xref: /freebsd/usr.bin/hexdump/parse.c (revision 2008043f386721d58158e37e0d7e50df8095942d)
1 /*-
2  * SPDX-License-Identifier: BSD-3-Clause
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. Neither the name of the University nor the names of its contributors
16  *    may be used to endorse or promote products derived from this software
17  *    without specific prior written permission.
18  *
19  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
20  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
23  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29  * SUCH DAMAGE.
30  */
31 
32 #ifndef lint
33 #if 0
34 static char sccsid[] = "@(#)parse.c	8.1 (Berkeley) 6/6/93";
35 #endif
36 #endif /* not lint */
37 #include <sys/cdefs.h>
38 #include <sys/types.h>
39 
40 #include <err.h>
41 #include <fcntl.h>
42 #include <stdio.h>
43 #include <stdlib.h>
44 #include <ctype.h>
45 #include <string.h>
46 #include "hexdump.h"
47 
48 FU *endfu;					/* format at end-of-data */
49 
50 void
51 addfile(const char *name)
52 {
53 	unsigned char *p;
54 	FILE *fp;
55 	int ch;
56 	char buf[2048 + 1];
57 
58 	if ((fp = fopen(name, "r")) == NULL)
59 		err(1, "%s", name);
60 	while (fgets(buf, sizeof(buf), fp)) {
61 		if (!(p = strchr(buf, '\n'))) {
62 			warnx("line too long");
63 			while ((ch = getchar()) != '\n' && ch != EOF);
64 			continue;
65 		}
66 		*p = '\0';
67 		for (p = buf; *p && isspace(*p); ++p);
68 		if (!*p || *p == '#')
69 			continue;
70 		add(p);
71 	}
72 	(void)fclose(fp);
73 }
74 
75 void
76 add(const char *fmt)
77 {
78 	unsigned const char *p, *savep;
79 	static FS **nextfs;
80 	FS *tfs;
81 	FU *tfu, **nextfu;
82 
83 	/* start new linked list of format units */
84 	if ((tfs = calloc(1, sizeof(FS))) == NULL)
85 		err(1, NULL);
86 	if (!fshead)
87 		fshead = tfs;
88 	else
89 		*nextfs = tfs;
90 	nextfs = &tfs->nextfs;
91 	nextfu = &tfs->nextfu;
92 
93 	/* take the format string and break it up into format units */
94 	for (p = fmt;;) {
95 		/* skip leading white space */
96 		for (; isspace(*p); ++p);
97 		if (!*p)
98 			break;
99 
100 		/* allocate a new format unit and link it in */
101 		if ((tfu = calloc(1, sizeof(FU))) == NULL)
102 			err(1, NULL);
103 		*nextfu = tfu;
104 		nextfu = &tfu->nextfu;
105 		tfu->reps = 1;
106 
107 		/* if leading digit, repetition count */
108 		if (isdigit(*p)) {
109 			for (savep = p; isdigit(*p); ++p);
110 			if (!isspace(*p) && *p != '/')
111 				badfmt(fmt);
112 			/* may overwrite either white space or slash */
113 			tfu->reps = atoi(savep);
114 			tfu->flags = F_SETREP;
115 			/* skip trailing white space */
116 			for (++p; isspace(*p); ++p);
117 		}
118 
119 		/* skip slash and trailing white space */
120 		if (*p == '/')
121 			while (isspace(*++p));
122 
123 		/* byte count */
124 		if (isdigit(*p)) {
125 			for (savep = p; isdigit(*p); ++p);
126 			if (!isspace(*p))
127 				badfmt(fmt);
128 			tfu->bcnt = atoi(savep);
129 			/* skip trailing white space */
130 			for (++p; isspace(*p); ++p);
131 		}
132 
133 		/* format */
134 		if (*p != '"')
135 			badfmt(fmt);
136 		for (savep = ++p; *p != '"';)
137 			if (*p++ == 0)
138 				badfmt(fmt);
139 		if (!(tfu->fmt = malloc(p - savep + 1)))
140 			err(1, NULL);
141 		(void) strlcpy(tfu->fmt, savep, p - savep + 1);
142 		escape(tfu->fmt);
143 		p++;
144 	}
145 }
146 
147 static const char *spec = ".#-+ 0123456789";
148 
149 int
150 size(FS *fs)
151 {
152 	FU *fu;
153 	int bcnt, cursize;
154 	unsigned char *fmt;
155 	int prec;
156 
157 	/* figure out the data block size needed for each format unit */
158 	for (cursize = 0, fu = fs->nextfu; fu; fu = fu->nextfu) {
159 		if (fu->bcnt) {
160 			cursize += fu->bcnt * fu->reps;
161 			continue;
162 		}
163 		for (bcnt = prec = 0, fmt = fu->fmt; *fmt; ++fmt) {
164 			if (*fmt != '%')
165 				continue;
166 			/*
167 			 * skip any special chars -- save precision in
168 			 * case it's a %s format.
