xref: /freebsd/usr.bin/printf/printf.c (revision ceaec73d406831b1251babb61675df0a1aa54a31)
1 /*
2  * Copyright (c) 1989, 1993
3  *	The Regents of the University of California.  All rights reserved.
4  *
5  * Redistribution and use in source and binary forms, with or without
6  * modification, are permitted provided that the following conditions
7  * are met:
8  * 1. Redistributions of source code must retain the above copyright
9  *    notice, this list of conditions and the following disclaimer.
10  * 2. Redistributions in binary form must reproduce the above copyright
11  *    notice, this list of conditions and the following disclaimer in the
12  *    documentation and/or other materials provided with the distribution.
13  * 3. All advertising materials mentioning features or use of this software
14  *    must display the following acknowledgement:
15  *	This product includes software developed by the University of
16  *	California, Berkeley and its contributors.
17  * 4. Neither the name of the University nor the names of its contributors
18  *    may be used to endorse or promote products derived from this software
19  *    without specific prior written permission.
20  *
21  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31  * SUCH DAMAGE.
32  */
33 
34 #if !defined(BUILTIN) && !defined(SHELL)
35 #ifndef lint
36 static char const copyright[] =
37 "@(#) Copyright (c) 1989, 1993\n\
38 	The Regents of the University of California.  All rights reserved.\n";
39 #endif /* not lint */
40 #endif
41 
42 #ifndef lint
43 #if 0
44 static char const sccsid[] = "@(#)printf.c	8.1 (Berkeley) 7/20/93";
45 #endif
46 static const char rcsid[] =
47   "$FreeBSD$";
48 #endif /* not lint */
49 
50 #include <sys/types.h>
51 
52 #include <err.h>
53 #include <errno.h>
54 #include <limits.h>
55 #include <stdio.h>
56 #include <stdlib.h>
57 #include <string.h>
58 #include <unistd.h>
59 
60 #ifdef SHELL
61 #define main printfcmd
62 #include "bltin/bltin.h"
63 #include "memalloc.h"
64 #else
65 #define	warnx1(a, b, c)		warnx(a)
66 #define	warnx2(a, b, c)		warnx(a, b)
67 #define	warnx3(a, b, c)		warnx(a, b, c)
68 #endif
69 
70 #ifndef BUILTIN
71 #include <locale.h>
72 #endif
73 
74 #define PF(f, func) do { \
75 	char *b = NULL; \
76 	if (havewidth) \
77 		if (haveprec) \
78 			(void)asprintf(&b, f, fieldwidth, precision, func); \
79 		else \
80 			(void)asprintf(&b, f, fieldwidth, func); \
81 	else if (haveprec) \
82 		(void)asprintf(&b, f, precision, func); \
83 	else \
84 		(void)asprintf(&b, f, func); \
85 	if (b) { \
86 		(void)fputs(b, stdout); \
87 		free(b); \
88 	} \
89 } while (0)
90 
91 static int	 asciicode(void);
92 static char	*doformat(char *, int *);
93 static int	 escape(char *, int, size_t *);
94 static int	 getchr(void);
95 static int	 getfloating(long double *, int);
96 static int	 getint(int *);
97 static int	 getquads(quad_t *, u_quad_t *, int);
98 static const char
99 		*getstr(void);
100 static char	*mkquad(char *, int);
101 static void	 usage(void);
102 
103 static char **gargv;
104 
105 int
106 #ifdef BUILTIN
107 progprintf(int argc, char *argv[])
108 #else
109 main(int argc, char *argv[])
110 #endif
111 {
112 	size_t len;
113 	int ch, chopped, end, rval;
114 	char *format, *fmt, *start;
115 
116 #ifndef BUILTIN
117 	(void) setlocale(LC_NUMERIC, "");
118 #endif
119 	while ((ch = getopt(argc, argv, "")) != -1)
120 		switch (ch) {
121 		case '?':
122 		default:
123 			usage();
124 			return (1);
125 		}
126 	argc -= optind;
127 	argv += optind;
128 
129 	if (argc < 1) {
130 		usage();
131 		return (1);
132 	}
133 
134 	/*
135 	 * Basic algorithm is to scan the format string for conversion
136 	 * specifications -- once one is found, find out if the field
137 	 * width or precision is a '*'; if it is, gather up value.  Note,
138 	 * format strings are reused as necessary to use up the provided
139 	 * arguments, arguments of zero/null string are provided to use
140 	 * up the format string.
