xref: /dragonfly/usr.bin/printf/printf.c (revision 73610d44)
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. Neither the name of the University nor the names of its contributors
14  *    may be used to endorse or promote products derived from this software
15  *    without specific prior written permission.
16  *
17  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27  * SUCH DAMAGE.
28  *
29  * @(#) Copyright (c) 1989, 1993 The Regents of the University of California.  All rights reserved.
30  * @(#)printf.c	8.1 (Berkeley) 7/20/93
31  * $FreeBSD: src/usr.bin/printf/printf.c,v 1.50 2010/12/29 21:38:00 jilles Exp $
32  */
33 
34 #include <sys/types.h>
35 
36 #include <err.h>
37 #include <errno.h>
38 #include <inttypes.h>
39 #include <limits.h>
40 #include <locale.h>
41 #include <stdio.h>
42 #include <stdlib.h>
43 #include <string.h>
44 #include <unistd.h>
45 
46 #ifdef SHELL
47 #define main printfcmd
48 #include "bltin/bltin.h"
49 #include "error.h"
50 #endif
51 
52 #define PF(f, func) do { \
53 	char *b = NULL; \
54 	if (havewidth) \
55 		if (haveprec) \
56 			asprintf(&b, f, fieldwidth, precision, func); \
57 		else \
58 			asprintf(&b, f, fieldwidth, func); \
59 	else if (haveprec) \
60 		asprintf(&b, f, precision, func); \
61 	else \
62 		asprintf(&b, f, func); \
63 	if (b) { \
64 		fputs(b, stdout); \
65 		free(b); \
66 	} \
67 } while (0)
68 
69 static int	 asciicode(void);
70 static char	*printf_doformat(char *, int *);
71 static int	 escape(char *, int, size_t *);
72 static int	 getchr(void);
73 static int	 getfloating(long double *, int);
74 static int	 getint(int *);
75 static int	 getnum(intmax_t *, uintmax_t *, int);
76 static const char
77 		*getstr(void);
78 static char	*mknum(char *, char);
79 static void	 usage(void);
80 
81 static char **gargv;
82 
83 int
84 main(int argc, char **argv)
85 {
86 	size_t len;
87 	int chopped, end, rval;
88 	char *format, *fmt, *start;
89 
90 #ifndef SHELL
91 	setlocale(LC_ALL, "");
92 #endif
93 
94 	/*
95 	 * We may not use getopt(3) because calling
96 	 * "printf -f%s oo" may not result in an invalid
97 	 * option error.
98 	 * However common usage and other implementations seem
99 	 * to indicate that we need to allow -- as a discardable
100 	 * option separator.
101 	 */
102 	if (argc > 1 && strcmp(argv[1], "--") == 0) {
103 		argc--;
104 		argv++;
105 	}
106 
107 	if (argc < 2) {
108 		usage();
109 		return (1);
110 	}
111 
112 	argv++;
113 
114 #ifdef SHELL
115 	INTOFF;
116 #endif
117 	/*
118 	 * Basic algorithm is to scan the format string for conversion
119 	 * specifications -- once one is found, find out if the field
120 	 * width or precision is a '*'; if it is, gather up value.  Note,
121 	 * format strings are reused as necessary to use up the provided
122 	 * arguments, arguments of zero/null string are provided to use
123 	 * up the format string.
124 	 */
125 	fmt = format = *argv;
126 	chopped = escape(fmt, 1, &len);		/* backslash interpretation */
127 	rval = end = 0;
128 	gargv = ++argv;
129 	for (;;) {
130 		start = fmt;
131 		while (fmt < format + len) {
132 			if (fmt[0] == '%') {
133 				fwrite(start, 1, fmt - start, stdout);
134 				if (fmt[1] == '%') {
135 					/* %% prints a % */
136 					putchar('%');
137 					fmt += 2;
138 				} else {
139 					fmt = printf_doformat(fmt, &rval);
140 					if (fmt == NULL) {
141 #ifdef SHELL
142 						INTON;
143 #endif
144 						return (1);
145 					}
146 					end = 0;
147 				}
148 				start = fmt;
149 			} else
150 				fmt++;
151 		}
152 
153 		if (end == 1) {
154 			warnx("missing format character");
155 #ifdef SHELL
156 			INTON;
157 #endif
158 			return (1);
159 		}
160 		fwrite(start, 1, fmt - start, stdout);
161 		if (chopped || !*gargv) {
162 #ifdef SHELL
163 			INTON;
164 #endif
165 			return (rval);
166 		}
167 		/* Restart at the beginning of the format string. */
168 		fmt = format;
169 		end = 1;
170 	}
171 	/* NOTREACHED */
172 }
173 
174 
175 static char *
