]> git.saurik.com Git - apple/xnu.git/blob - libkern/c++/OSUnserializeXML.y
xnu-7195.50.7.100.1.tar.gz
[apple/xnu.git] / libkern / c++ / OSUnserializeXML.y
1 /*
2 * Copyright (c) 1999-2019 Apple Inc. All rights reserved.
3 *
4 * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
5 *
6 * This file contains Original Code and/or Modifications of Original Code
7 * as defined in and that are subject to the Apple Public Source License
8 * Version 2.0 (the 'License'). You may not use this file except in
9 * compliance with the License. The rights granted to you under the License
10 * may not be used to create, or enable the creation or redistribution of,
11 * unlawful or unlicensed copies of an Apple operating system, or to
12 * circumvent, violate, or enable the circumvention or violation of, any
13 * terms of an Apple operating system software license agreement.
14 *
15 * Please obtain a copy of the License at
16 * http://www.opensource.apple.com/apsl/ and read it before using this file.
17 *
18 * The Original Code and all software distributed under the License are
19 * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
20 * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
21 * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
22 * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
23 * Please see the License for the specific language governing rights and
24 * limitations under the License.
25 *
26 * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
27 */
28
29 /*
30 * HISTORY
31 *
32 * OSUnserializeXML.y created by rsulack on Tue Oct 12 1999
33 */
34
35 // parser for unserializing OSContainer objects serialized to XML
36 //
37 // to build :
38 // bison -p OSUnserializeXML OSUnserializeXML.y
39 // head -50 OSUnserializeXML.y > OSUnserializeXML.cpp
40 // sed -e "s/#include <stdio.h>//" < OSUnserializeXML.tab.c >> OSUnserializeXML.cpp
41 //
42 // when changing code check in both OSUnserializeXML.y and OSUnserializeXML.cpp
43 //
44 //
45 //
46 //
47 //
48 // DO NOT EDIT OSUnserializeXML.cpp!
49 //
50 // this means you!
51 //
52 //
53 //
54 //
55 //
56 //
57
58
59 %pure_parser
60
61 %{
62 #include <string.h>
63 #include <libkern/c++/OSMetaClass.h>
64 #include <libkern/c++/OSContainers.h>
65 #include <libkern/c++/OSLib.h>
66
67 #define MAX_OBJECTS 131071
68 #define MAX_REFED_OBJECTS 65535
69
70 #define YYSTYPE object_t *
71 #define YYPARSE_PARAM state
72 #define YYLEX_PARAM (parser_state_t *)state
73
74 // this is the internal struct used to hold objects on parser stack
75 // it represents objects both before and after they have been created
76 typedef struct object {
77 struct object *next;
78 struct object *free;
79 struct object *elements;
80 OSObject *object;
81 OSSymbol *key; // for dictionary
82 int size;
83 void *data; // for data
84 char *string; // for string & symbol
85 long long number; // for number
86 int idref;
87 } object_t;
88
89 // this code is reentrant, this structure contains all
90 // state information for the parsing of a single buffer
91 typedef struct parser_state {
92 const char *parseBuffer; // start of text to be parsed
93 int parseBufferIndex; // current index into text
94 int lineNumber; // current line number
95 object_t *objects; // internal objects in use
96 object_t *freeObjects; // internal objects that are free
97 OSDictionary *tags; // used to remember "ID" tags
98 OSString **errorString; // parse error with line
99 OSObject *parsedObject; // resultant object of parsed text
100 int parsedObjectCount;
101 int retrievedObjectCount;
102 } parser_state_t;
103
104 #define STATE ((parser_state_t *)state)
105
106 #undef yyerror
107 #define yyerror(s) OSUnserializeerror(STATE, (s))
108 static int OSUnserializeerror(parser_state_t *state, const char *s);
109
110 static int yylex(YYSTYPE *lvalp, parser_state_t *state);
111
112 static object_t *newObject(parser_state_t *state);
113 static void freeObject(parser_state_t *state, object_t *o);
114 static void rememberObject(parser_state_t *state, int tag, OSObject *o);
115 static object_t *retrieveObject(parser_state_t *state, int tag);
116 static void cleanupObjects(parser_state_t *state);
117
118 static object_t *buildDictionary(parser_state_t *state, object_t *o);
119 static object_t *buildArray(parser_state_t *state, object_t *o);
120 static object_t *buildSet(parser_state_t *state, object_t *o);
121 static object_t *buildString(parser_state_t *state, object_t *o);
