xref: /freebsd-14.2/usr.bin/printf/printf.c (revision 6a6760db)
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  * 4. 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 
30 #ifndef SHELL
31 #ifndef lint
32 static char const copyright[] =
33 "@(#) Copyright (c) 1989, 1993\n\
34 	The Regents of the University of California.  All rights reserved.\n";
35 #endif /* not lint */
36 #endif
37 
38 #ifndef lint
39 #if 0
40 static char const sccsid[] = "@(#)printf.c	8.1 (Berkeley) 7/20/93";
41 #endif
42 static const char rcsid[] =
43   "$FreeBSD$";
44 #endif /* not lint */
45 
46 #include <sys/types.h>
47 
48 #include <err.h>
49 #include <errno.h>
50 #include <inttypes.h>
51 #include <limits.h>
52 #include <locale.h>
53 #include <stdio.h>
54 #include <stdlib.h>
55 #include <string.h>
56 #include <unistd.h>
57 
58 #ifdef SHELL
59 #define main printfcmd
60 #include "bltin/bltin.h"
61 #include "memalloc.h"
62 #include "error.h"
63 #endif
64 
65 #define PF(f, func) do { \
66 	char *b = NULL; \
67 	if (havewidth) \
68 		if (haveprec) \
69 			(void)asprintf(&b, f, fieldwidth, precision, func); \
70 		else \
71 			(void)asprintf(&b, f, fieldwidth, func); \
72 	else if (haveprec) \
73 		(void)asprintf(&b, f, precision, func); \
74 	else \
75 		(void)asprintf(&b, f, func); \
76 	if (b) { \
77 		(void)fputs(b, stdout); \
78 		free(b); \
79 	} \
80 } while (0)
81 
82 static int	 asciicode(void);
83 static char	*printf_doformat(char *, int *);
84 static int	 escape(char *, int, size_t *);
85 static int	 getchr(void);
86 static int	 getfloating(long double *, int);
87 static int	 getint(int *);
88 static int	 getnum(intmax_t *, uintmax_t *, int);
89 static const char
90 		*getstr(void);
91 static char	*mknum(char *, char);
92 static void	 usage(void);
93 
94 static char **gargv;
95 
96 int
97 main(int argc, char *argv[])
98 {
99 	size_t len;
100 	int ch, chopped, end, rval;
101 	char *format, *fmt, *start;
102 
103 #ifndef SHELL
104 	(void) setlocale(LC_ALL, "");
105 #endif
106 #ifdef SHELL
107 	optreset = 1; optind = 1; opterr = 0; /* initialize getopt */
108 #endif
109 	while ((ch = getopt(argc, argv, "")) != -1)
110 		switch (ch) {
111 		case '?':
112 		default:
113 			usage();
114 			return (1);
115 		}
116 	argc -= optind;
117 	argv += optind;
118 
119 	if (argc < 1) {
120 		usage();
121 		return (1);
122 	}
123 
124 #ifdef SHELL
125 	INTOFF;
126 #endif
127 	/*
128 	 * Basic algorithm is to scan the format string for conversion
129 	 * specifications -- once one is found, find out if the field
130 	 * width or precision is a '*'; if it is, gather up value.  Note,
131 	 * format strings are reused as necessary to use up the provided
132 	 * arguments, arguments of zero/null string are provided to use
133 	 * up the format string.
134 	 */
135 	fmt = format = *argv;
136 	chopped = escape(fmt, 1, &len);		/* backslash interpretation */
137 	rval = end = 0;
138 	gargv = ++argv;
139 	for (;;) {
140 		start = fmt;
141 		while (fmt < format + len) {
142 			if (fmt[0] == '%') {
143 				fwrite(start, 1, fmt - start, stdout);
144 				if (fmt[1] == '%') {
145 					/* %% prints a % */
146 					putchar('%');
147 					fmt += 2;
148 				} else {
149 					fmt = printf_doformat(fmt, &rval);
150 					if (fmt == NULL) {
151 #ifdef SHELL
152 						INTON;
153 #endif
154 						return (1);
155 					}
156 					end = 0;
157 				}
158 				start = fmt;
159 			} else
160 				fmt++;
161 		}
162 
163 		if (end == 1) {
164 			warnx("missing format character");
165 #ifdef SHELL
166 			INTON;
167 #endif
168 			return (1);
169 		}
170 		fwrite(start, 1, fmt - start, stdout);
171 		if (chopped || !*gargv) {
172 #ifdef SHELL
173 			INTON;
174 #endif
175 			return (rval);
176 		}
177 		/* Restart at the beginning of the format string. */
178 		fmt = format;
179 		end = 1;
180 	}
181 	/* NOTREACHED */
182 }
183 
184 
185 static char *
186 printf_doformat(char *start, int *rval)
187 {
188 	static const char skip1[] = "#'-+ 0";
189 	static const char skip2[] = "0123456789";
190 	char *fmt;
