xref: /freebsd/usr.bin/printf/printf.c (revision 10b9d77bf1ccf2f3affafa6261692cb92cf7e992)
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 "error.h"
62 #endif
63 
64 #define PF(f, func) do { \
65 	char *b = NULL; \
66 	if (havewidth) \
67 		if (haveprec) \
68 			(void)asprintf(&b, f, fieldwidth, precision, func); \
69 		else \
70 			(void)asprintf(&b, f, fieldwidth, func); \
71 	else if (haveprec) \
72 		(void)asprintf(&b, f, precision, func); \
73 	else \
74 		(void)asprintf(&b, f, func); \
75 	if (b) { \
76 		(void)fputs(b, stdout); \
77 		free(b); \
78 	} \
79 } while (0)
80 
81 static int	 asciicode(void);
82 static char	*printf_doformat(char *, int *);
83 static int	 escape(char *, int, size_t *);
84 static int	 getchr(void);
85 static int	 getfloating(long double *, int);
86 static int	 getint(int *);
87 static int	 getnum(intmax_t *, uintmax_t *, int);
88 static const char
89 		*getstr(void);
90 static char	*mknum(char *, char);
91 static void	 usage(void);
92 
93 static char **gargv;
94 
95 int
96 main(int argc, char *argv[])
97 {
98 	size_t len;
99 	int ch, chopped, end, rval;
100 	char *format, *fmt, *start;
101 
102 #ifndef SHELL
103 	(void) setlocale(LC_ALL, "");
104 #endif
105 #ifdef SHELL
106 	optreset = 1; optind = 1; opterr = 0; /* initialize getopt */
107 #endif
108 	while ((ch = getopt(argc, argv, "")) != -1)
109 		switch (ch) {
110 		case '?':
111 		default:
112 			usage();
113 			return (1);
114 		}
115 	argc -= optind;
116 	argv += optind;
117 
118 	if (argc < 1) {
119 		usage();
120 		return (1);
121 	}
122 
123 #ifdef SHELL
124 	INTOFF;
125 #endif
126 	/*
127 	 * Basic algorithm is to scan the format string for conversion
128 	 * specifications -- once one is found, find out if the field
129 	 * width or precision is a '*'; if it is, gather up value.  Note,
130 	 * format strings are reused as necessary to use up the provided
131 	 * arguments, arguments of zero/null string are provided to use
132 	 * up the format string.
133 	 */
134 	fmt = format = *argv;
135 	chopped = escape(fmt, 1, &len);		/* backslash interpretation */
136 	rval = end = 0;
137 	gargv = ++argv;
138 	for (;;) {
139 		start = fmt;
140 		while (fmt < format + len) {
141 			if (fmt[0] == '%') {
142 				fwrite(start, 1, fmt - start, stdout);
143 				if (fmt[1] == '%') {
144 					/* %% prints a % */
145 					putchar('%');
146 					fmt += 2;
147 				} else {
148 					fmt = printf_doformat(fmt, &rval);
149 					if (fmt == NULL) {
150 #ifdef SHELL
151 						INTON;
152 #endif
153 						return (1);
154 					}
155 					end = 0;
156 				}
157 				start = fmt;
158 			} else
159 				fmt++;
160 		}
161 
162 		if (end == 1) {
163 			warnx("missing format character");
164 #ifdef SHELL
165 			INTON;
166 #endif
167 			return (1);
168 		}
169 		fwrite(start, 1, fmt - start, stdout);
170 		if (chopped || !*gargv) {
171 #ifdef SHELL
172 			INTON;
173 #endif
174 			return (rval);
175 		}
176 		/* Restart at the beginning of the format string. */
177 		fmt = format;
178 		end = 1;
179 	}
180 	/* NOTREACHED */
181 }
182 
183 
184 static char *
185 printf_doformat(char *start, int *rval)
186 {
187 	static const char skip1[] = "#'-+ 0";
188 	static const char skip2[] = "0123456789";
189 	char *fmt;
190 	int fieldwidth, haveprec, havewidth, mod_ldbl, precision;
191 	char convch, nextch;
192 
193 	fmt = start + 1;
194 	/* skip to field width */
195 	fmt += strspn(fmt, skip1);
196 	if (*fmt == '*') {
197 		if (getint(&fieldwidth))
198 			return (NULL);
199 		havewidth = 1;
200 		++fmt;
201 	} else {
202 		havewidth = 0;
203 
204 		/* skip to possible '.', get following precision */
205 		fmt += strspn(fmt, skip2);
206 	}
207 	if (*fmt == '.') {