141 	 */
142 	fmt = format = *argv;
143 	chopped = escape(fmt, 1, &len);		/* backslash interpretation */
144 	rval = end = 0;
145 	gargv = ++argv;
146 	for (;;) {
147 		start = fmt;
148 		while (fmt < format + len) {
149 			if (fmt[0] == '%') {
150 				fwrite(start, 1, fmt - start, stdout);
151 				if (fmt[1] == '%') {
152 					/* %% prints a % */
153 					putchar('%');
154 					fmt += 2;
155 				} else {
156 					fmt = doformat(fmt, &rval);
157 					if (fmt == NULL)
158 						return (1);
159 					end = 0;
160 				}
161 				start = fmt;
162 			} else
163 				fmt++;
164 		}
165 
166 		if (end == 1) {
167 			warnx1("missing format character", NULL, NULL);
168 			return (1);
169 		}
170 		fwrite(start, 1, fmt - start, stdout);
171 		if (chopped || !*gargv)
172 			return (rval);
173 		/* Restart at the beginning of the format string. */
174 		fmt = format;
175 		end = 1;
176 	}
177 	/* NOTREACHED */
178 }
179 
180 
181 static char *
182 doformat(char *start, int *rval)
183 {
184 	static const char skip1[] = "#'-+ 0";
185 	static const char skip2[] = "0123456789";
186 	char *fmt;
187 	int fieldwidth, haveprec, havewidth, mod_ldbl, precision;
188 	char convch, nextch;
189 
190 	fmt = start + 1;
191 	/* skip to field width */
192 	fmt += strspn(fmt, skip1);
193 	if (*fmt == '*') {
194 		if (getint(&fieldwidth))
195 			return (NULL);
196 		havewidth = 1;
197 		++fmt;
198 	} else {
199 		havewidth = 0;
200 
201 		/* skip to possible '.', get following precision */
202 		fmt += strspn(fmt, skip2);
203 	}
204 	if (*fmt == '.') {
205 		/* precision present? */
206 		++fmt;
207 		if (*fmt == '*') {
208 			if (getint(&precision))
209 				return (NULL);
210 			haveprec = 1;
211 			++fmt;
212 		} else {
213 			haveprec = 0;
214 
215 			/* skip to conversion char */
216 			fmt += strspn(fmt, skip2);
217 		}
218 	} else
219 		haveprec = 0;
220 	if (!*fmt) {
221 		warnx1("missing format character", NULL, NULL);
222 		return (NULL);
223 	}
224 
225 	/*
226 	 * Look for a length modifier.  POSIX doesn't have these, so
227 	 * we only support them for floating-point conversions, which
228 	 * are extensions.  This is useful because the L modifier can
229 	 * be used to gain extra range and precision, while omitting
230 	 * it is more likely to produce consistent results on different
231 	 * architectures.  This is not so important for integers
232 	 * because overflow is the only bad thing that can happen to
233 	 * them, but consider the command  printf %a 1.1
234 	 */
235 	if (*fmt == 'L') {
236 		mod_ldbl = 1;
237 		fmt++;
238 		if (!strchr("aAeEfFgG", *fmt)) {
239 			warnx2("bad modifier L for %%%c", *fmt, NULL);
240 			return (NULL);
241 		}
242 	} else {
243 		mod_ldbl = 0;
244 	}
245 
246 	convch = *fmt;
247 	nextch = *++fmt;
248 	*fmt = '\0';
249 	switch (convch) {
250 	case 'b': {
251 		size_t len;
252 		char *p;
253 		int getout;
254 
255 #ifdef SHELL
256 		p = savestr(getstr());
257 #else
258 		p = strdup(getstr());
259 #endif
260 		if (p == NULL) {
261 			warnx2("%s", strerror(ENOMEM), NULL);
262 			return (NULL);
