xref: /PHP-5.5/ext/standard/iptc.c (revision 686a1789)
1 /*
2    +----------------------------------------------------------------------+
3    | PHP Version 5                                                        |
4    +----------------------------------------------------------------------+
5    | Copyright (c) 1997-2015 The PHP Group                                |
6    +----------------------------------------------------------------------+
7    | This source file is subject to version 3.01 of the PHP license,      |
8    | that is bundled with this package in the file LICENSE, and is        |
9    | available through the world-wide-web at the following url:           |
10    | http://www.php.net/license/3_01.txt                                  |
11    | If you did not receive a copy of the PHP license and are unable to   |
12    | obtain it through the world-wide-web, please send a note to          |
13    | license@php.net so we can mail you a copy immediately.               |
14    +----------------------------------------------------------------------+
15    | Author: Thies C. Arntzen <thies@thieso.net>                          |
16    +----------------------------------------------------------------------+
17  */
18 
19 /* $Id$ */
20 
21 /*
22  * Functions to parse & compse IPTC data.
23  * PhotoShop >= 3.0 can read and write textual data to JPEG files.
24  * ... more to come .....
25  *
26  * i know, parts of this is now duplicated in image.c
27  * but in this case i think it's okay!
28  */
29 
30 /*
31  * TODO:
32  *  - add IPTC translation table
33  */
34 
35 #include "php.h"
36 #include "php_iptc.h"
37 #include "ext/standard/head.h"
38 
39 #include <sys/stat.h>
40 
41 #ifdef PHP_WIN32
42 # include "win32/php_stdint.h"
43 #else
44 # if HAVE_INTTYPES_H
45 #  include <inttypes.h>
46 # elif HAVE_STDINT_H
47 #  include <stdint.h>
48 # endif
49 #endif
50 
51 /* some defines for the different JPEG block types */
52 #define M_SOF0  0xC0            /* Start Of Frame N */
53 #define M_SOF1  0xC1            /* N indicates which compression process */
54 #define M_SOF2  0xC2            /* Only SOF0-SOF2 are now in common use */
55 #define M_SOF3  0xC3
56 #define M_SOF5  0xC5            /* NB: codes C4 and CC are NOT SOF markers */
57 #define M_SOF6  0xC6
58 #define M_SOF7  0xC7
59 #define M_SOF9  0xC9
60 #define M_SOF10 0xCA
61 #define M_SOF11 0xCB
62 #define M_SOF13 0xCD
63 #define M_SOF14 0xCE
64 #define M_SOF15 0xCF
65 #define M_SOI   0xD8
66 #define M_EOI   0xD9            /* End Of Image (end of datastream) */
67 #define M_SOS   0xDA            /* Start Of Scan (begins compressed data) */
68 #define M_APP0  0xe0
69 #define M_APP1  0xe1
70 #define M_APP2  0xe2
71 #define M_APP3  0xe3
72 #define M_APP4  0xe4
73 #define M_APP5  0xe5
74 #define M_APP6  0xe6
75 #define M_APP7  0xe7
76 #define M_APP8  0xe8
77 #define M_APP9  0xe9
78 #define M_APP10 0xea
79 #define M_APP11 0xeb
80 #define M_APP12 0xec
81 #define M_APP13 0xed
