![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to funcs.c CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / php / ext / fileinfo / libmagic|
Return to funcs.c CVS log | Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / php / ext / fileinfo / libmagic |
1.1 misho 1: /*
2: * Copyright (c) Christos Zoulas 2003.
3: * 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 immediately at the beginning of the file, without modification,
10: * this list of conditions, and the following disclaimer.
11: * 2. Redistributions in binary form must reproduce the above copyright
12: * notice, this list of conditions and the following disclaimer in the
13: * documentation and/or other materials provided with the distribution.
14: *
15: * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
16: * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17: * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18: * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
19: * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
25: * SUCH DAMAGE.
26: */
27: #include "file.h"
28:
29: #ifndef lint
30: FILE_RCSID("@(#)$File: funcs.c,v 1.53 2009/04/07 11:07:00 christos Exp $")
31: #endif /* lint */
32:
33: #include "magic.h"
34: #include <stdarg.h>
35: #include <stdlib.h>
36: #include <string.h>
37: #include <ctype.h>
38: #if defined(HAVE_WCHAR_H)
39: #include <wchar.h>
40: #endif
41: #if defined(HAVE_WCTYPE_H)
42: #include <wctype.h>
43: #endif
44:
45: #ifndef SIZE_MAX
46: # define SIZE_MAX ((size_t) -1)
47: #endif
48:
49: /*
50: * Like printf, only we append to a buffer.
51: */
52: protected int
53: file_printf(struct magic_set *ms, const char *fmt, ...)
54: {
55: va_list ap;
56: int len;
57: char *buf = NULL, *newstr;
58:
59: va_start(ap, fmt);
60: len = vspprintf(&buf, 0, fmt, ap);
61: va_end(ap);
62:
63: if (ms->o.buf != NULL) {
64: len = spprintf(&newstr, 0, "%s%s", ms->o.buf, (buf ? buf : ""));
65: if (buf) {
66: efree(buf);
67: }
68: efree(ms->o.buf);
69: ms->o.buf = newstr;
70: } else {
71: ms->o.buf = buf;
72: }
73: return 0;
74: }
75:
76: /*
77: * error - print best error message possible
78: */
79: /*VARARGS*/
80: private void
81: file_error_core(struct magic_set *ms, int error, const char *f, va_list va,
82: uint32_t lineno)
83: {
84: char *buf = NULL;
85:
86: /* Only the first error is ok */
87: if (ms->event_flags & EVENT_HAD_ERR) {
88: return;
89: }
90:
91: if (lineno != 0) {
92: efree(ms->o.buf);
93: ms->o.buf = NULL;
94: file_printf(ms, "line %u: ", lineno);
95: }
96:
97: vspprintf(&buf, 0, f, va);
98: va_end(va);
99:
100: if (error > 0) {
101: file_printf(ms, "%s (%s)", (*buf ? buf : ""), strerror(error));
102: } else if (*buf) {
103: file_printf(ms, "%s", buf);
104: }
105:
106: if (buf) {
107: efree(buf);
108: }
109:
110: ms->event_flags |= EVENT_HAD_ERR;
111: ms->error = error;
112: }
113:
114: /*VARARGS*/
115: protected void
116: file_error(struct magic_set *ms, int error, const char *f, ...)
117: {
118: va_list va;
119: va_start(va, f);
120: file_error_core(ms, error, f, va, 0);
121: va_end(va);
122: }
123:
124: /*
125: * Print an error with magic line number.
126: */
127: /*VARARGS*/
128: protected void
129: file_magerror(struct magic_set *ms, const char *f, ...)
