1 /*-
2 * SPDX-License-Identifier: BSD-3-Clause
3 *
4 * Copyright (c) 1980, 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 * 3. 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
33 #ifndef lint
34 static const char copyright[] =
35 "@(#) Copyright (c) 1980, 1993\n\
36 The Regents of the University of California. All rights reserved.\n";
37 #endif
38
39 #ifndef lint
40 static const char sccsid[] = "@(#)xstr.c 8.1 (Berkeley) 6/9/93";
41 #endif
42
43 #include <sys/types.h>
44
45 #include <ctype.h>
46 #include <err.h>
47 #include <stdio.h>
48 #include <stdlib.h>
49 #include <signal.h>
50 #include <string.h>
51 #include <unistd.h>
52
53 #include "pathnames.h"
54
55 /*
56 * xstr - extract and hash strings in a C program
57 *
58 * Bill Joy UCB
59 * November, 1978
60 */
61
62 #define ignore(a) ((void) a)
63
64 static off_t tellpt;
65
66 static off_t mesgpt;
67 static char cstrings[] = "strings";
68 static char *strings = cstrings;
69
70 static int cflg;
71 static int vflg;
72 static int readstd;
73
74 static char lastchr(char *);
75
76 static int fgetNUL(char *, int, FILE *);
77 static int istail(char *, char *);
78 static int octdigit(char);
79 static int xgetc(FILE *);
80
81 static off_t hashit(char *, int);
82 static off_t yankstr(char **);
83
84 static void usage(void) __dead2;
85
86 static void flushsh(void);
87 static void found(int, off_t, char *);
88 static void inithash(void);
89 static void onintr(int);
90 static void process(const char *);
91 static void prstr(char *);
92 static void xsdotc(void);
93
94 int
main(int argc,char * argv[])95 main(int argc, char *argv[])
96 {
97 int c;
98 int fdesc;
99
100 while ((c = getopt(argc, argv, "-cv")) != -1)
101 switch (c) {
102 case '-':
103 readstd++;
104 break;
105 case 'c':
106 cflg++;
107 break;
108 case 'v':
109 vflg++;
110 break;
111 default:
112 usage();
113 }
114 argc -= optind;
115 argv += optind;
116
117 if (signal(SIGINT, SIG_IGN) == SIG_DFL)
118 signal(SIGINT, onintr);
119 if (cflg || (argc == 0 && !readstd))
120 inithash();
121 else {
122 strings = strdup(_PATH_TMP);
123 if (strings == NULL)
124 err(1, "strdup() failed");
125 fdesc = mkstemp(strings);
126 if (fdesc == -1)
127 err(1, "Unable to create temporary file");
128 close(fdesc);
129 }
130
131 while (readstd || argc > 0) {
132 if (freopen("x.c", "w", stdout) == NULL)
133 err(1, "x.c");
134 if (!readstd && freopen(argv[0], "r", stdin) == NULL)
135 err(2, "%s", argv[0]);
136 process("x.c");
137 if (readstd == 0)
138 argc--, argv++;
139 else
140 readstd = 0;
141 }
142 flushsh();
143 if (cflg == 0)
144 xsdotc();
145 if (strings[0] == '/')
146 ignore(unlink(strings));
147 exit(0);
148 }
149
150 static void
usage(void)151 usage(void)
152 {
153 fprintf(stderr, "usage: xstr [-cv] [-] [file ...]\n");
154 exit (1);
155 }
156
157 static char linebuf[BUFSIZ];
158
159 static void
process(const char * name)160 process(const char *name)
161 {
162 char *cp;
163 int c;
164 int incomm = 0;
165 int ret;
166
167 printf("extern char\txstr[];\n");
168 for (;;) {
169 if (fgets(linebuf, sizeof linebuf, stdin) == NULL) {
170 if (ferror(stdin))
171 err(3, "%s", name);