82 #define M_APP14 0xee
83 #define M_APP15 0xef
84 
85 /* {{{ php_iptc_put1
86  */
php_iptc_put1(FILE * fp,int spool,unsigned char c,unsigned char ** spoolbuf TSRMLS_DC)87 static int php_iptc_put1(FILE *fp, int spool, unsigned char c, unsigned char **spoolbuf TSRMLS_DC)
88 {
89 	if (spool > 0)
90 		PUTC(c);
91 
92 	if (spoolbuf) *(*spoolbuf)++ = c;
93 
94   	return c;
95 }
96 /* }}} */
97 
98 /* {{{ php_iptc_get1
99  */
php_iptc_get1(FILE * fp,int spool,unsigned char ** spoolbuf TSRMLS_DC)100 static int php_iptc_get1(FILE *fp, int spool, unsigned char **spoolbuf TSRMLS_DC)
101 {
102 	int c;
103 	char cc;
104 
105 	c = getc(fp);
106 
107 	if (c == EOF) return EOF;
108 
109 	if (spool > 0) {
110 		cc = c;
111 		PUTC(cc);
112 	}
113 
114 	if (spoolbuf) *(*spoolbuf)++ = c;
115 
116 	return c;
117 }
118 /* }}} */
119 
120 /* {{{ php_iptc_read_remaining
121  */
php_iptc_read_remaining(FILE * fp,int spool,unsigned char ** spoolbuf TSRMLS_DC)122 static int php_iptc_read_remaining(FILE *fp, int spool, unsigned char **spoolbuf TSRMLS_DC)
123 {
124   	while (php_iptc_get1(fp, spool, spoolbuf TSRMLS_CC) != EOF) continue;
125 
126 	return M_EOI;
127 }
128 /* }}} */
129 
130 /* {{{ php_iptc_skip_variable
131  */
php_iptc_skip_variable(FILE * fp,int spool,unsigned char ** spoolbuf TSRMLS_DC)132 static int php_iptc_skip_variable(FILE *fp, int spool, unsigned char **spoolbuf TSRMLS_DC)
133 {
134 	unsigned int  length;
135 	int c1, c2;
136 
137     if ((c1 = php_iptc_get1(fp, spool, spoolbuf TSRMLS_CC)) == EOF) return M_EOI;
138 
139     if ((c2 = php_iptc_get1(fp, spool, spoolbuf TSRMLS_CC)) == EOF) return M_EOI;
140 
141 	length = (((unsigned char) c1) << 8) + ((unsigned char) c2);
142 
143 	length -= 2;
144 
145 	while (length--)
146 		if (php_iptc_get1(fp, spool, spoolbuf TSRMLS_CC) == EOF) return M_EOI;
147 
148 	return 0;
149 }
150 /* }}} */
151 
152 /* {{{ php_iptc_next_marker
153  */
php_iptc_next_marker(FILE * fp,int spool,unsigned char ** spoolbuf TSRMLS_DC)154 static int php_iptc_next_marker(FILE *fp, int spool, unsigned char **spoolbuf TSRMLS_DC)
155 {
156     int c;
157 
158     /* skip unimportant stuff */
159 
160     c = php_iptc_get1(fp, spool, spoolbuf TSRMLS_CC);
161 
162 	if (c == EOF) return M_EOI;
163 
164     while (c != 0xff) {
165         if ((c = php_iptc_get1(fp, spool, spoolbuf TSRMLS_CC)) == EOF)
166             return M_EOI; /* we hit EOF */
167     }
168 
169     /* get marker byte, swallowing possible padding */
170     do {
171         c = php_iptc_get1(fp, 0, 0 TSRMLS_CC);
172 		if (c == EOF)
173             return M_EOI;       /* we hit EOF */
174 		else
175 		if (c == 0xff)
176 			php_iptc_put1(fp, spool, (unsigned char)c, spoolbuf TSRMLS_CC);
177     } while (c == 0xff);
178 
179     return (unsigned int) c;
180 }
181 /* }}} */
182 
183 static char psheader[] = "\xFF\xED\0\0Photoshop 3.0\08BIM\x04\x04\0\0\0\0";
184 
185 /* {{{ proto array iptcembed(string iptcdata, string jpeg_file_name [, int spool])
186    Embed binary IPTC data into a JPEG image. */