122 static object_t *buildSymbol(parser_state_t *state, object_t *o);
123 static object_t *buildData(parser_state_t *state, object_t *o);
124 static object_t *buildNumber(parser_state_t *state, object_t *o);
125 static object_t *buildBoolean(parser_state_t *state, object_t *o);
126
127 __BEGIN_DECLS
128 #include <kern/kalloc.h>
129 __END_DECLS
130
131 #define malloc(size) malloc_impl(size)
132 static inline void *
133 malloc_impl(size_t size)
134 {
135 if (size == 0) {
136 return NULL;
137 }
138 return kheap_alloc_tag_bt(KHEAP_DEFAULT, size,
139 (zalloc_flags_t) (Z_WAITOK | Z_ZERO),
140 VM_KERN_MEMORY_LIBKERN);
141 }
142
143 #define free(addr) free_impl(addr)
144 static inline void
145 free_impl(void *addr)
146 {
147 kheap_free_addr(KHEAP_DEFAULT, addr);
148 }
149 static inline void
150 safe_free(void *addr, size_t size)
151 {
152 if(addr) {
153 assert(size != 0);
154 kheap_free(KHEAP_DEFAULT, addr, size);
155 }
156 }
157
158 #define realloc(addr, osize, nsize) realloc_impl(addr, osize, nsize)
159 static inline void *
160 realloc_impl(void *addr, size_t osize, size_t nsize)
161 {
162 if (!addr) {
163 return malloc(nsize);
164 }
165 if (nsize == osize) {
166 return addr;
167 }
168 void *nmem = malloc(nsize);
169 if (!nmem) {
170 safe_free(addr, osize);
171 return NULL;
172 }
173 (void)memcpy(nmem, addr, (nsize > osize) ? osize : nsize);
174 safe_free(addr, osize);
175
176 return nmem;
177 }
178
179 %}
180 %token ARRAY
181 %token BOOLEAN
182 %token DATA
183 %token DICTIONARY
184 %token IDREF
185 %token KEY
186 %token NUMBER
187 %token SET
188 %token STRING
189 %token SYNTAX_ERROR
190 %% /* Grammar rules and actions follow */
191
192 input: /* empty */ { yyerror("unexpected end of buffer");
193 YYERROR;
194 }
195 | object { STATE->parsedObject = $1->object;
196 $1->object = 0;
197 freeObject(STATE, $1);
198 YYACCEPT;
199 }
200 | SYNTAX_ERROR { yyerror("syntax error");
201 YYERROR;
202 }
203 ;
204
205 object: dict { $$ = buildDictionary(STATE, $1);
206
207 if (!yyval->object) {
208 yyerror("buildDictionary");
209 YYERROR;
210 }
211 STATE->parsedObjectCount++;
212 if (STATE->parsedObjectCount > MAX_OBJECTS) {
213 yyerror("maximum object count");
214 YYERROR;
215 }
216 }
217 | array { $$ = buildArray(STATE, $1);
218
219 if (!yyval->object) {
220 yyerror("buildArray");
221 YYERROR;
222 }
223 STATE->parsedObjectCount++;
224 if (STATE->parsedObjectCount > MAX_OBJECTS) {
225 yyerror("maximum object count");
226 YYERROR;
227 }
228 }
229 | set { $$ = buildSet(STATE, $1);
230
231 if (!yyval->object) {
232 yyerror("buildSet");
233 YYERROR;
234 }
235 STATE->parsedObjectCount++;
236 if (STATE->parsedObjectCount > MAX_OBJECTS) {
237 yyerror("maximum object count");
238 YYERROR;
239 }
240 }
241 | string { $$ = buildString(STATE, $1);
242
243 if (!yyval->object) {
244 yyerror("buildString");
245 YYERROR;
246 }
247 STATE->parsedObjectCount++;
248 if (STATE->parsedObjectCount > MAX_OBJECTS) {
249 yyerror("maximum object count");
250 YYERROR;
251 }
252 }
253 | data { $$ = buildData(STATE, $1);
254
255 if (!yyval->object) {
256 yyerror("buildData");
257 YYERROR;
258 }
259 STATE->parsedObjectCount++;
260 if (STATE->parsedObjectCount > MAX_OBJECTS) {
261 yyerror("maximum object count");
262 YYERROR;
263 }
264 }
265 | number { $$ = buildNumber(STATE, $1);
266
267 if (!yyval->object) {
268 yyerror("buildNumber");
269 YYERROR;
270 }
271 STATE->parsedObjectCount++;
272 if (STATE->parsedObjectCount > MAX_OBJECTS) {
273 yyerror("maximum object count");
274 YYERROR;
275 }
276 }
277 | boolean { $$ = buildBoolean(STATE, $1);
278
279 if (!yyval->object) {
280 yyerror("buildBoolean");
281 YYERROR;
282 }
283 STATE->parsedObjectCount++;
284 if (STATE->parsedObjectCount > MAX_OBJECTS) {
285 yyerror("maximum object count");
286 YYERROR;
287 }
288 }
289 | idref { $$ = retrieveObject(STATE, $1->idref);
290 if ($$) {
291 STATE->retrievedObjectCount++;
292 $$->object->retain();
293 if (STATE->retrievedObjectCount > MAX_REFED_OBJECTS) {
294 yyerror("maximum object reference count");
295 YYERROR;
296 }
297 } else {
298 yyerror("forward reference detected");
299 YYERROR;
300 }
301 freeObject(STATE, $1);
302
303 STATE->parsedObjectCount++;
304 if (STATE->parsedObjectCount > MAX_OBJECTS) {
305 yyerror("maximum object count");
306 YYERROR;
307 }
308 }
309 ;
310
311 //------------------------------------------------------------------------------
312
313 dict: '{' '}' { $$ = $1;
314 $$->elements = NULL;
315 }
316 | '{' pairs '}' { $$ = $1;
317 $$->elements = $2;
