1 /*
2 * Copyright 2014 Garrett D'Amore <garrett@damore.org>
3 * Copyright 2010 Nexenta Systems, Inc. All rights reserved.
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 * 4. Neither the name of the University nor the names of its contributors
16 * may be used to endorse or promote products derived from this software
17 * without specific prior written permission.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * SUCH DAMAGE.
30 */
31
32 #include <sys/types.h>
33
34 #include <err.h>
35 #include <errno.h>
36 #include <inttypes.h>
37 #include <limits.h>
38 #include <stdio.h>
39 #include <stdlib.h>
40 #include <string.h>
41 #include <unistd.h>
42 #include <alloca.h>
43 #include <ctype.h>
44 #include <locale.h>
45 #include <note.h>
46
47 #define warnx1(a, b, c) warnx(a)
48 #define warnx2(a, b, c) warnx(a, b)
49 #define warnx3(a, b, c) warnx(a, b, c)
50
51 #define PTRDIFF(x, y) ((uintptr_t)(x) - (uintptr_t)(y))
52
53 #define _(x) gettext(x)
54
55 #define PF(f, func) do { \
56 char *b = NULL; \
57 if (havewidth) \
58 if (haveprec) \
59 (void) asprintf(&b, f, fieldwidth, precision, func); \
60 else \
61 (void) asprintf(&b, f, fieldwidth, func); \
62 else if (haveprec) \
63 (void) asprintf(&b, f, precision, func); \
64 else \
65 (void) asprintf(&b, f, func); \
66 if (b) { \
67 (void) fputs(b, stdout); \
68 free(b); \
69 } \
70 _NOTE(CONSTCOND) } while (0)
71
72 static int asciicode(void);
73 static char *doformat(char *, int *);
74 static int escape(char *, int, size_t *);
75 static int getchr(void);
76 static int getfloating(long double *, int);
77 static int getint(int *);
78 static int getnum(intmax_t *, uintmax_t *, int);
79 static const char
80 *getstr(void);
81 static char *mknum(char *, char);
82 static void usage(void);
83
84 static const char digits[] = "0123456789";
85
86 static int myargc;
87 static char **myargv;
88 static char **gargv;
89 static char **maxargv;
90
91 int
92 main(int argc, char *argv[])
93 {
94 size_t len;
95 int end, rval;
96 char *format, *fmt, *start;
97
98 (void) setlocale(LC_ALL, "");
99
100 argv++;
101 argc--;
102
103 /*
104 * POSIX says: Standard utilities that do not accept options,
105 * but that do accept operands, shall recognize "--" as a
106 * first argument to be discarded.
107 */
108 if (argc && strcmp(argv[0], "--") == 0) {
109 argc--;
110 argv++;
111 }
112
113 if (argc < 1) {
114 usage();
115 return (1);
116 }
117
118 /*
119 * Basic algorithm is to scan the format string for conversion
120 * specifications -- once one is found, find out if the field
121 * width or precision is a '*'; if it is, gather up value. Note,
122 * format strings are reused as necessary to use up the provided
123 * arguments, arguments of zero/null string are provided to use
124 * up the format string.