176 printf_doformat(char *start, int *rval)
177 {
178 	static const char skip1[] = "#'-+ 0";
179 	static const char skip2[] = "0123456789";
180 	char *fmt;
181 	int fieldwidth, haveprec, havewidth, mod_ldbl, precision;
182 	char convch, nextch;
183 
184 	fmt = start + 1;
185 	/* skip to field width */
186 	fmt += strspn(fmt, skip1);
187 	if (*fmt == '*') {
188 		if (getint(&fieldwidth))
189 			return (NULL);
190 		havewidth = 1;
191 		++fmt;
192 	} else {
193 		havewidth = 0;
194 
195 		/* skip to possible '.', get following precision */
196 		fmt += strspn(fmt, skip2);
197 	}
198 	if (*fmt == '.') {
199 		/* precision present? */
200 		++fmt;
201 		if (*fmt == '*') {
202 			if (getint(&precision))
203 				return (NULL);
204 			haveprec = 1;
205 			++fmt;
206 		} else {
207 			haveprec = 0;
208 
209 			/* skip to conversion char */
210 			fmt += strspn(fmt, skip2);
211 		}
212 	} else
213 		haveprec = 0;
214 	if (!*fmt) {
215 		warnx("missing format character");
216 		return (NULL);
217 	}
218 
219 	/*
220 	 * Look for a length modifier.  POSIX doesn't have these, so
221 	 * we only support them for floating-point conversions, which
222 	 * are extensions.  This is useful because the L modifier can
223 	 * be used to gain extra range and precision, while omitting
224 	 * it is more likely to produce consistent results on different
225 	 * architectures.  This is not so important for integers
226 	 * because overflow is the only bad thing that can happen to
227 	 * them, but consider the command  printf %a 1.1
228 	 */
229 	if (*fmt == 'L') {
230 		mod_ldbl = 1;
231 		fmt++;
232 		if (!strchr("aAeEfFgG", *fmt)) {
233 			warnx("bad modifier L for %%%c", *fmt);
234 			return (NULL);
235 		}
236 	} else {
237 		mod_ldbl = 0;
238 	}
239 
240 	convch = *fmt;
241 	nextch = *++fmt;
242 	*fmt = '\0';
243 	switch (convch) {
244 	case 'b': {
245 		size_t len;
246 		char *p;
247 		int getout;
248 
249 		p = strdup(getstr());
250 		if (p == NULL) {
251 			warnx("%s", strerror(ENOMEM));
252 			return (NULL);
253 		}
254 		getout = escape(p, 0, &len);
255 		*(fmt - 1) = 's';
256 		PF(start, p);
257 		*(fmt - 1) = 'b';
258 		free(p);
259 		if (getout)
260 			return (fmt);
261 		break;
262 	}
263 	case 'c': {
264 		char p;
265 
266 		p = getchr();
267 		PF(start, p);
268 		break;
269 	}
270 	case 's': {
271 		const char *p;
272 
273 		p = getstr();
274 		PF(start, p);
275 		break;
276 	}
277 	case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': {
278 		char *f;
279 		intmax_t val;
280 		uintmax_t uval;
281 		int signedconv;
282 
283 		signedconv = (convch == 'd' || convch == 'i');
284 		if ((f = mknum(start, convch)) == NULL)
285 			return (NULL);
286 		if (getnum(&val, &uval, signedconv))
287 			*rval = 1;
288 		if (signedconv)
289 			PF(f, val);
290 		else
291 			PF(f, uval);
292 		break;
293 	}
294 	case 'e': case 'E':
295 	case 'f': case 'F':
296 	case 'g': case 'G':
297 	case 'a': case 'A': {
298 		long double p;
299 
300 		if (getfloating(&p, mod_ldbl))
301 			*rval = 1;
302 		if (mod_ldbl)
303 			PF(start, p);
304 		else
305 			PF(start, (double)p);
306 		break;
307 	}
308 	default:
309 		warnx("illegal format character %c", convch);
310 		return (NULL);
311 	}
312 	*fmt = nextch;
313 	return (fmt);
314 }
315 
316 static char *
317 mknum(char *str, char ch)
318 {
319 	static char *copy;
320 	static size_t copy_size;
321 	char *newcopy;
322 	size_t len, newlen;
323 
324 	len = strlen(str) + 2;
325 	if (len > copy_size) {
326 		newlen = ((len + 1023) >> 10) << 10;
327 		if ((newcopy = realloc(copy, newlen)) == NULL)
328 		{
329 			warnx("%s", strerror(ENOMEM));
330 			return (NULL);
331 		}
332 		copy = newcopy;
333 		copy_size = newlen;
334 	}
335 
336 	memmove(copy, str, len - 3);
337 	copy[len - 3] = 'j';
338 	copy[len - 2] = ch;
339 	copy[len - 1] = '\0';
340 	return (copy);
341 }
342 
343 static int
344 escape(char *fmt, int percent, size_t *len)
345 {
346 	char *save, *store;
347 	int value, c;
348 