318 }
319 | DICTIONARY
320 ;
321
322 pairs: pair
323 | pairs pair { $$ = $2;
324 $$->next = $1;
325
326 object_t *o;
327 o = $$->next;
328 while (o) {
329 if (o->key == $$->key) {
330 yyerror("duplicate dictionary key");
331 YYERROR;
332 }
333 o = o->next;
334 }
335 }
336 ;
337
338 pair: key object { $$ = $1;
339 $$->key = (OSSymbol *)$$->object;
340 $$->object = $2->object;
341 $$->next = NULL;
342 $2->object = 0;
343 freeObject(STATE, $2);
344 }
345 ;
346
347 key: KEY { $$ = buildSymbol(STATE, $1);
348
349 // STATE->parsedObjectCount++;
350 // if (STATE->parsedObjectCount > MAX_OBJECTS) {
351 // yyerror("maximum object count");
352 // YYERROR;
353 // }
354 }
355 ;
356
357 //------------------------------------------------------------------------------
358
359 array: '(' ')' { $$ = $1;
360 $$->elements = NULL;
361 }
362 | '(' elements ')' { $$ = $1;
363 $$->elements = $2;
364 }
365 | ARRAY
366 ;
367
368 set: '[' ']' { $$ = $1;
369 $$->elements = NULL;
370 }
371 | '[' elements ']' { $$ = $1;
372 $$->elements = $2;
373 }
374 | SET
375 ;
376
377 elements: object { $$ = $1;
378 $$->next = NULL;
379 }
380 | elements object { $$ = $2;
381 $$->next = $1;
382 }
383 ;
384
385 //------------------------------------------------------------------------------
386
387 boolean: BOOLEAN
388 ;
389
390 data: DATA
391 ;
392
393 idref: IDREF
394 ;
395
396 number: NUMBER
397 ;
398
399 string: STRING
400 ;
401
402 %%
403
404 int
405 OSUnserializeerror(parser_state_t * state, const char *s) /* Called by yyparse on errors */
406 {
407 if (state->errorString) {
408 char tempString[128];
409 snprintf(tempString, 128, "OSUnserializeXML: %s near line %d\n", s, state->lineNumber);
410 *(state->errorString) = OSString::withCString(tempString);
411 }
412
413 return 0;
414 }
415
416 #define TAG_MAX_LENGTH 32
417 #define TAG_MAX_ATTRIBUTES 32
418 #define TAG_BAD 0
419 #define TAG_START 1
420 #define TAG_END 2
421 #define TAG_EMPTY 3
422 #define TAG_IGNORE 4
423
424 #define currentChar() (state->parseBuffer[state->parseBufferIndex])
425 #define nextChar() (state->parseBuffer[++state->parseBufferIndex])
426 #define prevChar() (state->parseBuffer[state->parseBufferIndex - 1])
427
428 #define isSpace(c) ((c) == ' ' || (c) == '\t')
429 #define isAlpha(c) (((c) >= 'A' && (c) <= 'Z') || ((c) >= 'a' && (c) <= 'z'))
430 #define isDigit(c) ((c) >= '0' && (c) <= '9')
431 #define isAlphaDigit(c) ((c) >= 'a' && (c) <= 'f')
432 #define isHexDigit(c) (isDigit(c) || isAlphaDigit(c))
433 #define isAlphaNumeric(c) (isAlpha(c) || isDigit(c) || ((c) == '-'))
434
435 static int
436 getTag(parser_state_t *state,
437 char tag[TAG_MAX_LENGTH],
438 int *attributeCount,
439 char attributes[TAG_MAX_ATTRIBUTES][TAG_MAX_LENGTH],
440 char values[TAG_MAX_ATTRIBUTES][TAG_MAX_LENGTH] )
441 {
442 int length = 0;
443 int c = currentChar();
444 int tagType = TAG_START;
445
446 *attributeCount = 0;
447
448 if (c != '<') {
449 return TAG_BAD;
450 }
451 c = nextChar(); // skip '<'
452
453
454 // <!TAG declarations >
455 // <!-- comments -->
456 if (c == '!') {
457 c = nextChar();
458 bool isComment = (c == '-') && ((c = nextChar()) != 0) && (c == '-');
459 if (!isComment && !isAlpha(c)) {
460 return TAG_BAD; // <!1, <!-A, <!eos
461 }
462 while (c && (c = nextChar()) != 0) {
463 if (c == '\n') {
464 state->lineNumber++;
465 }
466 if (isComment) {
467 if (c != '-') {
468 continue;
469 }
470 c = nextChar();
471 if (c != '-') {
472 continue;
473 }
474 c = nextChar();
475 }
476 if (c == '>') {
477 (void)nextChar();
478 return TAG_IGNORE;
479 }
480 if (isComment) {
481 break;
482 }
483 }
484 return TAG_BAD;
485 } else
486 // <? Processing Instructions ?>
487 if (c == '?') {
488 while ((c = nextChar()) != 0) {
489 if (c == '\n') {
490 state->lineNumber++;
491 }
492 if (c != '?') {
493 continue;
494 }
495 c = nextChar();
496 if (!c) {
497 return TAG_IGNORE;
498 }
499 if (c == '>') {
500 (void)nextChar();
501 return TAG_IGNORE;
502 }
503 }
504 return TAG_BAD;
505 } else
506 // </ end tag >
507 if (c == '/') {
508 c = nextChar(); // skip '/'
509 tagType = TAG_END;
510 }
511 if (!isAlpha(c)) {
512 return TAG_BAD;
513 }
514
515 /* find end of tag while copying it */
516 while (isAlphaNumeric(c)) {
517 tag[length++] = c;
518 c = nextChar();
519 if (length >= (TAG_MAX_LENGTH - 1)) {
520 return TAG_BAD;
521 }
522 }
523
524 tag[length] = 0;
525
526 // printf("tag %s, type %d\n", tag, tagType);
527
528 // look for attributes of the form attribute = "value" ...