191 	int fieldwidth, haveprec, havewidth, mod_ldbl, precision;
192 	char convch, nextch;
193 
194 	fmt = start + 1;
195 	/* skip to field width */
196 	fmt += strspn(fmt, skip1);
197 	if (*fmt == '*') {
198 		if (getint(&fieldwidth))
199 			return (NULL);
200 		havewidth = 1;
201 		++fmt;
202 	} else {
203 		havewidth = 0;
204 
205 		/* skip to possible '.', get following precision */
206 		fmt += strspn(fmt, skip2);
207 	}
208 	if (*fmt == '.') {
209 		/* precision present? */
210 		++fmt;
211 		if (*fmt == '*') {
212 			if (getint(&precision))
213 				return (NULL);
214 			haveprec = 1;
215 			++fmt;
216 		} else {
217 			haveprec = 0;
218 
219 			/* skip to conversion char */
220 			fmt += strspn(fmt, skip2);
221 		}
222 	} else
223 		haveprec = 0;
224 	if (!*fmt) {
225 		warnx("missing format character");
226 		return (NULL);
227 	}
228 
229 	/*
230 	 * Look for a length modifier.  POSIX doesn't have these, so
231 	 * we only support them for floating-point conversions, which
232 	 * are extensions.  This is useful because the L modifier can
233 	 * be used to gain extra range and precision, while omitting
234 	 * it is more likely to produce consistent results on different
235 	 * architectures.  This is not so important for integers
236 	 * because overflow is the only bad thing that can happen to
237 	 * them, but consider the command  printf %a 1.1
238 	 */
239 	if (*fmt == 'L') {
240 		mod_ldbl = 1;
241 		fmt++;
242 		if (!strchr("aAeEfFgG", *fmt)) {
243 			warnx("bad modifier L for %%%c", *fmt);
244 			return (NULL);
245 		}
246 	} else {
247 		mod_ldbl = 0;
248 	}
249 
250 	convch = *fmt;
251 	nextch = *++fmt;
252 	*fmt = '\0';
253 	switch (convch) {
254 	case 'b': {
255 		size_t len;
256 		char *p;
257 		int getout;
258 
259 #ifdef SHELL
260 		p = savestr(getstr());
261 #else
262 		p = strdup(getstr());
263 #endif
264 		if (p == NULL) {
265 			warnx("%s", strerror(ENOMEM));
266 			return (NULL);
267 		}
268 		getout = escape(p, 0, &len);
269 		*(fmt - 1) = 's';
270 		PF(start, p);
271 		*(fmt - 1) = 'b';
272 #ifdef SHELL
273 		ckfree(p);
274 #else
275 		free(p);
276 #endif
277 		if (getout)
278 			return (fmt);
279 		break;
280 	}
281 	case 'c': {
282 		char p;
283 
284 		p = getchr();
285 		PF(start, p);
286 		break;
287 	}
288 	case 's': {
289 		const char *p;
290 
291 		p = getstr();
292 		PF(start, p);
293 		break;
294 	}
295 	case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': {
296 		char *f;
297 		intmax_t val;
298 		uintmax_t uval;
299 		int signedconv;
300 
301 		signedconv = (convch == 'd' || convch == 'i');
302 		if ((f = mknum(start, convch)) == NULL)
303 			return (NULL);
304 		if (getnum(&val, &uval, signedconv))
305 			*rval = 1;
306 		if (signedconv)
307 			PF(f, val);
308 		else
309 			PF(f, uval);
310 		break;
311 	}
312 	case 'e': case 'E':
313 	case 'f': case 'F':
314 	case 'g': case 'G':
315 	case 'a': case 'A': {
316 		long double p;
317 
318 		if (getfloating(&p, mod_ldbl))
319 			*rval = 1;
320 		if (mod_ldbl)
321 			PF(start, p);
322 		else
323 			PF(start, (double)p);
324 		break;
325 	}
326 	default:
327 		warnx("illegal format character %c", convch);
328 		return (NULL);
329 	}
330 	*fmt = nextch;
331 	return (fmt);
332 }
333 
334 static char *
335 mknum(char *str, char ch)
336 {
337 	static char *copy;
338 	static size_t copy_size;
339 	char *newcopy;
340 	size_t len, newlen;
341 
342 	len = strlen(str) + 2;
343 	if (len > copy_size) {
344 		newlen = ((len + 1023) >> 10) << 10;
345 #ifdef SHELL
346 		if ((newcopy = ckrealloc(copy, newlen)) == NULL)
347 #else
348 		if ((newcopy = realloc(copy, newlen)) == NULL)
349 #endif
350 		{
351 			warnx("%s", strerror(ENOMEM));
352 			return (NULL);
353 		}
354 		copy = newcopy;
355 		copy_size = newlen;
356 	}
357 
358 	memmove(copy, str, len - 3);
359 	copy[len - 3] = 'j';
360 	copy[len - 2] = ch;
361 	copy[len - 1] = '\0';
362 	return (copy);
363 }
364 
365 static int
366 escape(char *fmt, int percent, size_t *len)
367 {
368 	char *save, *store;