172 break;
173 }
174 if (linebuf[0] == '#') {
175 if (linebuf[1] == ' ' && isdigit(linebuf[2]))
176 printf("#line%s", &linebuf[1]);
177 else
178 printf("%s", linebuf);
179 continue;
180 }
181 for (cp = linebuf; (c = *cp++);) switch (c) {
182
183 case '"':
184 if (incomm)
185 goto def;
186 if ((ret = (int) yankstr(&cp)) == -1)
187 goto out;
188 printf("(&xstr[%d])", ret);
189 break;
190
191 case '\'':
192 if (incomm)
193 goto def;
194 putchar(c);
195 if (*cp)
196 putchar(*cp++);
197 break;
198
199 case '/':
200 if (incomm || *cp != '*')
201 goto def;
202 incomm = 1;
203 cp++;
204 printf("/*");
205 continue;
206
207 case '*':
208 if (incomm && *cp == '/') {
209 incomm = 0;
210 cp++;
211 printf("*/");
212 continue;
213 }
214 goto def;
215
216 def:
217 default:
218 putchar(c);
219 break;
220 }
221 }
222 out:
223 if (ferror(stdout))
224 warn("x.c"), onintr(0);
225 }
226
227 static off_t
yankstr(char ** cpp)228 yankstr(char **cpp)
229 {
230 char *cp = *cpp;
231 int c, ch;
232 char dbuf[BUFSIZ];
233 char *dp = dbuf;
234 char *tp;
235 static char tmp[] = "b\bt\tr\rn\nf\f\\\\\"\"";
236
237 while ((c = *cp++)) {
238 if (dp == dbuf + sizeof(dbuf) - 3)
239 errx(1, "message too long");
240 switch (c) {
241
242 case '"':
243 cp++;
244 goto out;
245
246 case '\\':
247 c = *cp++;
248 if (c == 0)
249 break;
250 if (c == '\n') {
251 if (fgets(linebuf, sizeof linebuf, stdin)
252 == NULL) {
253 if (ferror(stdin))
254 err(3, "x.c");
255 return(-1);
256 }
257 cp = linebuf;
258 continue;
259 }
260 for (tp = tmp; (ch = *tp++); tp++)
261 if (c == ch) {
262 c = *tp;
263 goto gotc;
264 }
265 if (!octdigit(c)) {
266 *dp++ = '\\';
267 break;
268 }
269 c -= '0';
270 if (!octdigit(*cp))
271 break;
272 c <<= 3, c += *cp++ - '0';
273 if (!octdigit(*cp))
274 break;
275 c <<= 3, c += *cp++ - '0';
276 break;
277 }
278 gotc:
279 *dp++ = c;
280 }
281 out:
282 *cpp = --cp;
283 *dp = 0;
284 return (hashit(dbuf, 1));
285 }
286
287 static int
octdigit(char c)288 octdigit(char c)
289 {
290 return (isdigit(c) && c != '8' && c != '9');
291 }
292
293 static void
inithash(void)294 inithash(void)
295 {
296 char buf[BUFSIZ];
297 FILE *mesgread = fopen(strings, "r");
298
299 if (mesgread == NULL)
300 return;
301 for (;;) {
302 mesgpt = tellpt;
303 if (fgetNUL(buf, sizeof buf, mesgread) == 0)
304 break;
305 ignore(hashit(buf, 0));
306 }
307 ignore(fclose(mesgread));
308 }
309
310 static int
fgetNUL(char * obuf,int rmdr,FILE * file)311 fgetNUL(char *obuf, int rmdr, FILE *file)
312 {
313 int c;
314 char *buf = obuf;
315
316 while (--rmdr > 0 && (c = xgetc(file)) != 0 && c != EOF)
317 *buf++ = c;
318 *buf++ = 0;
319 return ((feof(file) || ferror(file)) ? 0 : 1);
320 }
321
322 static int
xgetc(FILE * file)323 xgetc(FILE *file)
324 {
325
326 tellpt++;
327 return (getc(file));
328 }
329
330 #define BUCKETS 128
331
332 static struct hash {
333 off_t hpt;
334 char *hstr;
335 struct hash *hnext;
336 short hnew;
337 } bucket[BUCKETS];
338
339 static off_t
hashit(char * str,int new)340 hashit(char *str, int new)
341 {
342 int i;
343 struct hash *hp, *hp0;
344
345 hp = hp0 = &bucket[lastchr(str) & 0177];