529 while ((c != '>') && (c != '/')) {
530 while (isSpace(c)) {
531 c = nextChar();
532 }
533
534 length = 0;
535 while (isAlphaNumeric(c)) {
536 attributes[*attributeCount][length++] = c;
537 if (length >= (TAG_MAX_LENGTH - 1)) {
538 return TAG_BAD;
539 }
540 c = nextChar();
541 }
542 attributes[*attributeCount][length] = 0;
543
544 while (isSpace(c)) {
545 c = nextChar();
546 }
547
548 if (c != '=') {
549 return TAG_BAD;
550 }
551 c = nextChar();
552
553 while (isSpace(c)) {
554 c = nextChar();
555 }
556
557 if (c != '"') {
558 return TAG_BAD;
559 }
560 c = nextChar();
561 length = 0;
562 while (c != '"') {
563 values[*attributeCount][length++] = c;
564 if (length >= (TAG_MAX_LENGTH - 1)) {
565 return TAG_BAD;
566 }
567 c = nextChar();
568 if (!c) {
569 return TAG_BAD;
570 }
571 }
572 values[*attributeCount][length] = 0;
573
574 c = nextChar(); // skip closing quote
575
576 // printf(" attribute '%s' = '%s', nextchar = '%c'\n",
577 // attributes[*attributeCount], values[*attributeCount], c);
578
579 (*attributeCount)++;
580 if (*attributeCount >= TAG_MAX_ATTRIBUTES) {
581 return TAG_BAD;
582 }
583 }
584
585 if (c == '/') {
586 c = nextChar(); // skip '/'
587 tagType = TAG_EMPTY;
588 }
589 if (c != '>') {
590 return TAG_BAD;
591 }
592 c = nextChar(); // skip '>'
593
594 return tagType;
595 }
596
597 static char *
598 getString(parser_state_t *state, int *alloc_lengthp)
599 {
600 int c = currentChar();
601 int start, length, i, j;
602 char * tempString;
603
604 start = state->parseBufferIndex;
605 /* find end of string */
606
607 while (c != 0) {
608 if (c == '\n') {
609 state->lineNumber++;
610 }
611 if (c == '<') {
612 break;
613 }
614 c = nextChar();
615 }
616
617 if (c != '<') {
618 return 0;
619 }
620
621 length = state->parseBufferIndex - start;
622
623 /* copy to null terminated buffer */
624 tempString = (char *)malloc(length + 1);
625 if (tempString == NULL) {
626 printf("OSUnserializeXML: can't alloc temp memory\n");
627 goto error;
628 }
629 if (alloc_lengthp != NULL) {
630 *alloc_lengthp = length + 1;
631 }
632
633 // copy out string in tempString
634 // "&amp;" -> '&', "&lt;" -> '<', "&gt;" -> '>'
635
636 i = j = 0;
637 while (i < length) {
638 c = state->parseBuffer[start + i++];
639 if (c != '&') {
640 tempString[j++] = c;
641 } else {
642 if ((i + 3) > length) {
643 goto error;
644 }
645 c = state->parseBuffer[start + i++];
646 if (c == 'l') {
647 if (state->parseBuffer[start + i++] != 't') {
648 goto error;
649 }
650 if (state->parseBuffer[start + i++] != ';') {
651 goto error;
652 }
653 tempString[j++] = '<';
654 continue;
655 }
656 if (c == 'g') {
657 if (state->parseBuffer[start + i++] != 't') {
658 goto error;
659 }
660 if (state->parseBuffer[start + i++] != ';') {
661 goto error;
662 }
663 tempString[j++] = '>';
664 continue;
665 }
666 if ((i + 3) > length) {
667 goto error;
668 }
669 if (c == 'a') {
670 if (state->parseBuffer[start + i++] != 'm') {
671 goto error;
672 }
673 if (state->parseBuffer[start + i++] != 'p') {
674 goto error;
675 }
676 if (state->parseBuffer[start + i++] != ';') {
677 goto error;
678 }
679 tempString[j++] = '&';
680 continue;
681 }
682 goto error;
683 }
684 }
685 tempString[j] = 0;
686
687 // printf("string %s\n", tempString);
688
689 return tempString;
690
691 error:
692 if (tempString) {
693 safe_free(tempString, length + 1);
694 if (alloc_lengthp != NULL) {
695 *alloc_lengthp = 0;
696 }
697 }
698 return 0;
699 }
700
701 static long long
702 getNumber(parser_state_t *state)
703 {
704 unsigned long long n = 0;
705 int base = 10;
706 bool negate = false;
707 int c = currentChar();
708
709 if (c == '0') {
710 c = nextChar();
711 if (c == 'x') {
712 base = 16;
713 c = nextChar();
714 }
715 }
716 if (base == 10) {
717 if (c == '-') {
718 negate = true;
719 c = nextChar();
720 }
721 while (isDigit(c)) {
722 n = (n * base + c - '0');
723 c = nextChar();
724 }
725 if (negate) {
726 n = (unsigned long long)((long long)n * (long long)-1);
727 }
728 } else {
729 while (isHexDigit(c)) {
730 if (isDigit(c)) {
731 n = (n * base + c - '0');
732 } else {
733 n = (n * base + 0xa + c - 'a');
734 }