125 */
126 fmt = format = *argv;
127 (void) escape(fmt, 1, &len); /* backslash interpretation */
128 rval = end = 0;
129 gargv = ++argv;
130
131 for (;;) {
132 maxargv = gargv;
133
134 myargv = gargv;
135 for (myargc = 0; gargv[myargc]; myargc++)
136 /* nop */;
137 start = fmt;
138 while (fmt < format + len) {
139 if (fmt[0] == '%') {
140 (void) fwrite(start, 1, PTRDIFF(fmt, start),
141 stdout);
142 if (fmt[1] == '%') {
143 /* %% prints a % */
144 (void) putchar('%');
145 fmt += 2;
146 } else {
147 fmt = doformat(fmt, &rval);
148 if (fmt == NULL)
149 return (1);
150 end = 0;
151 }
152 start = fmt;
153 } else
154 fmt++;
155 if (gargv > maxargv)
156 maxargv = gargv;
157 }
158 gargv = maxargv;
159
160 if (end == 1) {
161 warnx1(_("missing format character"), NULL, NULL);
162 return (1);
163 }
164 (void) fwrite(start, 1, PTRDIFF(fmt, start), stdout);
165 if (!*gargv)
166 return (rval);
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 doformat(char *fmt, int *rval)
177 {
178 static const char skip1[] = "#'-+ 0";
179 int fieldwidth, haveprec, havewidth, mod_ldbl, precision;
180 char convch, nextch;
181 char *start;
182 char **fargv;
183 char *dptr;
184 int l;
185
186 start = alloca(strlen(fmt) + 1);
187
188 dptr = start;
189 *dptr++ = '%';
190 *dptr = 0;
191
192 fmt++;
193
194 /* look for "n$" field index specifier */
195 l = strspn(fmt, digits);
196 if ((l > 0) && (fmt[l] == '$')) {
197 int idx = atoi(fmt);
198 if (idx <= myargc) {
199 gargv = &myargv[idx - 1];
200 } else {
201 gargv = &myargv[myargc];
202 }
203 if (gargv > maxargv) {
204 maxargv = gargv;
205 }
206 fmt += l + 1;
207
208 /* save format argument */
209 fargv = gargv;
210 } else {
211 fargv = NULL;
212 }
213
214 /* skip to field width */
215 while (strchr(skip1, *fmt) != NULL) {
216 *dptr++ = *fmt++;
217 *dptr = 0;
218 }
219
220
221 if (*fmt == '*') {
222
223 fmt++;
224 l = strspn(fmt, digits);
225 if ((l > 0) && (fmt[l] == '$')) {
226 int idx = atoi(fmt);
227 if (idx <= myargc) {
228 gargv = &myargv[idx - 1];
229 } else {
230 gargv = &myargv[myargc];
231 }
232 fmt += l + 1;
233 }
234
235 if (getint(&fieldwidth))
236 return (NULL);
237 if (gargv > maxargv) {
238 maxargv = gargv;
239 }
240 havewidth = 1;
241
242 *dptr++ = '*';
243 *dptr = 0;
244 } else {
245 havewidth = 0;
246
247 /* skip to possible '.', get following precision */
248 while (isdigit(*fmt)) {
249 *dptr++ = *fmt++;
250 *dptr = 0;
251 }
252 }
253
254 if (*fmt == '.') {
255 /* precision present? */
256 fmt++;
257 *dptr++ = '.';
258
259 if (*fmt == '*') {
260
261 fmt++;
262 l = strspn(fmt, digits);
263 if ((l > 0) && (fmt[l] == '$')) {
264 int idx = atoi(fmt);
265 if (idx <= myargc) {
266 gargv = &myargv[idx - 1];
267 } else {
268 gargv = &myargv[myargc];
269 }
270 fmt += l + 1;
271 }
272
273 if (getint(&precision))
274 return (NULL);
275 if (gargv > maxargv) {
276 maxargv = gargv;
277 }
278 haveprec = 1;
279 *dptr++ = '*';
280 *dptr = 0;
281 } else {
282 haveprec = 0;
283
284 /* skip to conversion char */
285 while (isdigit(*fmt)) {
286 *dptr++ = *fmt++;
287 *dptr = 0;
288 }
289 }
290 } else
291 haveprec = 0;
292 if (!*fmt) {
293 warnx1(_("missing format character"), NULL, NULL);
294 return (NULL);
295 }
296 *dptr++ = *fmt;
297 *dptr = 0;
298
299 /*
300 * Look for a length modifier. POSIX doesn't have these, so