130: {
131: va_list va;
132: va_start(va, f);
133: file_error_core(ms, 0, f, va, ms->line);
134: va_end(va);
135: }
136:
137: protected void
138: file_oomem(struct magic_set *ms, size_t len)
139: {
140: file_error(ms, errno, "cannot allocate %zu bytes", len);
141: }
142:
143: protected void
144: file_badseek(struct magic_set *ms)
145: {
146: file_error(ms, errno, "error seeking");
147: }
148:
149: protected void
150: file_badread(struct magic_set *ms)
151: {
152: file_error(ms, errno, "error reading");
153: }
154:
155: protected int
156: file_buffer(struct magic_set *ms, php_stream *stream, const char *inname, const void *buf,
157: size_t nb)
158: {
159: int m = 0, rv = 0, looks_text = 0;
160: int mime = ms->flags & MAGIC_MIME;
161: const unsigned char *ubuf = buf;
162: unichar *u8buf = NULL;
163: size_t ulen;
164: const char *code = NULL;
165: const char *code_mime = "binary";
166: const char *type = NULL;
167:
168:
169:
170: if (nb == 0) {
171: if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
172: file_printf(ms, mime ? "application/x-empty" :
173: "empty") == -1)
174: return -1;
175: return 1;
176: } else if (nb == 1) {
177: if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
178: file_printf(ms, mime ? "application/octet-stream" :
179: "very short file (no magic)") == -1)
180: return -1;
181: return 1;
182: }
183:
184: if ((ms->flags & MAGIC_NO_CHECK_ENCODING) == 0) {
185: looks_text = file_encoding(ms, ubuf, nb, &u8buf, &ulen,
186: &code, &code_mime, &type);
187: }
188:
189: #if defined(__EMX__)
190: if ((ms->flags & MAGIC_NO_CHECK_APPTYPE) == 0 && inname) {
191: switch (file_os2_apptype(ms, inname, buf, nb)) {
192: case -1:
193: return -1;
194: case 0:
195: break;
196: default:
197: return 1;
198: }
199: }
200: #endif
201:
202: #if PHP_FILEINFO_UNCOMPRESS
203: if ((ms->flags & MAGIC_NO_CHECK_COMPRESS) == 0)
204: if ((m = file_zmagic(ms, stream, inname, ubuf, nb)) != 0) {
205: if ((ms->flags & MAGIC_DEBUG) != 0)
206: (void)fprintf(stderr, "zmagic %d\n", m);
207: goto done;
208: }
209: #endif
210:
211: /* Check if we have a tar file */
212: if ((ms->flags & MAGIC_NO_CHECK_TAR) == 0)
213: if ((m = file_is_tar(ms, ubuf, nb)) != 0) {
214: if ((ms->flags & MAGIC_DEBUG) != 0)
215: (void)fprintf(stderr, "tar %d\n", m);
216: goto done;
217: }
218:
219: /* Check if we have a CDF file */
220: if ((ms->flags & MAGIC_NO_CHECK_CDF) == 0) {
221: int fd;
222: TSRMLS_FETCH();
223: if (stream && SUCCESS == php_stream_cast(stream, PHP_STREAM_AS_FD, (void **)&fd, 0)) {
224: if ((m = file_trycdf(ms, fd, ubuf, nb)) != 0) {
225: if ((ms->flags & MAGIC_DEBUG) != 0)
226: (void)fprintf(stderr, "cdf %d\n", m);
227: goto done;
228: }
229: }
230: }
231:
232: /* try soft magic tests */
233: if ((ms->flags & MAGIC_NO_CHECK_SOFT) == 0)
234: if ((m = file_softmagic(ms, ubuf, nb, BINTEST)) != 0) {
235: if ((ms->flags & MAGIC_DEBUG) != 0)
236: (void)fprintf(stderr, "softmagic %d\n", m);
237: #ifdef BUILTIN_ELF
238: if ((ms->flags & MAGIC_NO_CHECK_ELF) == 0 && m == 1 &&
239: nb > 5 && fd != -1) {
240: /*
241: * We matched something in the file, so this
242: * *might* be an ELF file, and the file is at
243: * least 5 bytes long, so if it's an ELF file
244: * it has at least one byte past the ELF magic
245: * number - try extracting information from the
246: * ELF headers that cannot easily * be
247: * extracted with rules in the magic file.