735 c = nextChar();
736 }
737 }
738 // printf("number 0x%x\n", (unsigned long)n);
739 return n;
740 }
741
742 // taken from CFXMLParsing/CFPropertyList.c
743
744 static const signed char __CFPLDataDecodeTable[128] = {
745 /* 000 */ -1, -1, -1, -1, -1, -1, -1, -1,
746 /* 010 */ -1, -1, -1, -1, -1, -1, -1, -1,
747 /* 020 */ -1, -1, -1, -1, -1, -1, -1, -1,
748 /* 030 */ -1, -1, -1, -1, -1, -1, -1, -1,
749 /* ' ' */ -1, -1, -1, -1, -1, -1, -1, -1,
750 /* '(' */ -1, -1, -1, 62, -1, -1, -1, 63,
751 /* '0' */ 52, 53, 54, 55, 56, 57, 58, 59,
752 /* '8' */ 60, 61, -1, -1, -1, 0, -1, -1,
753 /* '@' */ -1, 0, 1, 2, 3, 4, 5, 6,
754 /* 'H' */ 7, 8, 9, 10, 11, 12, 13, 14,
755 /* 'P' */ 15, 16, 17, 18, 19, 20, 21, 22,
756 /* 'X' */ 23, 24, 25, -1, -1, -1, -1, -1,
757 /* '`' */ -1, 26, 27, 28, 29, 30, 31, 32,
758 /* 'h' */ 33, 34, 35, 36, 37, 38, 39, 40,
759 /* 'p' */ 41, 42, 43, 44, 45, 46, 47, 48,
760 /* 'x' */ 49, 50, 51, -1, -1, -1, -1, -1
761 };
762
763 #define DATA_ALLOC_SIZE 4096
764
765 static void *
766 getCFEncodedData(parser_state_t *state, unsigned int *size)
767 {
768 int numeq = 0, cntr = 0;
769 unsigned int acc = 0;
770 int tmpbufpos = 0;
771 size_t tmpbuflen = DATA_ALLOC_SIZE;
772 unsigned char *tmpbuf = (unsigned char *)malloc(tmpbuflen);
773
774 int c = currentChar();
775 *size = 0;
776
777 while (c != '<') {
778 c &= 0x7f;
779 if (c == 0) {
780 safe_free(tmpbuf, tmpbuflen);
781 return 0;
782 }
783 if (c == '=') {
784 numeq++;
785 } else {
786 numeq = 0;
787 }
788 if (c == '\n') {
789 state->lineNumber++;
790 }
791 if (__CFPLDataDecodeTable[c] < 0) {
792 c = nextChar();
793 continue;
794 }
795 cntr++;
796 acc <<= 6;
797 acc += __CFPLDataDecodeTable[c];
798 if (0 == (cntr & 0x3)) {
799 if (tmpbuflen <= tmpbufpos + 2) {
800 size_t oldsize = tmpbuflen;
801 tmpbuflen += DATA_ALLOC_SIZE;
802 tmpbuf = (unsigned char *)realloc(tmpbuf, oldsize, tmpbuflen);
803 }
804 tmpbuf[tmpbufpos++] = (acc >> 16) & 0xff;
805 if (numeq < 2) {
806 tmpbuf[tmpbufpos++] = (acc >> 8) & 0xff;
807 }
808 if (numeq < 1) {
809 tmpbuf[tmpbufpos++] = acc & 0xff;
810 }
811 }
812 c = nextChar();
813 }
814 *size = tmpbufpos;
815 if (*size == 0) {
816 safe_free(tmpbuf, tmpbuflen);
817 return 0;
818 }
819 return tmpbuf;
820 }
821
822 static void *
823 getHexData(parser_state_t *state, unsigned int *size)
824 {
825 int c;
826 unsigned char *d, *start, *lastStart;
827
828 size_t buflen = DATA_ALLOC_SIZE;
829 start = lastStart = d = (unsigned char *)malloc(buflen);
830 c = currentChar();
831
832 while (c != '<') {
833 if (isSpace(c)) {
834 while ((c = nextChar()) != 0 && isSpace(c)) {
835 }
836 }
837 ;
838 if (c == '\n') {
839 state->lineNumber++;
840 c = nextChar();
841 continue;
842 }
843
844 // get high nibble
845 if (isDigit(c)) {
846 *d = (c - '0') << 4;
847 } else if (isAlphaDigit(c)) {
848 *d = (0xa + (c - 'a')) << 4;
849 } else {
850 goto error;
851 }
852
853 // get low nibble
854 c = nextChar();
855 if (isDigit(c)) {
856 *d |= c - '0';
857 } else if (isAlphaDigit(c)) {
858 *d |= 0xa + (c - 'a');
859 } else {
860 goto error;
861 }
862
863 d++;
864 if ((d - lastStart) >= DATA_ALLOC_SIZE) {
865 int oldsize = d - start;
866 assert(oldsize == buflen);
867 buflen += DATA_ALLOC_SIZE;
868 start = (unsigned char *)realloc(start, oldsize, buflen);
869 d = lastStart = start + oldsize;
870 }
871 c = nextChar();
872 }
873
874 *size = d - start;
875 return start;
876
877 error:
878
879 *size = 0;
880 safe_free(start, buflen);
881 return 0;
882 }
883
884 static int
885 yylex(YYSTYPE *lvalp, parser_state_t *state)
886 {
887 int c, i;
888 int tagType;
889 char tag[TAG_MAX_LENGTH];
890 int attributeCount;
891 char attributes[TAG_MAX_ATTRIBUTES][TAG_MAX_LENGTH];
892 char values[TAG_MAX_ATTRIBUTES][TAG_MAX_LENGTH];
893 object_t *object;
894 int alloc_length;
895
896 top:
897 c = currentChar();
898
899 /* skip white space */
900 if (isSpace(c)) {
901 while ((c = nextChar()) != 0 && isSpace(c)) {
902 }
903 }
904 ;
905
906 /* keep track of line number, don't return \n's */
907 if (c == '\n') {
908 STATE->lineNumber++;
909 (void)nextChar();
910 goto top;
911 }
912
913 // end of the buffer?