263 		}
264 		getout = escape(p, 0, &len);
265 		*(fmt - 1) = 's';
266 		PF(start, p);
267 		*(fmt - 1) = 'b';
268 #ifdef SHELL
269 		ckfree(p);
270 #else
271 		free(p);
272 #endif
273 		if (getout)
274 			return (fmt);
275 		break;
276 	}
277 	case 'c': {
278 		char p;
279 
280 		p = getchr();
281 		PF(start, p);
282 		break;
283 	}
284 	case 's': {
285 		const char *p;
286 
287 		p = getstr();
288 		PF(start, p);
289 		break;
290 	}
291 	case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': {
292 		char *f;
293 		quad_t val;
294 		u_quad_t uval;
295 		int signedconv;
296 
297 		signedconv = (convch == 'd' || convch == 'i');
298 		if ((f = mkquad(start, convch)) == NULL)
299 			return (NULL);
300 		if (getquads(&val, &uval, signedconv))
301 			*rval = 1;
302 		if (signedconv)
303 			PF(f, val);
304 		else
305 			PF(f, uval);
306 		break;
307 	}
308 	case 'e': case 'E':
309 	case 'f': case 'F':
310 	case 'g': case 'G':
311 	case 'a': case 'A': {
312 		long double p;
313 
314 		if (getfloating(&p, mod_ldbl))
315 			*rval = 1;
316 		if (mod_ldbl)
317 			PF(start, p);
318 		else
319 			PF(start, (double)p);
320 		break;
321 	}
322 	default:
323 		warnx2("illegal format character %c", convch, NULL);
324 		return (NULL);
325 	}
326 	*fmt = nextch;
327 	return (fmt);
328 }
329 
330 static char *
331 mkquad(char *str, int ch)
332 {
333 	static char *copy;
334 	static size_t copy_size;
335 	char *newcopy;
336 	size_t len, newlen;
337 
338 	len = strlen(str) + 2;
339 	if (len > copy_size) {
340 		newlen = ((len + 1023) >> 10) << 10;
341 #ifdef SHELL
342 		if ((newcopy = ckrealloc(copy, newlen)) == NULL)
343 #else
344 		if ((newcopy = realloc(copy, newlen)) == NULL)
345 #endif
346 		{
347 			warnx2("%s", strerror(ENOMEM), NULL);
348 			return (NULL);
349 		}
350 		copy = newcopy;
351 		copy_size = newlen;
352 	}
353 
354 	memmove(copy, str, len - 3);
355 	copy[len - 3] = 'q';
356 	copy[len - 2] = ch;
357 	copy[len - 1] = '\0';
358 	return (copy);
359 }
360 
361 static int
362 escape(char *fmt, int percent, size_t *len)
363 {
364 	char *save, *store;
365 	int value, c;
366 
367 	for (save = store = fmt; (c = *fmt); ++fmt, ++store) {
368 		if (c != '\\') {
369 			*store = c;
370 			continue;
371 		}
372 		switch (*++fmt) {
373 		case '\0':		/* EOS, user error */
374 			*store = '\\';
375 			*++store = '\0';
376 			*len = store - save;
377 			return (0);
378 		case '\\':		/* backslash */
379 		case '\'':		/* single quote */
380 			*store = *fmt;
381 			break;
382 		case 'a':		/* bell/alert */
383 			*store = '\a';
384 			break;
385 		case 'b':		/* backspace */
386 			*store = '\b';
387 			break;
388 		case 'c':
389 			*store = '\0';
390 			*len = store - save;
391 			return (1);
392 		case 'f':		/* form-feed */
393 			*store = '\f';
394 			break;
395 		case 'n':		/* newline */
396 			*store = '\n';
397 			break;
398 		case 'r':		/* carriage-return */
399 			*store = '\r';
400 			break;
401 		case 't':		/* horizontal tab */
402 			*store = '\t';
403 			break;
404 		case 'v':		/* vertical tab */