349 	for (save = store = fmt; (c = *fmt); ++fmt, ++store) {
350 		if (c != '\\') {
351 			*store = c;
352 			continue;
353 		}
354 		switch (*++fmt) {
355 		case '\0':		/* EOS, user error */
356 			*store = '\\';
357 			*++store = '\0';
358 			*len = store - save;
359 			return (0);
360 		case '\\':		/* backslash */
361 		case '\'':		/* single quote */
362 			*store = *fmt;
363 			break;
364 		case 'a':		/* bell/alert */
365 			*store = '\a';
366 			break;
367 		case 'b':		/* backspace */
368 			*store = '\b';
369 			break;
370 		case 'c':
371 			*store = '\0';
372 			*len = store - save;
373 			return (1);
374 		case 'f':		/* form-feed */
375 			*store = '\f';
376 			break;
377 		case 'n':		/* newline */
378 			*store = '\n';
379 			break;
380 		case 'r':		/* carriage-return */
381 			*store = '\r';
382 			break;
383 		case 't':		/* horizontal tab */
384 			*store = '\t';
385 			break;
386 		case 'v':		/* vertical tab */
387 			*store = '\v';
388 			break;
389 					/* octal constant */
390 		case '0': case '1': case '2': case '3':
391 		case '4': case '5': case '6': case '7':
392 			c = (!percent && *fmt == '0') ? 4 : 3;
393 			for (value = 0;
394 			    c-- && *fmt >= '0' && *fmt <= '7'; ++fmt) {
395 				value <<= 3;
396 				value += *fmt - '0';
397 			}
398 			--fmt;
399 			if (percent && value == '%') {
400 				*store++ = '%';
401 				*store = '%';
402 			} else
403 				*store = value;
404 			break;
405 		default:
406 			*store = *fmt;
407 			break;
408 		}
409 	}
410 	*store = '\0';
411 	*len = store - save;
412 	return (0);
413 }
414 
415 static int
416 getchr(void)
417 {
418 	if (!*gargv)
419 		return ('\0');
420 	return ((int)**gargv++);
421 }
422 
423 static const char *
424 getstr(void)
425 {
426 	if (!*gargv)
427 		return ("");
428 	return (*gargv++);
429 }
430 
431 static int
432 getint(int *ip)
433 {
434 	intmax_t val;
435 	uintmax_t uval;
436 	int rval;
437 
438 	if (getnum(&val, &uval, 1))
439 		return (1);
440 	rval = 0;
441 	if (val < INT_MIN || val > INT_MAX) {
442 		warnx("%s: %s", *gargv, strerror(ERANGE));
443 		rval = 1;
444 	}
445 	*ip = (int)val;
446 	return (rval);
447 }
448 
449 static int
450 getnum(intmax_t *ip, uintmax_t *uip, int signedconv)
451 {
452 	char *ep;
453 	int rval;
454 
455 	if (!*gargv) {
456 		*ip = 0;
457 		return (0);
458 	}
459 	if (**gargv == '"' || **gargv == '\'') {
460 		if (signedconv)
461 			*ip = asciicode();
462 		else
463 			*uip = asciicode();
464 		return (0);
465 	}
466 	rval = 0;
467 	errno = 0;
468 	if (signedconv)
469 		*ip = strtoimax(*gargv, &ep, 0);
470 	else
471 		*uip = strtoumax(*gargv, &ep, 0);
472 	if (ep == *gargv) {
473 		warnx("%s: expected numeric value", *gargv);
474 		rval = 1;
475 	}
476 	else if (*ep != '\0') {
477 		warnx("%s: not completely converted", *gargv);
478 		rval = 1;
479 	}
480 	if (errno == ERANGE) {
481 		warnx("%s: %s", *gargv, strerror(ERANGE));
482 		rval = 1;
483 	}
484 	++gargv;
485 	return (rval);
486 }
487 
488 static int
489 getfloating(long double *dp, int mod_ldbl)
490 {
491 	char *ep;
492 	int rval;
493 
494 	if (!*gargv) {
495 		*dp = 0.0;
496 		return (0);
497 	}
498 	if (**gargv == '"' || **gargv == '\'') {
499 		*dp = asciicode();
500 		return (0);
501 	}
502 	rval = 0;
503 	errno = 0;
504 	if (mod_ldbl)
505 		*dp = strtold(*gargv, &ep);
506 	else
507 		*dp = strtod(*gargv, &ep);
508 	if (ep == *gargv) {
509 		warnx("%s: expected numeric value", *gargv);
510 		rval = 1;
511 	} else if (*ep != '\0') {
512 		warnx("%s: not completely converted", *gargv);
513 		rval = 1;
514 	}
515 	if (errno == ERANGE) {
516 		warnx("%s: %s", *gargv, strerror(ERANGE));
517 		rval = 1;
518 	}
519 	++gargv;
520 	return (rval);
521 }
522 
523 static int
524 asciicode(void)
525 {
526 	int ch;
527 
528 	ch = **gargv;
529 	if (ch == '\'' || ch == '"')
530 		ch = (*gargv)[1];
531 	++gargv;
532 	return (ch);
533 }
534 
535 static void
536 usage(void)
537 {
538 	fprintf(stderr, "usage: printf format [arguments ...]\n");
539 }
540