914 if (!c) {
915 return 0;
916 }
917
918 tagType = getTag(STATE, tag, &attributeCount, attributes, values);
919 if (tagType == TAG_BAD) {
920 return SYNTAX_ERROR;
921 }
922 if (tagType == TAG_IGNORE) {
923 goto top;
924 }
925
926 // handle allocation and check for "ID" and "IDREF" tags up front
927 *lvalp = object = newObject(STATE);
928 object->idref = -1;
929 for (i = 0; i < attributeCount; i++) {
930 if (attributes[i][0] == 'I' && attributes[i][1] == 'D') {
931 // check for idref's, note: we ignore the tag, for
932 // this to work correctly, all idrefs must be unique
933 // across the whole serialization
934 if (attributes[i][2] == 'R' && attributes[i][3] == 'E' &&
935 attributes[i][4] == 'F' && !attributes[i][5]) {
936 if (tagType != TAG_EMPTY) {
937 return SYNTAX_ERROR;
938 }
939 object->idref = strtol(values[i], NULL, 0);
940 return IDREF;
941 }
942 // check for id's
943 if (!attributes[i][2]) {
944 object->idref = strtol(values[i], NULL, 0);
945 } else {
946 return SYNTAX_ERROR;
947 }
948 }
949 }
950
951 switch (*tag) {
952 case 'a':
953 if (!strcmp(tag, "array")) {
954 if (tagType == TAG_EMPTY) {
955 object->elements = NULL;
956 return ARRAY;
957 }
958 return (tagType == TAG_START) ? '(' : ')';
959 }
960 break;
961 case 'd':
962 if (!strcmp(tag, "dict")) {
963 if (tagType == TAG_EMPTY) {
964 object->elements = NULL;
965 return DICTIONARY;
966 }
967 return (tagType == TAG_START) ? '{' : '}';
968 }
969 if (!strcmp(tag, "data")) {
970 unsigned int size;
971 if (tagType == TAG_EMPTY) {
972 object->data = NULL;
973 object->size = 0;
974 return DATA;
975 }
976
977 bool isHexFormat = false;
978 for (i = 0; i < attributeCount; i++) {
979 if (!strcmp(attributes[i], "format") && !strcmp(values[i], "hex")) {
980 isHexFormat = true;
981 break;
982 }
983 }
984 // CF encoded is the default form
985 if (isHexFormat) {
986 object->data = getHexData(STATE, &size);
987 } else {
988 object->data = getCFEncodedData(STATE, &size);
989 }
990 object->size = size;
991 if ((getTag(STATE, tag, &attributeCount, attributes, values) != TAG_END) || strcmp(tag, "data")) {
992 return SYNTAX_ERROR;
993 }
994 return DATA;
995 }
996 break;
997 case 'f':
998 if (!strcmp(tag, "false")) {
999 if (tagType == TAG_EMPTY) {
1000 object->number = 0;
1001 return BOOLEAN;
1002 }
1003 }
1004 break;
1005 case 'i':
1006 if (!strcmp(tag, "integer")) {
1007 object->size = 64; // default
1008 for (i = 0; i < attributeCount; i++) {
1009 if (!strcmp(attributes[i], "size")) {
1010 object->size = strtoul(values[i], NULL, 0);
1011 }
1012 }
1013 if (tagType == TAG_EMPTY) {
1014 object->number = 0;
1015 return NUMBER;
1016 }
1017 object->number = getNumber(STATE);
1018 if ((getTag(STATE, tag, &attributeCount, attributes, values) != TAG_END) || strcmp(tag, "integer")) {
1019 return SYNTAX_ERROR;
1020 }
1021 return NUMBER;
1022 }
1023 break;
1024 case 'k':
1025 if (!strcmp(tag, "key")) {
1026 if (tagType == TAG_EMPTY) {
1027 return SYNTAX_ERROR;
1028 }
1029 object->string = getString(STATE, &alloc_length);
1030 if (!object->string) {
1031 return SYNTAX_ERROR;
1032 }
1033 object->string_alloc_length = alloc_length;
1034 if ((getTag(STATE, tag, &attributeCount, attributes, values) != TAG_END)
1035 || strcmp(tag, "key")) {
1036 return SYNTAX_ERROR;
1037 }
1038 return KEY;
1039 }
1040 break;
1041 case 'p':
1042 if (!strcmp(tag, "plist")) {
1043 freeObject(STATE, object);
1044 goto top;
1045 }
1046 break;
1047 case 's':
1048 if (!strcmp(tag, "string")) {
1049 if (tagType == TAG_EMPTY) {
1050 object->string = (char *)malloc(1);
1051 object->string[0] = 0;
1052 object->string_alloc_length = 1;
1053 return STRING;
1054 }
1055 object->string = getString(STATE, &alloc_length);
1056 if (!object->string) {
1057 return SYNTAX_ERROR;
1058 }
1059 object->string_alloc_length = alloc_length;
1060 if ((getTag(STATE, tag, &attributeCount, attributes, values) != TAG_END)
1061 || strcmp(tag, "string")) {
1062 return SYNTAX_ERROR;
1063 }
1064 return STRING;
1065 }