405 			*store = '\v';
406 			break;
407 					/* octal constant */
408 		case '0': case '1': case '2': case '3':
409 		case '4': case '5': case '6': case '7':
410 			for (c = *fmt == '0' ? 4 : 3, value = 0;
411 			    c-- && *fmt >= '0' && *fmt <= '7'; ++fmt) {
412 				value <<= 3;
413 				value += *fmt - '0';
414 			}
415 			--fmt;
416 			if (percent && value == '%') {
417 				*store++ = '%';
418 				*store = '%';
419 			} else
420 				*store = value;
421 			break;
422 		default:
423 			*store = *fmt;
424 			break;
425 		}
426 	}
427 	*store = '\0';
428 	*len = store - save;
429 	return (0);
430 }
431 
432 static int
433 getchr(void)
434 {
435 	if (!*gargv)
436 		return ('\0');
437 	return ((int)**gargv++);
438 }
439 
440 static const char *
441 getstr(void)
442 {
443 	if (!*gargv)
444 		return ("");
445 	return (*gargv++);
446 }
447 
448 static int
449 getint(int *ip)
450 {
451 	quad_t val;
452 	u_quad_t uval;
453 	int rval;
454 
455 	if (getquads(&val, &uval, 1))
456 		return (1);
457 	rval = 0;
458 	if (val < INT_MIN || val > INT_MAX) {
459 		warnx3("%s: %s", *gargv, strerror(ERANGE));
460 		rval = 1;
461 	}
462 	*ip = (int)val;
463 	return (rval);
464 }
465 
466 static int
467 getquads(quad_t *qp, u_quad_t *uqp, int signedconv)
468 {
469 	char *ep;
470 	int rval;
471 
472 	if (!*gargv) {
473 		*qp = 0;
474 		return (0);
475 	}
476 	if (**gargv == '"' || **gargv == '\'') {
477 		if (signedconv)
478 			*qp = asciicode();
479 		else
480 			*uqp = asciicode();
481 		return (0);
482 	}
483 	rval = 0;
484 	errno = 0;
485 	if (signedconv)
486 		*qp = strtoq(*gargv, &ep, 0);
487 	else
488 		*uqp = strtouq(*gargv, &ep, 0);
489 	if (ep == *gargv) {
490 		warnx2("%s: expected numeric value", *gargv, NULL);
491 		rval = 1;
492 	}
493 	else if (*ep != '\0') {
494 		warnx2("%s: not completely converted", *gargv, NULL);
495 		rval = 1;
496 	}
497 	if (errno == ERANGE) {
498 		warnx3("%s: %s", *gargv, strerror(ERANGE));
499 		rval = 1;
500 	}
501 	++gargv;
502 	return (rval);
503 }
504 
505 static int
506 getfloating(long double *dp, int mod_ldbl)
507 {
508 	char *ep;
509 	int rval;
510 
511 	if (!*gargv) {
512 		*dp = 0.0;
513 		return (0);
514 	}
515 	if (**gargv == '"' || **gargv == '\'') {
516 		*dp = asciicode();
517 		return (0);
518 	}
519 	rval = 0;
520 	errno = 0;
521 	if (mod_ldbl)
522 		*dp = strtold(*gargv, &ep);
523 	else
524 		*dp = strtod(*gargv, &ep);
525 	if (ep == *gargv) {
526 		warnx2("%s: expected numeric value", *gargv, NULL);
527 		rval = 1;
528 	} else if (*ep != '\0') {
529 		warnx2("%s: not completely converted", *gargv, NULL);
530 		rval = 1;
531 	}
532 	if (errno == ERANGE) {
533 		warnx3("%s: %s", *gargv, strerror(ERANGE));
534 		rval = 1;
535 	}
536 	++gargv;
537 	return (rval);
538 }
539 
540 static int
541 asciicode(void)
542 {
543 	int ch;
544 
545 	ch = **gargv;
546 	if (ch == '\'' || ch == '"')
547 		ch = (*gargv)[1];
548 	++gargv;
549 	return (ch);
550 }
551 
552 static void
553 usage(void)
554 {
555 	(void)fprintf(stderr, "usage: printf format [arg ...]\n");
556 }
557