xref: /netbsd-src/usr.bin/hexdump/parse.c (revision 2a399c6883d870daece976daec6ffa7bb7f934ce)
1 /*	$NetBSD: parse.c,v 1.7 1997/10/19 02:34:10 lukem 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 #include <sys/cdefs.h>
37 #ifndef lint
38 #if 0
39 static char sccsid[] = "@(#)parse.c	8.1 (Berkeley) 6/6/93";
40 #else
41 __RCSID("$NetBSD: parse.c,v 1.7 1997/10/19 02:34:10 lukem 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(name)
62 	char *name;
63 {
64 	char *p;
65 	FILE *fp;
66 	int ch;
67 	char buf[2048 + 1];
68 
69 	if ((fp = fopen(name, "r")) == NULL)
70 		err(1, "fopen %s", name);
71 	while (fgets(buf, sizeof(buf), fp)) {
72 		if (!(p = strchr(buf, '\n'))) {
73 			warnx("line too long.");
74 			while ((ch = getchar()) != '\n' && ch != EOF);
75 			continue;
76 		}
77 		*p = '\0';
78 		for (p = buf; *p && isspace(*p); ++p);
79 		if (!*p || *p == '#')
80 			continue;
81 		add(p);
82 	}
83 	(void)fclose(fp);
84 }
85 
86 void
87 add(fmt)
88 	char *fmt;
89 {
90 	char *p;
91 	static FS **nextfs;
92 	FS *tfs;
93 	FU *tfu, **nextfu;
94 	char *savep;
95 
96 	/* start new linked list of format units */
97 	tfs = emalloc(sizeof(FS));
98 	if (!fshead)
99 		fshead = tfs;
100 	else
101 		*nextfs = tfs;
102 	nextfs = &tfs->nextfs;
103 	nextfu = &tfs->nextfu;
104 
105 	/* take the format string and break it up into format units */
106 	for (p = fmt;;) {
107 		/* skip leading white space */
108 		for (; isspace(*p); ++p);
109 		if (!*p)
110 			break;
111 
112 		/* allocate a new format unit and link it in */
113 		tfu = emalloc(sizeof(FU));
114 		*nextfu = tfu;
115 		nextfu = &tfu->nextfu;
116 		tfu->reps = 1;
117 
118 		/* if leading digit, repetition count */
119 		if (isdigit(*p)) {
120 			for (savep = p; isdigit(*p); ++p);
121 			if (!isspace(*p) && *p != '/')
122 				badfmt(fmt);
123 			/* may overwrite either white space or slash */
124 			tfu->reps = atoi(savep);
125 			tfu->flags = F_SETREP;
126 			/* skip trailing white space */
127 			for (++p; isspace(*p); ++p);
128 		}
129 
130 		/* skip slash and trailing white space */
131 		if (*p == '/')
132 			while (isspace(*++p));
133 
134 		/* byte count */
135 		if (isdigit(*p)) {
136 			for (savep = p; isdigit(*p); ++p);
137 			if (!isspace(*p))
138 				badfmt(fmt);
139 			tfu->bcnt = atoi(savep);
140 			/* skip trailing white space */
141 			for (++p; isspace(*p); ++p);
142 		}
143 
144 		/* format */
145 		if (*p != '"')
146 			badfmt(fmt);
147 		for (savep = ++p; *p != '"';)
148 			if (*p++ == 0)
149 				badfmt(fmt);
150 		if (!(tfu->fmt = malloc(p - savep + 1)))
151 			nomem();
152 		(void) strncpy(tfu->fmt, savep, p - savep);
153 		tfu->fmt[p - savep] = '\0';
154 		escape(tfu->fmt);
155 		p++;
156 	}
157 }
158 
159 static const char *spec = ".#-+ 0123456789";
160 
161 int
162 size(fs)
163 	FS *fs;
164 {
165 	FU *fu;
166 	int bcnt, cursize;
167 	char *fmt;
168 	int prec;
169 
170 	/* figure out the data block size needed for each format unit */
171 	for (cursize = 0, fu = fs->nextfu; fu; fu = fu->nextfu) {
172 		if (fu->bcnt) {
173 			cursize += fu->bcnt * fu->reps;
174 			continue;
175 		}
176 		for (bcnt = prec = 0, fmt = fu->fmt; *fmt; ++fmt) {
177 			if (*fmt != '%')
178 				continue;
179 			/*
180 			 * skip any special chars -- save precision in
181 			 * case it's a %s format.