1066 if (!strcmp(tag, "set")) {
1067 if (tagType == TAG_EMPTY) {
1068 object->elements = NULL;
1069 return SET;;
1070 }
1071 if (tagType == TAG_START) {
1072 return '[';
1073 } else {
1074 return ']';
1075 }
1076 }
1077 break;
1078 case 't':
1079 if (!strcmp(tag, "true")) {
1080 if (tagType == TAG_EMPTY) {
1081 object->number = 1;
1082 return BOOLEAN;
1083 }
1084 }
1085 break;
1086 }
1087
1088 return SYNTAX_ERROR;
1089 }
1090
1091 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
1092 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
1093 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
1094
1095 // "java" like allocation, if this code hits a syntax error in the
1096 // the middle of the parsed string we just bail with pointers hanging
1097 // all over place, this code helps keeps it all together
1098
1099 //static int object_count = 0;
1100
1101 object_t *
1102 newObject(parser_state_t *state)
1103 {
1104 object_t *o;
1105
1106 if (state->freeObjects) {
1107 o = state->freeObjects;
1108 state->freeObjects = state->freeObjects->next;
1109 } else {
1110 o = (object_t *)malloc(sizeof(object_t));
1111 // object_count++;
1112 o->free = state->objects;
1113 state->objects = o;
1114 }
1115
1116 return o;
1117 }
1118
1119 void
1120 freeObject(parser_state_t * state, object_t *o)
1121 {
1122 o->next = state->freeObjects;
1123 state->freeObjects = o;
1124 }
1125
1126 void
1127 cleanupObjects(parser_state_t *state)
1128 {
1129 object_t *t, *o = state->objects;
1130
1131 while (o) {
1132 if (o->object) {
1133 // printf("OSUnserializeXML: releasing object o=%x object=%x\n", (int)o, (int)o->object);
1134 o->object->release();
1135 }
1136 if (o->data) {
1137 // printf("OSUnserializeXML: freeing object o=%x data=%x\n", (int)o, (int)o->data);
1138 free(o->data);
1139 }
1140 if (o->key) {
1141 // printf("OSUnserializeXML: releasing object o=%x key=%x\n", (int)o, (int)o->key);
1142 o->key->release();
1143 }
1144 if (o->string) {
1145 // printf("OSUnserializeXML: freeing object o=%x string=%x\n", (int)o, (int)o->string);
1146 free(o->string);
1147 }
1148
1149 t = o;
1150 o = o->free;
1151 safe_free(t, sizeof(object_t));
1152 // object_count--;
1153 }
1154 // printf("object_count = %d\n", object_count);
1155 }
1156
1157 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
1158 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
1159 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
1160
1161 static void
1162 rememberObject(parser_state_t *state, int tag, OSObject *o)
1163 {
1164 char key[16];
1165 snprintf(key, 16, "%u", tag);
1166
1167 // printf("remember key %s\n", key);
1168
1169 state->tags->setObject(key, o);
1170 }
1171
1172 static object_t *
1173 retrieveObject(parser_state_t *state, int tag)
1174 {
1175 OSObject *ref;
1176 object_t *o;
1177 char key[16];
1178 snprintf(key, 16, "%u", tag);
1179
1180 // printf("retrieve key '%s'\n", key);
1181
1182 ref = state->tags->getObject(key);
1183 if (!ref) {
1184 return 0;
1185 }
1186
1187 o = newObject(state);
1188 o->object = ref;
1189 return o;
1190 }
1191
1192 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
1193 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
1194 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
1195
1196 object_t *
1197 buildDictionary(parser_state_t *state, object_t * header)
1198 {
1199 object_t *o, *t;
1200 int count = 0;
1201 OSDictionary *dict;
1202
1203 // get count and reverse order
1204 o = header->elements;
1205 header->elements = 0;
1206 while (o) {
1207 count++;
1208 t = o;
1209 o = o->next;
1210
1211 t->next = header->elements;
1212 header->elements = t;
1213 }
1214
1215 dict = OSDictionary::withCapacity(count);
1216 if (header->idref >= 0) {
1217 rememberObject(state, header->idref, dict);
1218 }
1219
1220 o = header->elements;
1221 while (o) {
1222 dict->setObject(o->key, o->object);
1223
1224 o->key->release();
1225 o->object->release();
1226 o->key = 0;
1227 o->object = 0;
1228
1229 t = o;
1230 o = o->next;