169 			 */
170 			while (*++fmt != 0 && strchr(spec + 1, *fmt) != NULL)
171 				;
172 			if (*fmt == 0)
173 				badnoconv();
174 			if (*fmt == '.' && isdigit(*++fmt)) {
175 				prec = atoi(fmt);
176 				while (isdigit(*++fmt));
177 			}
178 			switch(*fmt) {
179 			case 'c':
180 				bcnt += 1;
181 				break;
182 			case 'd': case 'i': case 'o': case 'u':
183 			case 'x': case 'X':
184 				bcnt += 4;
185 				break;
186 			case 'e': case 'E': case 'f': case 'g': case 'G':
187 				bcnt += 8;
188 				break;
189 			case 's':
190 				bcnt += prec;
191 				break;
192 			case '_':
193 				switch(*++fmt) {
194 				case 'c': case 'p': case 'u':
195 					bcnt += 1;
196 					break;
197 				}
198 			}
199 		}
200 		cursize += bcnt * fu->reps;
201 	}
202 	return (cursize);
203 }
204 
205 void
206 rewrite(FS *fs)
207 {
208 	enum { NOTOKAY, USEBCNT, USEPREC } sokay;
209 	PR *pr, **nextpr;
210 	FU *fu;
211 	unsigned char *p1, *p2, *fmtp;
212 	char savech, cs[3];
213 	int nconv, prec;
214 
215 	prec = 0;
216 
217 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
218 		/*
219 		 * Break each format unit into print units; each conversion
220 		 * character gets its own.
221 		 */
222 		nextpr = &fu->nextpr;
223 		for (nconv = 0, fmtp = fu->fmt; *fmtp; nextpr = &pr->nextpr) {
224 			if ((pr = calloc(1, sizeof(PR))) == NULL)
225 				err(1, NULL);
226 			*nextpr = pr;
227 
228 			/* Skip preceding text and up to the next % sign. */
229 			for (p1 = fmtp; *p1 && *p1 != '%'; ++p1);
230 
231 			/* Only text in the string. */
232 			if (!*p1) {
233 				pr->fmt = fmtp;
234 				pr->flags = F_TEXT;
235 				break;
236 			}
237 
238 			/*
239 			 * Get precision for %s -- if have a byte count, don't
240 			 * need it.
241 			 */
242 			if (fu->bcnt) {
243 				sokay = USEBCNT;
244 				/* Skip to conversion character. */
245 				while (*++p1 != 0 && strchr(spec, *p1) != NULL)
246 					;
247 				if (*p1 == 0)
248 					badnoconv();
249 			} else {
250 				/* Skip any special chars, field width. */
251 				while (*++p1 != 0 && strchr(spec + 1, *p1) != NULL)
252 					;
253 				if (*p1 == 0)
254 					badnoconv();
255 				if (*p1 == '.' && isdigit(*++p1)) {
256 					sokay = USEPREC;
257 					prec = atoi(p1);
258 					while (isdigit(*++p1));
259 				} else
260 					sokay = NOTOKAY;
261 			}
262 
263 			p2 = *p1 ? p1 + 1 : p1;	/* Set end pointer -- make sure
264 						 * that it's non-NUL/-NULL first
265 						 * though. */
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 				case 8:
305 					pr->bcnt = 8;
306 					break;
307 				default:
308 					p1[1] = '\0';
309 					badcnt(p1);
310 				}
311 				break;
312 			case 'e': case 'E': case 'f': case 'g': case 'G':
313 				pr->flags = F_DBL;
314 				switch(fu->bcnt) {
315 				case 0: case 8:
316 					pr->bcnt = 8;
317 					break;
318 				case 4:
319 					pr->bcnt = 4;
320 					break;
321 				default:
322 					if (fu->bcnt == sizeof(long double)) {
323 						cs[2] = '\0';
324 						cs[1] = cs[0];
325 						cs[0] = 'L';
326 						pr->bcnt = sizeof(long double);
327 					} else {
328 						p1[1] = '\0';
329 						badcnt(p1);
330 					}
331 				}
332 				break;
333 			case 's':
334 				pr->flags = F_STR;
335 				switch(sokay) {
336 				case NOTOKAY:
337 					badsfmt();
338 				case USEBCNT:
339 					pr->bcnt = fu->bcnt;
340 					break;
341 				case USEPREC:
342 					pr->bcnt = prec;
343 					break;
344 				}
345 				break;
346 			case '_':
347 				++p2;
348 				switch(p1[1]) {
349 				case 'A':
350 					endfu = fu;
351 					fu->flags |= F_IGNORE;
352 					/* FALLTHROUGH */
353 				case 'a':
354 					pr->flags = F_ADDRESS;
355 					++p2;
356 					switch(p1[2]) {
357 					case 'd': case 'o': case'x':
358 						cs[0] = 'q';
359 						cs[1] = p1[2];
360 						cs[2] = '\0';
361 						break;
362 					default:
363 						p1[3] = '\0';
364 						badconv(p1);
365 					}
366 					break;
367 				case 'c':
368 					pr->flags = F_C;
369 					/* cs[0] = 'c';	set in conv_c */
370 					goto isint2;
371 				case 'p':
372 					pr->flags = F_P;
373 					cs[0] = 'c';
374 					goto isint2;
375 				case 'u':
376 					pr->flags = F_U;
377 					/* cs[0] = 'c';	set in conv_u */
378 isint2:					switch(fu->bcnt) {
379 					case 0: case 1:
380 						pr->bcnt = 1;
381 						break;
382 					default:
383 						p1[2] = '\0';
384 						badcnt(p1);
385 					}
386 					break;
387 				default:
388 					p1[2] = '\0';
389 					badconv(p1);
390 				}
391 				break;
392 			default:
393 				p1[1] = '\0';
394 				badconv(p1);
395 			}
396 
397 			/*
398 			 * Copy to PR format string, set conversion character
399 			 * pointer, update original.
