xref: /netbsd-src/usr.bin/hexdump/parse.c (revision 27527e67bbdf8d9ec84fd58803048ed6d181ece2)
1 /*	$NetBSD: parse.c,v 1.18 2006/01/04 01:30:21 perry 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. 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 #if HAVE_NBTOOL_CONFIG_H
33 #include "nbtool_config.h"
34 #endif
35 
36 #include <sys/cdefs.h>
37 #if !defined(lint)
38 #if 0
39 static char sccsid[] = "@(#)parse.c	8.1 (Berkeley) 6/6/93";
40 #else
41 __RCSID("$NetBSD: parse.c,v 1.18 2006/01/04 01:30:21 perry Exp $");
42 #endif
43 #endif /* not lint */
44 
45 #include <sys/types.h>
46 #include <sys/file.h>
47 
48 #include <ctype.h>
49 #include <err.h>
50 #include <errno.h>
51 #include <fcntl.h>
52 #include <stdio.h>
53 #include <stdlib.h>
54 #include <string.h>
55 
56 #include "hexdump.h"
57 
58 FU *endfu;					/* format at end-of-data */
59 
60 void
61 addfile(char *name)
62 {
63 	char *p;
64 	FILE *fp;
65 	int ch;
66 	char buf[2048 + 1];
67 
68 	if ((fp = fopen(name, "r")) == NULL)
69 		err(1, "fopen %s", name);
70 	while (fgets(buf, sizeof(buf), fp)) {
71 		if (!(p = strchr(buf, '\n'))) {
72 			warnx("line too long.");
73 			while ((ch = getchar()) != '\n' && ch != EOF);
74 			continue;
75 		}
76 		*p = '\0';
77 		for (p = buf; *p && isspace((unsigned char)*p); ++p);
78 		if (!*p || *p == '#')
79 			continue;
80 		add(p);
81 	}
82 	(void)fclose(fp);
83 }
84 
85 void
86 add(const char *fmt)
87 {
88 	const char *p;
89 	static FS **nextfs;
90 	FS *tfs;
91 	FU *tfu, **nextfu;
92 	const 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((unsigned char)*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((unsigned char)*p)) {
118 			for (savep = p; isdigit((unsigned char)*p); ++p);
119 			if (!isspace((unsigned char)*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((unsigned char)*p); ++p);
126 		}
127 
128 		/* skip slash and trailing white space */
129 		if (*p == '/')
130 			while (isspace((unsigned char)*++p));
131 
132 		/* byte count */
133 		if (isdigit((unsigned char)*p)) {
134 			for (savep = p; isdigit((unsigned char)*p); ++p);
135 			if (!isspace((unsigned char)*p))
136 				badfmt(fmt);
137 			tfu->bcnt = atoi(savep);
138 			/* skip trailing white space */
139 			for (++p; isspace((unsigned char)*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 *fs)
161 {
162 	FU *fu;
163 	int bcnt, cursize;
164 	char *fmt;
165 	int prec;
166 
167 	/* figure out the data block size needed for each format unit */
168 	for (cursize = 0, fu = fs->nextfu; fu; fu = fu->nextfu) {
169 		if (fu->bcnt) {
170 			cursize += fu->bcnt * fu->reps;
171 			continue;
172 		}
173 		for (bcnt = prec = 0, fmt = fu->fmt; *fmt; ++fmt) {
174 			if (*fmt != '%')
175 				continue;
176 			/*
177 			 * skip any special chars -- save precision in
178 			 * case it's a %s format.
179 			 */
180 			while (strchr(spec + 1, *++fmt));
181 			if (*fmt == '.' && isdigit((unsigned char)*++fmt)) {
182 				prec = atoi(fmt);
183 				while (isdigit((unsigned char)*++fmt));
184 			}
185 			switch(*fmt) {
186 			case 'c':
187 				bcnt += 1;
188 				break;
189 			case 'd': case 'i': case 'o': case 'u':
190 			case 'x': case 'X':
191 				bcnt += 4;
192 				break;
193 			case 'e': case 'E': case 'f': case 'g': case 'G':
194 				bcnt += 8;
195 				break;
196 			case 's':
197 				bcnt += prec;
198 				break;
199 			case '_':
200 				switch(*++fmt) {
201 				case 'c': case 'p': case 'u':
202 					bcnt += 1;
203 					break;
204 				}
205 			}
206 		}
207 		cursize += bcnt * fu->reps;
208 	}
209 	return (cursize);
210 }
211 
212 void
213 rewrite(FS *fs)
214 {
215 	enum { NOTOKAY, USEBCNT, USEPREC } sokay;
216 	PR *pr, **nextpr;
217 	FU *fu;
218 	char *p1, *p2;
219 	char savech, *fmtp, cs[3];
220 	int nconv, prec;
221 
222 	nextpr = NULL;
223 	prec = 0;
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 == '.' &&
258 				    isdigit((unsigned char)*++p1)) {
259 					sokay = USEPREC;
260 					prec = atoi(p1);
261 					while (isdigit((unsigned char)*++p1))
262 						continue;
263 				} else
264 					sokay = NOTOKAY;
265 			}
266 
267 			p2 = p1 + 1;		/* Set end pointer. */
268 			cs[0] = *p1;		/* Set conversion string. */
269 			cs[1] = '\0';
270 
271 			/*
272 			 * Figure out the byte count for each conversion;
273 			 * rewrite the format as necessary, set up blank-
274 			 * padding for end of data.