PHP_FUNCTION(iptcembed)187 PHP_FUNCTION(iptcembed)
188 {
189 	char *iptcdata, *jpeg_file;
190 	int iptcdata_len, jpeg_file_len;
191 	long spool = 0;
192 	FILE *fp;
193 	unsigned int marker, done = 0;
194 	int inx;
195 	unsigned char *spoolbuf = NULL, *poi = NULL;
196 	struct stat sb;
197 	zend_bool written = 0;
198 
199 	if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "sp|l", &iptcdata, &iptcdata_len, &jpeg_file, &jpeg_file_len, &spool) != SUCCESS) {
200 		return;
201 	}
202 
203 	if (php_check_open_basedir(jpeg_file TSRMLS_CC)) {
204 		RETURN_FALSE;
205 	}
206 
207 	if ((size_t)iptcdata_len >= SIZE_MAX - sizeof(psheader) - 1025) {
208 		php_error_docref(NULL TSRMLS_CC, E_WARNING, "IPTC data too large");
209 		RETURN_FALSE;
210 	}
211 
212 	if ((fp = VCWD_FOPEN(jpeg_file, "rb")) == 0) {
213 		php_error_docref(NULL TSRMLS_CC, E_WARNING, "Unable to open %s", jpeg_file);
214 		RETURN_FALSE;
215 	}
216 
217 	if (spool < 2) {
218 		fstat(fileno(fp), &sb);
219 
220 		poi = spoolbuf = safe_emalloc(1, (size_t)iptcdata_len + sizeof(psheader) + 1024 + 1, sb.st_size);
221 		memset(poi, 0, iptcdata_len + sizeof(psheader) + sb.st_size + 1024 + 1);
222 	}
223 
224 	if (php_iptc_get1(fp, spool, poi?&poi:0 TSRMLS_CC) != 0xFF) {
225 		fclose(fp);
226 		if (spoolbuf) {
227 			efree(spoolbuf);
228 		}
229 		RETURN_FALSE;
230 	}
231 
232 	if (php_iptc_get1(fp, spool, poi?&poi:0 TSRMLS_CC) != 0xD8) {
233 		fclose(fp);
234 		if (spoolbuf) {
235 			efree(spoolbuf);
236 		}
237 		RETURN_FALSE;
238 	}
239 
240 	while (!done) {
241 		marker = php_iptc_next_marker(fp, spool, poi?&poi:0 TSRMLS_CC);
242 
243 		if (marker == M_EOI) { /* EOF */
244 			break;
245 		} else if (marker != M_APP13) {
246 			php_iptc_put1(fp, spool, (unsigned char)marker, poi?&poi:0 TSRMLS_CC);
247 		}
248 
249 		switch (marker) {
250 			case M_APP13:
251 				/* we are going to write a new APP13 marker, so don't output the old one */
252 				php_iptc_skip_variable(fp, 0, 0 TSRMLS_CC);
253 				php_iptc_read_remaining(fp, spool, poi?&poi:0 TSRMLS_CC);
254 				done = 1;
255 				break;
256 
257 			case M_APP0:
258 				/* APP0 is in each and every JPEG, so when we hit APP0 we insert our new APP13! */
259 			case M_APP1:
260 				if (written) {
261 					/* don't try to write the data twice */
262 					break;
263 				}
264 				written = 1;
265 
266 				php_iptc_skip_variable(fp, spool, poi?&poi:0 TSRMLS_CC);
267 
268 				if (iptcdata_len & 1) {
269 					iptcdata_len++; /* make the length even */
270 				}
271 
272 				psheader[ 2 ] = (iptcdata_len+28)>>8;
273 				psheader[ 3 ] = (iptcdata_len+28)&0xff;
274 
275 				for (inx = 0; inx < 28; inx++) {
276 					php_iptc_put1(fp, spool, psheader[inx], poi?&poi:0 TSRMLS_CC);
277 				}
278 
279 				php_iptc_put1(fp, spool, (unsigned char)(iptcdata_len>>8), poi?&poi:0 TSRMLS_CC);
280 				php_iptc_put1(fp, spool, (unsigned char)(iptcdata_len&0xff), poi?&poi:0 TSRMLS_CC);
281 
282 				for (inx = 0; inx < iptcdata_len; inx++) {