369 	int value, c;
370 
371 	for (save = store = fmt; (c = *fmt); ++fmt, ++store) {
372 		if (c != '\\') {
373 			*store = c;
374 			continue;
375 		}
376 		switch (*++fmt) {
377 		case '\0':		/* EOS, user error */
378 			*store = '\\';
379 			*++store = '\0';
380 			*len = store - save;
381 			return (0);
382 		case '\\':		/* backslash */
383 		case '\'':		/* single quote */
384 			*store = *fmt;
385 			break;
386 		case 'a':		/* bell/alert */
387 			*store = '\a';
388 			break;
389 		case 'b':		/* backspace */
390 			*store = '\b';
391 			break;
392 		case 'c':
393 			*store = '\0';
394 			*len = store - save;
395 			return (1);
396 		case 'f':		/* form-feed */
397 			*store = '\f';
398 			break;
399 		case 'n':		/* newline */
400 			*store = '\n';
401 			break;
402 		case 'r':		/* carriage-return */
403 			*store = '\r';
404 			break;
405 		case 't':		/* horizontal tab */
406 			*store = '\t';
407 			break;
408 		case 'v':		/* vertical tab */
409 			*store = '\v';
410 			break;
411 					/* octal constant */
412 		case '0': case '1': case '2': case '3':
413 		case '4': case '5': case '6': case '7':
414 			c = (!percent && *fmt == '0') ? 4 : 3;
415 			for (value = 0;
416 			    c-- && *fmt >= '0' && *fmt <= '7'; ++fmt) {
417 				value <<= 3;
418 				value += *fmt - '0';
419 			}
420 			--fmt;
421 			if (percent && value == '%') {
422 				*store++ = '%';
423 				*store = '%';
424 			} else
425 				*store = value;
426 			break;
427 		default:
428 			*store = *fmt;
429 			break;
430 		}
431 	}
432 	*store = '\0';
433 	*len = store - save;
434 	return (0);
435 }
436 
437 static int
438 getchr(void)
439 {
440 	if (!*gargv)
441 		return ('\0');
442 	return ((int)**gargv++);
443 }
444 
445 static const char *
446 getstr(void)
447 {
448 	if (!*gargv)
449 		return ("");
450 	return (*gargv++);
451 }
452 
453 static int
454 getint(int *ip)
455 {
456 	intmax_t val;
457 	uintmax_t uval;
458 	int rval;
459 
460 	if (getnum(&val, &uval, 1))
461 		return (1);
462 	rval = 0;
463 	if (val < INT_MIN || val > INT_MAX) {
464 		warnx("%s: %s", *gargv, strerror(ERANGE));
465 		rval = 1;
466 	}
467 	*ip = (int)val;
468 	return (rval);
469 }
470 
471 static int
472 getnum(intmax_t *ip, uintmax_t *uip, int signedconv)
473 {
474 	char *ep;
475 	int rval;
476 
477 	if (!*gargv) {
478 		*ip = 0;
479 		return (0);
480 	}
481 	if (**gargv == '"' || **gargv == '\'') {
482 		if (signedconv)
483 			*ip = asciicode();
484 		else
485 			*uip = asciicode();
486 		return (0);
487 	}
488 	rval = 0;
489 	errno = 0;
490 	if (signedconv)
491 		*ip = strtoimax(*gargv, &ep, 0);
492 	else
493 		*uip = strtoumax(*gargv, &ep, 0);
494 	if (ep == *gargv) {
495 		warnx("%s: expected numeric value", *gargv);
496 		rval = 1;
497 	}
498 	else if (*ep != '\0') {
499 		warnx("%s: not completely converted", *gargv);
500 		rval = 1;
501 	}
502 	if (errno == ERANGE) {
503 		warnx("%s: %s", *gargv, strerror(ERANGE));
504 		rval = 1;
505 	}
506 	++gargv;
507 	return (rval);
508 }
509 
510 static int
511 getfloating(long double *dp, int mod_ldbl)
512 {
513 	char *ep;
514 	int rval;
515 
516 	if (!*gargv) {
517 		*dp = 0.0;
518 		return (0);
519 	}
520 	if (**gargv == '"' || **gargv == '\'') {
521 		*dp = asciicode();
522 		return (0);
523 	}
524 	rval = 0;
525 	errno = 0;
526 	if (mod_ldbl)
527 		*dp = strtold(*gargv, &ep);
528 	else
529 		*dp = strtod(*gargv, &ep);
530 	if (ep == *gargv) {
531 		warnx("%s: expected numeric value", *gargv);
532 		rval = 1;
533 	} else if (*ep != '\0') {
534 		warnx("%s: not completely converted", *gargv);
535 		rval = 1;
536 	}
537 	if (errno == ERANGE) {
538 		warnx("%s: %s", *gargv, strerror(ERANGE));
539 		rval = 1;
540 	}
541 	++gargv;
542 	return (rval);
543 }
544 
545 static int
546 asciicode(void)
547 {
548 	int ch;
549 
550 	ch = **gargv;
551 	if (ch == '\'' || ch == '"')
552 		ch = (*gargv)[1];
553 	++gargv;
554 	return (ch);
555 }
556 
557 static void
558 usage(void)
559 {
560 	(void)fprintf(stderr, "usage: printf format [arguments ...]\n");
561 }
562