400 			 */
401 			savech = *p2;
402 			p1[0] = '\0';
403 			if (asprintf(&pr->fmt, "%s%s", fmtp, cs) == -1)
404 				err(1, NULL);
405 			*p2 = savech;
406 			pr->cchar = pr->fmt + (p1 - fmtp);
407 			fmtp = p2;
408 
409 			/* Only one conversion character if byte count. */
410 			if (!(pr->flags&F_ADDRESS) && fu->bcnt && nconv++)
411 	    errx(1, "byte count with multiple conversion characters");
412 		}
413 		/*
414 		 * If format unit byte count not specified, figure it out
415 		 * so can adjust rep count later.
416 		 */
417 		if (!fu->bcnt)
418 			for (pr = fu->nextpr; pr; pr = pr->nextpr)
419 				fu->bcnt += pr->bcnt;
420 	}
421 	/*
422 	 * If the format string interprets any data at all, and it's
423 	 * not the same as the blocksize, and its last format unit
424 	 * interprets any data at all, and has no iteration count,
425 	 * repeat it as necessary.
426 	 *
427 	 * If, rep count is greater than 1, no trailing whitespace
428 	 * gets output from the last iteration of the format unit.
429 	 */
430 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
431 		if (!fu->nextfu && fs->bcnt < blocksize &&
432 		    !(fu->flags&F_SETREP) && fu->bcnt)
433 			fu->reps += (blocksize - fs->bcnt) / fu->bcnt;
434 		if (fu->reps > 1) {
435 			for (pr = fu->nextpr;; pr = pr->nextpr)
436 				if (!pr->nextpr)
437 					break;
438 			for (p1 = pr->fmt, p2 = NULL; *p1; ++p1)
439 				p2 = isspace(*p1) ? p1 : NULL;
440 			if (p2)
441 				pr->nospace = p2;
442 		}
443 	}
444 #ifdef DEBUG
445 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
446 		(void)printf("fmt:");
447 		for (pr = fu->nextpr; pr; pr = pr->nextpr)
448 			(void)printf(" {%s}", pr->fmt);
449 		(void)printf("\n");
450 	}
451 #endif
452 }
453 
454 void
455 escape(char *p1)
456 {
457 	char *p2;
458 
459 	/* alphabetic escape sequences have to be done in place */
460 	for (p2 = p1;; p1++, p2++) {
461 		if (*p1 == '\\') {
462 			p1++;
463 			switch(*p1) {
464 			case '\0':
465 				*p2 = '\\';
466 				*++p2 = '\0';
467 				return;
468 			case 'a':
469 			     /* *p2 = '\a'; */
470 				*p2 = '\007';
471 				break;
472 			case 'b':
473 				*p2 = '\b';
474 				break;
475 			case 'f':
476 				*p2 = '\f';
477 				break;
478 			case 'n':
479 				*p2 = '\n';
480 				break;
481 			case 'r':
482 				*p2 = '\r';
483 				break;
484 			case 't':
485 				*p2 = '\t';
486 				break;
487 			case 'v':
488 				*p2 = '\v';
489 				break;
490 			default:
491 				*p2 = *p1;
492 				break;
493 			}
494 		} else {
495 			*p2 = *p1;
496 			if (*p1 == '\0')
497 				return;
498 		}
499 	}
500 }
501 
502 void
503 badcnt(const char *s)
504 {
505 	errx(1, "%s: bad byte count", s);
506 }
507 
508 void
509 badsfmt(void)
510 {
511 	errx(1, "%%s: requires a precision or a byte count");
512 }
513 
514 void
515 badfmt(const char *fmt)
516 {
517 	errx(1, "\"%s\": bad format", fmt);
518 }
519 
520 void
521 badconv(const char *ch)
522 {
523 	errx(1, "%%%s: bad conversion character", ch);
524 }
525 
526 void
527 badnoconv(void)
528 {
529 	errx(1, "missing conversion character");
530 }
531