182 			 */
183 			while (strchr(spec + 1, *++fmt));
184 			if (*fmt == '.' && isdigit(*++fmt)) {
185 				prec = atoi(fmt);
186 				while (isdigit(*++fmt));
187 			}
188 			switch(*fmt) {
189 			case 'c':
190 				bcnt += 1;
191 				break;
192 			case 'd': case 'i': case 'o': case 'u':
193 			case 'x': case 'X':
194 				bcnt += 4;
195 				break;
196 			case 'e': case 'E': case 'f': case 'g': case 'G':
197 				bcnt += 8;
198 				break;
199 			case 's':
200 				bcnt += prec;
201 				break;
202 			case '_':
203 				switch(*++fmt) {
204 				case 'c': case 'p': case 'u':
205 					bcnt += 1;
206 					break;
207 				}
208 			}
209 		}
210 		cursize += bcnt * fu->reps;
211 	}
212 	return (cursize);
213 }
214 
215 void
216 rewrite(fs)
217 	FS *fs;
218 {
219 	enum { NOTOKAY, USEBCNT, USEPREC } sokay;
220 	PR *pr, **nextpr;
221 	FU *fu;
222 	char *p1, *p2;
223 	char savech, *fmtp, cs[3];
224 	int nconv, prec;
225 
226 	nextpr = NULL;
227 	prec = 0;
228 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
229 		/*
230 		 * Break each format unit into print units; each conversion
231 		 * character gets its own.
232 		 */
233 		for (nconv = 0, fmtp = fu->fmt; *fmtp; nextpr = &pr->nextpr) {
234 			pr = emalloc(sizeof(PR));
235 			if (!fu->nextpr)
236 				fu->nextpr = pr;
237 			else
238 				*nextpr = pr;
239 
240 			/* Skip preceding text and up to the next % sign. */
241 			for (p1 = fmtp; *p1 && *p1 != '%'; ++p1);
242 
243 			/* Only text in the string. */
244 			if (!*p1) {
245 				pr->fmt = fmtp;
246 				pr->flags = F_TEXT;
247 				break;
248 			}
249 
250 			/*
251 			 * Get precision for %s -- if have a byte count, don't
252 			 * need it.
253 			 */
254 			if (fu->bcnt) {
255 				sokay = USEBCNT;
256 				/* Skip to conversion character. */
257 				for (++p1; strchr(spec, *p1); ++p1);
258 			} else {
259 				/* Skip any special chars, field width. */
260 				while (strchr(spec + 1, *++p1));
261 				if (*p1 == '.' && isdigit(*++p1)) {
262 					sokay = USEPREC;
263 					prec = atoi(p1);
264 					while (isdigit(*++p1));
265 				} else
266 					sokay = NOTOKAY;
267 			}
268 
269 			p2 = p1 + 1;		/* Set end pointer. */
270 			cs[0] = *p1;		/* Set conversion string. */
271 			cs[1] = '\0';
272 
273 			/*
274 			 * Figure out the byte count for each conversion;
275 			 * rewrite the format as necessary, set up blank-
276 			 * padding for end of data.
277 			 */
278 			switch(cs[0]) {
279 			case 'c':
280 				pr->flags = F_CHAR;
281 				switch(fu->bcnt) {
282 				case 0: case 1:
283 					pr->bcnt = 1;
284 					break;
285 				default:
286 					p1[1] = '\0';
287 					badcnt(p1);
288 				}
289 				break;
290 			case 'd': case 'i':
291 				pr->flags = F_INT;
292 				goto isint;
293 			case 'o': case 'u': case 'x': case 'X':
294 				pr->flags = F_UINT;
295 isint:				cs[2] = '\0';
296 				cs[1] = cs[0];
297 				cs[0] = 'q';
298 				switch(fu->bcnt) {
299 				case 0: case 4:
300 					pr->bcnt = 4;
301 					break;
302 				case 1:
303 					pr->bcnt = 1;
304 					break;
305 				case 2:
306 					pr->bcnt = 2;
307 					break;
308 				default:
309 					p1[1] = '\0';
310 					badcnt(p1);
311 				}
312 				break;
313 			case 'e': case 'E': case 'f': case 'g': case 'G':
314 				pr->flags = F_DBL;
315 				switch(fu->bcnt) {
316 				case 0: case 8:
317 					pr->bcnt = 8;
318 					break;
319 				case 4:
320 					pr->bcnt = 4;
321 					break;
322 				default:
323 					p1[1] = '\0';
324 					badcnt(p1);
325 				}
326 				break;
327 			case 's':
328 				pr->flags = F_STR;
329 				switch(sokay) {
330 				case NOTOKAY:
331 					badsfmt();
332 				case USEBCNT:
333 					pr->bcnt = fu->bcnt;
334 					break;
335 				case USEPREC:
336 					pr->bcnt = prec;
337 					break;
338 				}
339 				break;