283 					php_iptc_put1(fp, spool, iptcdata[inx], poi?&poi:0 TSRMLS_CC);
284 				}
285 				break;
286 
287 			case M_SOS:
288 				/* we hit data, no more marker-inserting can be done! */
289 				php_iptc_read_remaining(fp, spool, poi?&poi:0 TSRMLS_CC);
290 				done = 1;
291 				break;
292 
293 			default:
294 				php_iptc_skip_variable(fp, spool, poi?&poi:0 TSRMLS_CC);
295 				break;
296 		}
297 	}
298 
299 	fclose(fp);
300 
301 	if (spool < 2) {
302 		RETVAL_STRINGL(spoolbuf, poi - spoolbuf, 0);
303 	} else {
304 		RETURN_TRUE;
305 	}
306 }
307 /* }}} */
308 
309 /* {{{ proto array iptcparse(string iptcdata)
310    Parse binary IPTC-data into associative array */
PHP_FUNCTION(iptcparse)311 PHP_FUNCTION(iptcparse)
312 {
313 	int inx = 0, len;
314 	unsigned int tagsfound = 0;
315 	unsigned char *buffer, recnum, dataset, key[ 16 ];
316 	char *str;
317 	int str_len;
318 	zval *values, **element;
319 
320 	if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "s", &str, &str_len) != SUCCESS) {
321 		return;
322 	}
323 
324 	buffer = (unsigned char *)str;
325 
326 	while (inx < str_len) { /* find 1st tag */
327 		if ((buffer[inx] == 0x1c) && ((buffer[inx+1] == 0x01) || (buffer[inx+1] == 0x02))){
328 			break;
329 		} else {
330 			inx++;
331 		}
332 	}
333 
334 	while (inx < str_len) {
335 		if (buffer[ inx++ ] != 0x1c) {
336 			break;   /* we ran against some data which does not conform to IPTC - stop parsing! */
337 		}
338 
339 		if ((inx + 4) >= str_len)
340 			break;
341 
342 		dataset = buffer[ inx++ ];
343 		recnum = buffer[ inx++ ];
344 
345 		if (buffer[ inx ] & (unsigned char) 0x80) { /* long tag */
346 			if((inx+6) >= str_len) {
347 				break;
348 			}
349 			len = (((long) buffer[ inx + 2 ]) << 24) + (((long) buffer[ inx + 3 ]) << 16) +
350 				  (((long) buffer[ inx + 4 ]) <<  8) + (((long) buffer[ inx + 5 ]));
351 			inx += 6;
352 		} else { /* short tag */
353 			len = (((unsigned short) buffer[ inx ])<<8) | (unsigned short)buffer[ inx+1 ];
354 			inx += 2;
355 		}
356 
357 		if ((len < 0) || (len > str_len) || (inx + len) > str_len) {
358 			break;
359 		}
360 
361 		snprintf(key, sizeof(key), "%d#%03d", (unsigned int) dataset, (unsigned int) recnum);
362 
363 		if (tagsfound == 0) { /* found the 1st tag - initialize the return array */
364 			array_init(return_value);
365 		}
366 
367 		if (zend_hash_find(Z_ARRVAL_P(return_value), key, strlen(key) + 1, (void **) &element) == FAILURE) {
368 			MAKE_STD_ZVAL(values);
369 			array_init(values);
370 
371 			zend_hash_update(Z_ARRVAL_P(return_value), key, strlen(key) + 1, (void *) &values, sizeof(zval*), (void **) &element);
372 		}
373 
374 		add_next_index_stringl(*element, buffer+inx, len, 1);
375 		inx += len;
376 		tagsfound++;
377 	}
378 
379 	if (! tagsfound) {
380 		RETURN_FALSE;
381 	}
382 }
383 /* }}} */
384 
385 /*
386  * Local variables:
387  * tab-width: 4
388  * c-basic-offset: 4
389  * End:
390  * vim600: sw=4 ts=4 fdm=marker
391  * vim<600: sw=4 ts=4
392  */
393