275 			 */
276 			switch(cs[0]) {
277 			case 'c':
278 				pr->flags = F_CHAR;
279 				switch(fu->bcnt) {
280 				case 0: case 1:
281 					pr->bcnt = 1;
282 					break;
283 				default:
284 					p1[1] = '\0';
285 					badcnt(p1);
286 				}
287 				break;
288 			case 'd': case 'i':
289 				pr->flags = F_INT;
290 				goto isint;
291 			case 'o': case 'u': case 'x': case 'X':
292 				pr->flags = F_UINT;
293 isint:				cs[2] = '\0';
294 				cs[1] = cs[0];
295 				cs[0] = 'q';
296 				switch(fu->bcnt) {
297 				case 0: case 4:
298 					pr->bcnt = 4;
299 					break;
300 				case 1:
301 					pr->bcnt = 1;
302 					break;
303 				case 2:
304 					pr->bcnt = 2;
305 					break;
306 				case 8:
307 					pr->bcnt = 8;
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 						cs[0] = 'q';
354 						cs[1] = p1[2];
355 						cs[2] = '\0';
356 						break;
357 					default:
358 						p1[3] = '\0';
359 						badconv(p1);
360 					}
361 					break;
362 				case 'c':
363 					pr->flags = F_C;
364 					/* cs[0] = 'c';	set in conv_c */
365 					goto isint2;
366 				case 'p':
367 					pr->flags = F_P;
368 					cs[0] = 'c';
369 					goto isint2;
370 				case 'u':
371 					pr->flags = F_U;
372 					/* cs[0] = 'c';	set in conv_u */
373 isint2:					switch(fu->bcnt) {
374 					case 0: case 1:
375 						pr->bcnt = 1;
376 						break;
377 					default:
378 						p1[2] = '\0';
379 						badcnt(p1);
380 					}
381 					break;
382 				default:
383 					p1[2] = '\0';
384 					badconv(p1);
385 				}
386 				break;
387 			default:
388 				p1[1] = '\0';
389 				badconv(p1);
390 			}
391 
392 			/*
393 			 * Copy to PR format string, set conversion character
394 			 * pointer, update original.
395 			 */
396 			savech = *p2;
397 			p1[0] = '\0';
398 			pr->fmt = emalloc(strlen(fmtp) + strlen(cs) + 1);
399 			(void)strcpy(pr->fmt, fmtp);
400 			(void)strcat(pr->fmt, cs);
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 				errx(1,
408 			    "byte count with multiple conversion characters");
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 = 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((unsigned char)*p1) ? p1 : NULL;
437 			if (p2)
438 				pr->nospace = p2;
439 		}
440 	}
441 #ifdef DEBUG
442 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
443 		(void)printf("fmt:");
444 		for (pr = fu->nextpr; pr; pr = pr->nextpr)
445 			(void)printf(" {%s}", pr->fmt);
446 		(void)printf("\n");
447 	}
448 #endif
449 }
450 
451 void
452 escape(char *p1)
453 {
454 	char *p2;
455 
456 	/* alphabetic escape sequences have to be done in place */
457 	for (p2 = p1;; ++p1, ++p2) {
458 		if (!*p1) {
459 			*p2 = *p1;
460 			break;
461 		}
462 		if (*p1 == '\\')
463 			switch(*++p1) {
464 			case 'a':
465 			     /* *p2 = '\a'; */
466 				*p2 = '\007';
467 				break;
468 			case 'b':
469 				*p2 = '\b';
470 				break;
471 			case 'f':
472 				*p2 = '\f';
473 				break;
474 			case 'n':
475 				*p2 = '\n';
476 				break;
477 			case 'r':
478 				*p2 = '\r';
479 				break;
480 			case 't':
481 				*p2 = '\t';
482 				break;
483 			case 'v':
484 				*p2 = '\v';
485 				break;
486 			default:
487 				*p2 = *p1;
488 				break;
489 			}
490 	}
491 }
492 
493 void
494 badcnt(char *s)
495 {
496 	errx(1, "%s: bad byte count", s);
497 }
498 
499 void
500 badsfmt(void)
501 {
502 	errx(1, "%%s: requires a precision or a byte count");
503 }
504 
505 void
506 badfmt(const char *fmt)
507 {
508 	errx(1, "\"%s\": bad format", fmt);
509 }
510 
511 void
512 badconv(char *ch)
513 {
514 	errx(1, "%%%s: bad conversion character", ch);
515 }
516