340 			case '_':
341 				++p2;
342 				switch(p1[1]) {
343 				case 'A':
344 					endfu = fu;
345 					fu->flags |= F_IGNORE;
346 					/* FALLTHROUGH */
347 				case 'a':
348 					pr->flags = F_ADDRESS;
349 					++p2;
350 					switch(p1[2]) {
351 					case 'd': case 'o': case'x':
352 						cs[0] = 'q';
353 						cs[1] = p1[2];
354 						cs[2] = '\0';
355 						break;
356 					default:
357 						p1[3] = '\0';
358 						badconv(p1);
359 					}
360 					break;
361 				case 'c':
362 					pr->flags = F_C;
363 					/* cs[0] = 'c';	set in conv_c */
364 					goto isint2;
365 				case 'p':
366 					pr->flags = F_P;
367 					cs[0] = 'c';
368 					goto isint2;
369 				case 'u':
370 					pr->flags = F_U;
371 					/* cs[0] = 'c';	set in conv_u */
372 isint2:					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[0] = '\0';
397 			pr->fmt = emalloc(strlen(fmtp) + 2);
398 			(void)strcpy(pr->fmt, fmtp);
399 			(void)strcat(pr->fmt, cs);
400 			*p2 = savech;
401 			pr->cchar = pr->fmt + (p1 - fmtp);
402 			fmtp = p2;
403 
404 			/* Only one conversion character if byte count. */
405 			if (!(pr->flags&F_ADDRESS) && fu->bcnt && nconv++)
406 				errx(1,
407 			    "byte count with multiple conversion characters");
408 		}
409 		/*
410 		 * If format unit byte count not specified, figure it out
411 		 * so can adjust rep count later.
412 		 */
413 		if (!fu->bcnt)
414 			for (pr = fu->nextpr; pr; pr = pr->nextpr)
415 				fu->bcnt += pr->bcnt;
416 	}
417 	/*
418 	 * If the format string interprets any data at all, and it's
419 	 * not the same as the blocksize, and its last format unit
420 	 * interprets any data at all, and has no iteration count,
421 	 * repeat it as necessary.
422 	 *
423 	 * If, rep count is greater than 1, no trailing whitespace
424 	 * gets output from the last iteration of the format unit.
425 	 */
426 	for (fu = fs->nextfu;; fu = fu->nextfu) {
427 		if (!fu->nextfu && fs->bcnt < blocksize &&
428 		    !(fu->flags&F_SETREP) && fu->bcnt)
429 			fu->reps += (blocksize - fs->bcnt) / fu->bcnt;
430 		if (fu->reps > 1) {
431 			for (pr = fu->nextpr;; pr = pr->nextpr)
432 				if (!pr->nextpr)
433 					break;
434 			for (p1 = pr->fmt, p2 = NULL; *p1; ++p1)
435 				p2 = isspace(*p1) ? p1 : NULL;
436 			if (p2)
437 				pr->nospace = p2;
438 		}
439 		if (!fu->nextfu)
440 			break;
441 	}
442 #ifdef DEBUG
443 	for (fu = fs->nextfu; fu; fu = fu->nextfu) {
444 		(void)printf("fmt:");
445 		for (pr = fu->nextpr; pr; pr = pr->nextpr)
446 			(void)printf(" {%s}", pr->fmt);
447 		(void)printf("\n");
448 	}
449 #endif
450 }
451 
452 void
453 escape(p1)
454 	char *p1;
455 {
456 	char *p2;
457 
458 	/* alphabetic escape sequences have to be done in place */
459 	for (p2 = p1;; ++p1, ++p2) {
460 		if (!*p1) {
461 			*p2 = *p1;
462 			break;
463 		}
464 		if (*p1 == '\\')
465 			switch(*++p1) {
466 			case 'a':
467 			     /* *p2 = '\a'; */
468 				*p2 = '\007';
469 				break;
470 			case 'b':
471 				*p2 = '\b';
472 				break;
473 			case 'f':
474 				*p2 = '\f';
475 				break;
476 			case 'n':
477 				*p2 = '\n';
478 				break;
479 			case 'r':
480 				*p2 = '\r';
481 				break;
482 			case 't':
483 				*p2 = '\t';
484 				break;
485 			case 'v':
486 				*p2 = '\v';
487 				break;
488 			default:
489 				*p2 = *p1;
490 				break;
491 			}
492 	}
493 }
494 
495 void
496 badcnt(s)
497 	char *s;
498 {
499 	errx(1, "%s: bad byte count", s);
500 }
501 
502 void
503 badsfmt()
504 {
505 	errx(1, "%%s: requires a precision or a byte count\n");
506 }
507 
508 void
509 badfmt(fmt)
510 	char *fmt;
511 {
512 	errx(1, "\"%s\": bad format\n", fmt);
513 }
514 
515 void
516 badconv(ch)
517 	char *ch;
518 {
519 	errx(1, "%%%s: bad conversion character\n", ch);
520 }
521