346 while (hp->hnext) {
347 hp = hp->hnext;
348 i = istail(str, hp->hstr);
349 if (i >= 0)
350 return (hp->hpt + i);
351 }
352 if ((hp = (struct hash *) calloc(1, sizeof (*hp))) == NULL)
353 errx(8, "calloc");
354 hp->hpt = mesgpt;
355 if (!(hp->hstr = strdup(str)))
356 err(1, NULL);
357 mesgpt += strlen(hp->hstr) + 1;
358 hp->hnext = hp0->hnext;
359 hp->hnew = new;
360 hp0->hnext = hp;
361 return (hp->hpt);
362 }
363
364 static void
flushsh(void)365 flushsh(void)
366 {
367 int i;
368 struct hash *hp;
369 FILE *mesgwrit;
370 int old = 0, new = 0;
371
372 for (i = 0; i < BUCKETS; i++)
373 for (hp = bucket[i].hnext; hp != NULL; hp = hp->hnext)
374 if (hp->hnew)
375 new++;
376 else
377 old++;
378 if (new == 0 && old != 0)
379 return;
380 mesgwrit = fopen(strings, old ? "r+" : "w");
381 if (mesgwrit == NULL)
382 err(4, "%s", strings);
383 for (i = 0; i < BUCKETS; i++)
384 for (hp = bucket[i].hnext; hp != NULL; hp = hp->hnext) {
385 found(hp->hnew, hp->hpt, hp->hstr);
386 if (hp->hnew) {
387 fseek(mesgwrit, hp->hpt, 0);
388 ignore(fwrite(hp->hstr, strlen(hp->hstr) + 1, 1, mesgwrit));
389 if (ferror(mesgwrit))
390 err(4, "%s", strings);
391 }
392 }
393 if (fclose(mesgwrit) == EOF)
394 err(4, "%s", strings);
395 }
396
397 static void
found(int new,off_t off,char * str)398 found(int new, off_t off, char *str)
399 {
400 if (vflg == 0)
401 return;
402 if (!new)
403 fprintf(stderr, "found at %d:", (int) off);
404 else
405 fprintf(stderr, "new at %d:", (int) off);
406 prstr(str);
407 fprintf(stderr, "\n");
408 }
409
410 static void
prstr(char * cp)411 prstr(char *cp)
412 {
413 int c;
414
415 while ((c = (*cp++ & 0377)))
416 if (c < ' ')
417 fprintf(stderr, "^%c", c + '`');
418 else if (c == 0177)
419 fprintf(stderr, "^?");
420 else if (c > 0200)
421 fprintf(stderr, "\\%03o", c);
422 else
423 fprintf(stderr, "%c", c);
424 }
425
426 static void
xsdotc(void)427 xsdotc(void)
428 {
429 FILE *strf = fopen(strings, "r");
430 FILE *xdotcf;
431
432 if (strf == NULL)
433 err(5, "%s", strings);
434 xdotcf = fopen("xs.c", "w");
435 if (xdotcf == NULL)
436 err(6, "xs.c");
437 fprintf(xdotcf, "char\txstr[] = {\n");
438 for (;;) {
439 int i, c;
440
441 for (i = 0; i < 8; i++) {
442 c = getc(strf);
443 if (ferror(strf)) {
444 warn("%s", strings);
445 onintr(0);
446 }
447 if (feof(strf)) {
448 fprintf(xdotcf, "\n");
449 goto out;
450 }
451 fprintf(xdotcf, "0x%02x,", c);
452 }
453 fprintf(xdotcf, "\n");
454 }
455 out:
456 fprintf(xdotcf, "};\n");
457 ignore(fclose(xdotcf));
458 ignore(fclose(strf));
459 }
460
461 static char
lastchr(char * cp)462 lastchr(char *cp)
463 {
464
465 while (cp[0] && cp[1])
466 cp++;
467 return (*cp);
468 }
469
470 static int
istail(char * str,char * of)471 istail(char *str, char *of)
472 {
473 int d = strlen(of) - strlen(str);
474
475 if (d < 0 || strcmp(&of[d], str) != 0)
476 return (-1);
477 return (d);
478 }
479
480 static void
onintr(int dummy __unused)481 onintr(int dummy __unused)
482 {
483
484 ignore(signal(SIGINT, SIG_IGN));
485 if (strings[0] == '/')
486 ignore(unlink(strings));
487 ignore(unlink("x.c"));
488 ignore(unlink("xs.c"));
489 exit(7);
490 }
491