208 		/* precision present? */
209 		++fmt;
210 		if (*fmt == '*') {
211 			if (getint(&precision))
212 				return (NULL);
213 			haveprec = 1;
214 			++fmt;
215 		} else {
216 			haveprec = 0;
217 
218 			/* skip to conversion char */
219 			fmt += strspn(fmt, skip2);
220 		}
221 	} else
222 		haveprec = 0;
223 	if (!*fmt) {
224 		warnx("missing format character");
225 		return (NULL);
226 	}
227 
228 	/*
229 	 * Look for a length modifier.  POSIX doesn't have these, so
230 	 * we only support them for floating-point conversions, which
231 	 * are extensions.  This is useful because the L modifier can
232 	 * be used to gain extra range and precision, while omitting
233 	 * it is more likely to produce consistent results on different
234 	 * architectures.  This is not so important for integers
235 	 * because overflow is the only bad thing that can happen to
236 	 * them, but consider the command  printf %a 1.1
237 	 */
238 	if (*fmt == 'L') {
239 		mod_ldbl = 1;
240 		fmt++;
241 		if (!strchr("aAeEfFgG", *fmt)) {
242 			warnx("bad modifier L for %%%c", *fmt);
243 			return (NULL);
244 		}
245 	} else {
246 		mod_ldbl = 0;
247 	}
248 
249 	convch = *fmt;
250 	nextch = *++fmt;
251 	*fmt = '\0';
252 	switch (convch) {
253 	case 'b': {
254 		size_t len;
255 		char *p;
256 		int getout;
257 
258 		p = strdup(getstr());
259 		if (p == NULL) {
260 			warnx("%s", strerror(ENOMEM));
261 			return (NULL);
262 		}
263 		getout = escape(p, 0, &len);
264 		*(fmt - 1) = 's';
265 		PF(start, p);
266 		*(fmt - 1) = 'b';
267 		free(p);
268 		if (getout)
269 			return (fmt);
270 		break;
271 	}
272 	case 'c': {
273 		char p;
274 
275 		p = getchr();
276 		PF(start, p);
277 		break;
278 	}
279 	case 's': {
280 		const char *p;
281 
282 		p = getstr();
283 		PF(start, p);
284 		break;
285 	}
286 	case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': {
287 		char *f;
288 		intmax_t val;
289 		uintmax_t uval;
290 		int signedconv;
291 
292 		signedconv = (convch == 'd' || convch == 'i');
293 		if ((f = mknum(start, convch)) == NULL)
294 			return (NULL);
295 		if (getnum(&val, &uval, signedconv))
296 			*rval = 1;
297 		if (signedconv)
298 			PF(f, val);
299 		else
300 			PF(f, uval);
301 		break;
302 	}
303 	case 'e': case 'E':
304 	case 'f': case 'F':
305 	case 'g': case 'G':
306 	case 'a': case 'A': {
307 		long double p;
308 
309 		if (getfloating(&p, mod_ldbl))
310 			*rval = 1;
311 		if (mod_ldbl)
312 			PF(start, p);
313 		else
314 			PF(start, (double)p);
315 		break;
316 	}
317 	default:
318 		warnx("illegal format character %c", convch);
319 		return (NULL);
320 	}
321 	*fmt = nextch;
322 	return (fmt);
323 }
324 
325 static char *
326 mknum(char *str, char ch)
327 {
328 	static char *copy;
329 	static size_t copy_size;
330 	char *newcopy;
331 	size_t len, newlen;
332 
333 	len = strlen(str) + 2;
334 	if (len > copy_size) {
335 		newlen = ((len + 1023) >> 10) << 10;
336 		if ((newcopy = realloc(copy, newlen)) == NULL)
337 		{
338 			warnx("%s", strerror(ENOMEM));
339 			return (NULL);
340 		}
341 		copy = newcopy;
342 		copy_size = newlen;
343 	}
344 
345 	memmove(copy, str, len - 3);
346 	copy[len - 3] = 'j';
347 	copy[len - 2] = ch;
348 	copy[len - 1] = '\0';
349 	return (copy);
350 }
351 
352 static int
353 escape(char *fmt, int percent, size_t *len)
354 {
355 	char *save, *store;
356 	int value, c;
357 
358 	for (save = store = fmt; (c = *fmt); ++fmt, ++store) {
359 		if (c != '\\') {
360 			*store = c;
361 			continue;
362 		}
363 		switch (*++fmt) {
364 		case '\0':		/* EOS, user error */
365 			*store = '\\';
366 			*++store = '\0';
367 			*len = store - save;
368 			return (0);
369 		case '\\':		/* backslash */