301 * we only support them for floating-point conversions, which
302 * are extensions. This is useful because the L modifier can
303 * be used to gain extra range and precision, while omitting
304 * it is more likely to produce consistent results on different
305 * architectures. This is not so important for integers
306 * because overflow is the only bad thing that can happen to
307 * them, but consider the command printf %a 1.1
308 */
309 if (*fmt == 'L') {
310 mod_ldbl = 1;
311 fmt++;
312 if (!strchr("aAeEfFgG", *fmt)) {
313 warnx2(_("bad modifier L for %%%c"), *fmt, NULL);
314 return (NULL);
315 }
316 } else {
317 mod_ldbl = 0;
318 }
319
320 /* save the current arg offset, and set to the format arg */
321 if (fargv != NULL) {
322 gargv = fargv;
323 }
324
325 convch = *fmt;
326 nextch = *++fmt;
327
328 *fmt = '\0';
329 switch (convch) {
330 case 'b': {
331 size_t len;
332 char *p;
333 int getout;
334
335 p = strdup(getstr());
336 if (p == NULL) {
337 warnx2("%s", strerror(ENOMEM), NULL);
338 return (NULL);
339 }
340 getout = escape(p, 0, &len);
341 (void) fputs(p, stdout);
342 free(p);
343
344 if (getout)
345 exit(*rval);
346 break;
347 }
348 case 'c': {
349 char p;
350
351 p = getchr();
352 PF(start, p);
353 break;
354 }
355 case 's': {
356 const char *p;
357
358 p = getstr();
359 PF(start, p);
360 break;
361 }
362 case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': {
363 char *f;
364 intmax_t val;
365 uintmax_t uval;
366 int signedconv;
367
368 signedconv = (convch == 'd' || convch == 'i');
369 if ((f = mknum(start, convch)) == NULL)
370 return (NULL);
371 if (getnum(&val, &uval, signedconv))
372 *rval = 1;
373 if (signedconv)
374 PF(f, val);
375 else
376 PF(f, uval);
377 break;
378 }
379 case 'e': case 'E':
380 case 'f': case 'F':
381 case 'g': case 'G':
382 case 'a': case 'A': {
383 long double p;
384
385 if (getfloating(&p, mod_ldbl))
386 *rval = 1;
387 if (mod_ldbl)
388 PF(start, p);
389 else
390 PF(start, (double)p);
391 break;
392 }
393 default:
394 warnx2(_("illegal format character %c"), convch, NULL);
395 return (NULL);
396 }
397 *fmt = nextch;
398
399 /* return the gargv to the next element */
400 return (fmt);
401 }
402
403 static char *
404 mknum(char *str, char ch)
405 {
406 static char *copy;
407 static size_t copy_size;
408 char *newcopy;
409 size_t len, newlen;
410
411 len = strlen(str) + 2;
412 if (len > copy_size) {
413 newlen = ((len + 1023) >> 10) << 10;
414 if ((newcopy = realloc(copy, newlen)) == NULL) {
415 warnx2("%s", strerror(ENOMEM), NULL);
416 return (NULL);
417 }
418 copy = newcopy;
419 copy_size = newlen;
420 }
421
422 (void) memmove(copy, str, len - 3);
423 copy[len - 3] = 'j';
424 copy[len - 2] = ch;
425 copy[len - 1] = '\0';
426 return (copy);
427 }
428
429 static int
430 escape(char *fmt, int percent, size_t *len)
431 {
432 char *save, *store, c;
433 int value;
434
435 for (save = store = fmt; ((c = *fmt) != 0); ++fmt, ++store) {
436 if (c != '\\') {
437 *store = c;
438 continue;
439 }
440 switch (*++fmt) {
441 case '\0': /* EOS, user error */
442 *store = '\\';
443 *++store = '\0';
444 *len = PTRDIFF(store, save);
445 return (0);
446 case '\\': /* backslash */
447 case '\'': /* single quote */
448 *store = *fmt;
449 break;
450 case 'a': /* bell/alert */
451 *store = '\a';
452 break;
453 case 'b': /* backspace */
454 *store = '\b';
