439 		}
440 		if (!fu->nextfu)
441 			break;
442 	}
443 }
444 
445 
446 escape(p1)
447 	register char *p1;
448 {
449 	register char *p2;
450 
451 	/* alphabetic escape sequences have to be done in place */
452 	for (p2 = p1;; ++p1, ++p2) {
453 		if (!*p1) {
454 			*p2 = *p1;
455 			break;
456 		}
457 		if (*p1 == '\\')
458 			switch(*++p1) {
459 			case 'a':
460 			     /* *p2 = '\a'; */
461 				*p2 = '\007';
462 				break;
463 			case 'b':
464 				*p2 = '\b';
465 				break;
466 			case 'f':
467 				*p2 = '\f';
468 				break;
469 			case 'n':
470 				*p2 = '\n';
471 				break;
472 			case 'r':
473 				*p2 = '\r';
474 				break;
475 			case 't':
476 				*p2 = '\t';
477 				break;
478 			case 'v':
479 				*p2 = '\v';
480 				break;
481 			default:
482 				*p2 = *p1;
483 				break;
484 			}
485 	}
486 }
487 
488 badcnt(s)
489 	char *s;
490 {
491 	(void)fprintf(stderr,
492 	    "hexdump: bad byte count for conversion character %s.\n", s);
493 	exit(1);
494 }
495 
496 badsfmt()
497 {
498 	(void)fprintf(stderr,
499 	    "hexdump: %%s requires a precision or a byte count.\n");
500 	exit(1);
501 }
502 
503 badfmt(fmt)
504 	char *fmt;
505 {
506 	(void)fprintf(stderr, "hexdump: bad format {%s}\n", fmt);
507 	exit(1);
508 }
509 
510 badconv(ch)
511 	char *ch;
512 {
513 	(void)fprintf(stderr, "hexdump: bad conversion character %%%s.\n", ch);
514 	exit(1);
515 }
516