370 		case '\'':		/* single quote */
371 			*store = *fmt;
372 			break;
373 		case 'a':		/* bell/alert */
374 			*store = '\a';
375 			break;
376 		case 'b':		/* backspace */
377 			*store = '\b';
378 			break;
379 		case 'c':
380 			*store = '\0';
381 			*len = store - save;
382 			return (1);
383 		case 'f':		/* form-feed */
384 			*store = '\f';
385 			break;
386 		case 'n':		/* newline */
387 			*store = '\n';
388 			break;
389 		case 'r':		/* carriage-return */
390 			*store = '\r';
391 			break;
392 		case 't':		/* horizontal tab */
393 			*store = '\t';
394 			break;
395 		case 'v':		/* vertical tab */
396 			*store = '\v';
397 			break;
398 					/* octal constant */
399 		case '0': case '1': case '2': case '3':
400 		case '4': case '5': case '6': case '7':
401 			c = (!percent && *fmt == '0') ? 4 : 3;
402 			for (value = 0;
403 			    c-- && *fmt >= '0' && *fmt <= '7'; ++fmt) {
404 				value <<= 3;
405 				value += *fmt - '0';
406 			}
407 			--fmt;
408 			if (percent && value == '%') {
409 				*store++ = '%';
410 				*store = '%';
411 			} else
412 				*store = value;
413 			break;
414 		default:
415 			*store = *fmt;
416 			break;
417 		}
418 	}
419 	*store = '\0';
420 	*len = store - save;
421 	return (0);
422 }
423 
424 static int
425 getchr(void)
426 {
427 	if (!*gargv)
428 		return ('\0');
429 	return ((int)**gargv++);
430 }
431 
432 static const char *
433 getstr(void)
434 {
435 	if (!*gargv)
436 		return ("");
437 	return (*gargv++);
438 }
439 
440 static int
441 getint(int *ip)
442 {
443 	intmax_t val;
444 	uintmax_t uval;
445 	int rval;
446 
447 	if (getnum(&val, &uval, 1))
448 		return (1);
449 	rval = 0;
450 	if (val < INT_MIN || val > INT_MAX) {
451 		warnx("%s: %s", *gargv, strerror(ERANGE));
452 		rval = 1;
453 	}
454 	*ip = (int)val;
455 	return (rval);
456 }
457 
458 static int
459 getnum(intmax_t *ip, uintmax_t *uip, int signedconv)
460 {
461 	char *ep;
462 	int rval;
463 
464 	if (!*gargv) {
465 		*ip = 0;
466 		return (0);
467 	}
468 	if (**gargv == '"' || **gargv == '\'') {
469 		if (signedconv)
470 			*ip = asciicode();
471 		else
472 			*uip = asciicode();
473 		return (0);
474 	}
475 	rval = 0;
476 	errno = 0;
477 	if (signedconv)
478 		*ip = strtoimax(*gargv, &ep, 0);
479 	else
480 		*uip = strtoumax(*gargv, &ep, 0);
481 	if (ep == *gargv) {
482 		warnx("%s: expected numeric value", *gargv);
483 		rval = 1;
484 	}
485 	else if (*ep != '\0') {
486 		warnx("%s: not completely converted", *gargv);
487 		rval = 1;
488 	}
489 	if (errno == ERANGE) {
490 		warnx("%s: %s", *gargv, strerror(ERANGE));
491 		rval = 1;
492 	}
493 	++gargv;
494 	return (rval);
495 }
496 
497 static int
498 getfloating(long double *dp, int mod_ldbl)
499 {
500 	char *ep;
501 	int rval;
502 
503 	if (!*gargv) {
504 		*dp = 0.0;
505 		return (0);
506 	}
507 	if (**gargv == '"' || **gargv == '\'') {
508 		*dp = asciicode();
509 		return (0);
510 	}
511 	rval = 0;
512 	errno = 0;
513 	if (mod_ldbl)
514 		*dp = strtold(*gargv, &ep);
515 	else
516 		*dp = strtod(*gargv, &ep);
517 	if (ep == *gargv) {
518 		warnx("%s: expected numeric value", *gargv);
519 		rval = 1;
520 	} else if (*ep != '\0') {
521 		warnx("%s: not completely converted", *gargv);
522 		rval = 1;
523 	}
524 	if (errno == ERANGE) {
525 		warnx("%s: %s", *gargv, strerror(ERANGE));
526 		rval = 1;
527 	}
528 	++gargv;
529 	return (rval);
530 }
531 
532 static int
533 asciicode(void)
534 {
535 	int ch;
536 
537 	ch = **gargv;
538 	if (ch == '\'' || ch == '"')
539 		ch = (*gargv)[1];
540 	++gargv;
541 	return (ch);
542 }
543 
544 static void
545 usage(void)
546 {
547 	(void)fprintf(stderr, "usage: printf format [arguments ...]\n");
548 }
549