455 break;
456 case 'c':
457 if (!percent) {
458 *store = '\0';
459 *len = PTRDIFF(store, save);
460 return (1);
461 }
462 *store = 'c';
463 break;
464 case 'f': /* form-feed */
465 *store = '\f';
466 break;
467 case 'n': /* newline */
468 *store = '\n';
469 break;
470 case 'r': /* carriage-return */
471 *store = '\r';
472 break;
473 case 't': /* horizontal tab */
474 *store = '\t';
475 break;
476 case 'v': /* vertical tab */
477 *store = '\v';
478 break;
479 /* octal constant */
480 case '0': case '1': case '2': case '3':
481 case '4': case '5': case '6': case '7':
482 c = (!percent && *fmt == '0') ? 4 : 3;
483 for (value = 0;
484 c-- && *fmt >= '0' && *fmt <= '7'; ++fmt) {
485 value <<= 3;
486 value += *fmt - '0';
487 }
488 --fmt;
489 if (percent && value == '%') {
490 *store++ = '%';
491 *store = '%';
492 } else
493 *store = (char)value;
494 break;
495 default:
496 *store = *fmt;
497 break;
498 }
499 }
500 *store = '\0';
501 *len = PTRDIFF(store, save);
502 return (0);
503 }
504
505 static int
506 getchr(void)
507 {
508 if (!*gargv)
509 return ('\0');
510 return ((int)**gargv++);
511 }
512
513 static const char *
514 getstr(void)
515 {
516 if (!*gargv)
517 return ("");
518 return (*gargv++);
519 }
520
521 static int
522 getint(int *ip)
523 {
524 intmax_t val;
525 uintmax_t uval;
526 int rval;
527
528 if (getnum(&val, &uval, 1))
529 return (1);
530 rval = 0;
531 if (val < INT_MIN || val > INT_MAX) {
532 warnx3("%s: %s", *gargv, strerror(ERANGE));
533 rval = 1;
534 }
535 *ip = (int)val;
536 return (rval);
537 }
538
539 static int
540 getnum(intmax_t *ip, uintmax_t *uip, int signedconv)
541 {
542 char *ep;
543 int rval;
544
545 if (!*gargv) {
546 *ip = 0;
547 return (0);
548 }
549 if (**gargv == '"' || **gargv == '\'') {
550 if (signedconv)
551 *ip = asciicode();
552 else
553 *uip = asciicode();
554 return (0);
555 }
556 rval = 0;
557 errno = 0;
558 if (signedconv)
559 *ip = strtoimax(*gargv, &ep, 0);
560 else
561 *uip = strtoumax(*gargv, &ep, 0);
562 if (ep == *gargv) {
563 warnx2(_("%s: expected numeric value"), *gargv, NULL);
564 rval = 1;
565 } else if (*ep != '\0') {
566 warnx2(_("%s: not completely converted"), *gargv, NULL);
567 rval = 1;
568 }
569 if (errno == ERANGE) {
570 warnx3("%s: %s", *gargv, strerror(ERANGE));
571 rval = 1;
572 }
573 ++gargv;
574 return (rval);
575 }
576
577 static int
578 getfloating(long double *dp, int mod_ldbl)
579 {
580 char *ep;
581 int rval;
582
583 if (!*gargv) {
584 *dp = 0.0;
585 return (0);
586 }
587 if (**gargv == '"' || **gargv == '\'') {
588 *dp = asciicode();
589 return (0);
590 }
591 rval = 0;
592 errno = 0;
593 if (mod_ldbl)
594 *dp = strtold(*gargv, &ep);
595 else
596 *dp = strtod(*gargv, &ep);
597 if (ep == *gargv) {
598 warnx2(_("%s: expected numeric value"), *gargv, NULL);
599 rval = 1;
600 } else if (*ep != '\0') {
601 warnx2(_("%s: not completely converted"), *gargv, NULL);
602 rval = 1;
603 }
604 if (errno == ERANGE) {
605 warnx3("%s: %s", *gargv, strerror(ERANGE));
606 rval = 1;
607 }
608 ++gargv;
609 return (rval);
610 }
611
612 static int
613 asciicode(void)
614 {
615 int ch;
616
617 ch = **gargv;
618 if (ch == '\'' || ch == '"')
619 ch = (*gargv)[1];
620 ++gargv;
621 return (ch);
622 }
623
624 static void
625 usage(void)
626 {
627 (void) fprintf(stderr, _("usage: printf format [arguments ...]\n"));
628 }