248: */
249: if ((m = file_tryelf(ms, fd, ubuf, nb)) != 0)
250: if ((ms->flags & MAGIC_DEBUG) != 0)
251: (void)fprintf(stderr,
252: "elf %d\n", m);
253: }
254: #endif
255: goto done;
256: }
257:
258: /* try text properties (and possibly text tokens) */
259: if ((ms->flags & MAGIC_NO_CHECK_TEXT) == 0) {
260:
261: if ((m = file_ascmagic(ms, ubuf, nb)) != 0) {
262: if ((ms->flags & MAGIC_DEBUG) != 0)
263: (void)fprintf(stderr, "ascmagic %d\n", m);
264: goto done;
265: }
266:
267: /* try to discover text encoding */
268: if ((ms->flags & MAGIC_NO_CHECK_ENCODING) == 0) {
269: if (looks_text == 0)
270: if ((m = file_ascmagic_with_encoding( ms, ubuf,
271: nb, u8buf, ulen, code, type)) != 0) {
272: if ((ms->flags & MAGIC_DEBUG) != 0)
273: (void)fprintf(stderr,
274: "ascmagic/enc %d\n", m);
275: goto done;
276: }
277: }
278: }
279:
280: /* give up */
281: m = 1;
282: if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
283: file_printf(ms, mime ? "application/octet-stream" : "data") == -1) {
284: rv = -1;
285: }
286: done:
287: if ((ms->flags & MAGIC_MIME_ENCODING) != 0) {
288: if (ms->flags & MAGIC_MIME_TYPE)
289: if (file_printf(ms, "; charset=") == -1)
290: rv = -1;
291: if (file_printf(ms, "%s", code_mime) == -1)
292: rv = -1;
293: }
294: if (u8buf)
295: free(u8buf);
296: if (rv)
297: return rv;
298:
299: return m;
300: }
301:
302: protected int
303: file_reset(struct magic_set *ms)
304: {
305: if (ms->mlist == NULL) {
306: file_error(ms, 0, "no magic files loaded");
307: return -1;
308: }
309: if (ms->o.buf) {
310: efree(ms->o.buf);
311: ms->o.buf = NULL;
312: }
313: if (ms->o.pbuf) {
314: efree(ms->o.pbuf);
315: ms->o.pbuf = NULL;
316: }
317: ms->event_flags &= ~EVENT_HAD_ERR;
318: ms->error = -1;
319: return 0;
320: }
321:
322: #define OCTALIFY(n, o) \
323: /*LINTED*/ \
324: (void)(*(n)++ = '\\', \
325: *(n)++ = (((uint32_t)*(o) >> 6) & 3) + '0', \
326: *(n)++ = (((uint32_t)*(o) >> 3) & 7) + '0', \
327: *(n)++ = (((uint32_t)*(o) >> 0) & 7) + '0', \
328: (o)++)
329:
330: protected const char *
331: file_getbuffer(struct magic_set *ms)
332: {
333: char *pbuf, *op, *np;
334: size_t psize, len;
335:
336: if (ms->event_flags & EVENT_HAD_ERR)
337: return NULL;
338:
339: if (ms->flags & MAGIC_RAW)
340: return ms->o.buf;
341:
342: if (ms->o.buf == NULL)
343: return NULL;
344:
345: /* * 4 is for octal representation, + 1 is for NUL */
346: len = strlen(ms->o.buf);
347: if (len > (SIZE_MAX - 1) / 4) {
348: return NULL;
349: }
350: psize = len * 4 + 1;
351: pbuf = erealloc(ms->o.pbuf, psize);
352: ms->o.pbuf = pbuf;
353:
354: #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
355: {
356: mbstate_t state;
357: wchar_t nextchar;
358: int mb_conv = 1;
359: size_t bytesconsumed;
360: char *eop;
361: (void)memset(&state, 0, sizeof(mbstate_t));
362:
363: np = ms->o.pbuf;
364: op = ms->o.buf;
365: eop = op + len;
366:
367: while (op < eop) {
368: bytesconsumed = mbrtowc(&nextchar, op,
369: (size_t)(eop - op), &state);
370: if (bytesconsumed == (size_t)(-1) ||
371: bytesconsumed == (size_t)(-2)) {
372: mb_conv = 0;
373: break;
374: }
375:
376: if (iswprint(nextchar)) {
377: (void)memcpy(np, op, bytesconsumed);
378: op += bytesconsumed;
379: np += bytesconsumed;
380: } else {
381: while (bytesconsumed-- > 0)
382: OCTALIFY(np, op);
383: }
384: }
385: *np = '\0';
386:
387: /* Parsing succeeded as a multi-byte sequence */
388: if (mb_conv != 0)
389: return ms->o.pbuf;
390: }
391: #endif
392:
393: for (np = ms->o.pbuf, op = ms->o.buf; *op; op++) {
394: if (isprint((unsigned char)*op)) {
395: *np++ = *op;
396: } else {
397: OCTALIFY(np, op);
398: }
399: }
400: *np = '\0';
401: return ms->o.pbuf;
402: }
403:
404: protected int
405: file_check_mem(struct magic_set *ms, unsigned int level)
406: {
407: size_t len;
408:
409: if (level >= ms->c.len) {
410: len = (ms->c.len += 20) * sizeof(*ms->c.li);
411: ms->c.li = (ms->c.li == NULL) ? emalloc(len) : erealloc(ms->c.li, len);
412: }
413: ms->c.li[level].got_match = 0;
414: #ifdef ENABLE_CONDITIONALS
415: ms->c.li[level].last_match = 0;
416: ms->c.li[level].last_cond = COND_NONE;
417: #endif /* ENABLE_CONDITIONALS */
418: return 0;
419: }