1231 freeObject(state, t);
1232 }
1233 o = header;
1234 o->object = dict;
1235 return o;
1236 };
1237
1238 object_t *
1239 buildArray(parser_state_t *state, object_t * header)
1240 {
1241 object_t *o, *t;
1242 int count = 0;
1243 OSArray *array;
1244
1245 // get count and reverse order
1246 o = header->elements;
1247 header->elements = 0;
1248 while (o) {
1249 count++;
1250 t = o;
1251 o = o->next;
1252
1253 t->next = header->elements;
1254 header->elements = t;
1255 }
1256
1257 array = OSArray::withCapacity(count);
1258 if (header->idref >= 0) {
1259 rememberObject(state, header->idref, array);
1260 }
1261
1262 o = header->elements;
1263 while (o) {
1264 array->setObject(o->object);
1265
1266 o->object->release();
1267 o->object = 0;
1268
1269 t = o;
1270 o = o->next;
1271 freeObject(state, t);
1272 }
1273 o = header;
1274 o->object = array;
1275 return o;
1276 };
1277
1278 object_t *
1279 buildSet(parser_state_t *state, object_t *header)
1280 {
1281 object_t *o = buildArray(state, header);
1282
1283 OSArray *array = (OSArray *)o->object;
1284 OSSet *set = OSSet::withArray(array, array->getCapacity());
1285
1286 // write over the reference created in buildArray
1287 if (header->idref >= 0) {
1288 rememberObject(state, header->idref, set);
1289 }
1290
1291 array->release();
1292 o->object = set;
1293 return o;
1294 };
1295
1296 object_t *
1297 buildString(parser_state_t *state, object_t *o)
1298 {
1299 OSString *string;
1300
1301 string = OSString::withCString(o->string);
1302 if (o->idref >= 0) {
1303 rememberObject(state, o->idref, string);
1304 }
1305
1306 free(o->string);
1307 o->string = 0;
1308 o->object = string;
1309
1310 return o;
1311 };
1312
1313 object_t *
1314 buildSymbol(parser_state_t *state, object_t *o)
1315 {
1316 OSSymbol *symbol;
1317
1318 symbol = const_cast < OSSymbol * > (OSSymbol::withCString(o->string));
1319 if (o->idref >= 0) {
1320 rememberObject(state, o->idref, symbol);
1321 }
1322
1323 safe_free(o->string, strlen(o->string) + 1);
1324 o->string = 0;
1325 o->object = symbol;
1326
1327 return o;
1328 };
1329
1330 object_t *
1331 buildData(parser_state_t *state, object_t *o)
1332 {
1333 OSData *data;
1334
1335 if (o->size) {
1336 data = OSData::withBytes(o->data, o->size);
1337 } else {
1338 data = OSData::withCapacity(0);
1339 }
1340 if (o->idref >= 0) {
1341 rememberObject(state, o->idref, data);
1342 }
1343
1344 if (o->size) {
1345 free(o->data);
1346 }
1347 o->data = 0;
1348 o->object = data;
1349 return o;
1350 };
1351
1352 object_t *
1353 buildNumber(parser_state_t *state, object_t *o)
1354 {
1355 OSNumber *number = OSNumber::withNumber(o->number, o->size);
1356
1357 if (o->idref >= 0) {
1358 rememberObject(state, o->idref, number);
1359 }
1360
1361 o->object = number;
1362 return o;
1363 };
1364
1365 object_t *
1366 buildBoolean(parser_state_t *state __unused, object_t *o)
1367 {
1368 o->object = ((o->number == 0) ? kOSBooleanFalse : kOSBooleanTrue);
1369 o->object->retain();
1370 return o;
1371 };
1372
1373 OSObject*
1374 OSUnserializeXML(const char *buffer, OSString **errorString)
1375 {
1376 OSObject *object;
1377
1378 if (!buffer) {
1379 return 0;
1380 }
1381 parser_state_t *state = (parser_state_t *)malloc(sizeof(parser_state_t));
1382 if (!state) {
1383 return 0;
1384 }
1385
1386 // just in case
1387 if (errorString) {
1388 *errorString = NULL;
1389 }
1390
1391 state->parseBuffer = buffer;
1392 state->parseBufferIndex = 0;
1393 state->lineNumber = 1;
1394 state->objects = 0;
1395 state->freeObjects = 0;
1396 state->tags = OSDictionary::withCapacity(128);
1397 state->errorString = errorString;
1398 state->parsedObject = 0;
1399 state->parsedObjectCount = 0;
1400 state->retrievedObjectCount = 0;
1401
1402 (void)yyparse((void *)state);
1403
1404 object = state->parsedObject;
1405
1406 cleanupObjects(state);
1407 state->tags->release();
1408 safe_free(state, sizeof(parser_state_t));
1409
1410 return object;
1411 }
1412
1413 #include <libkern/OSSerializeBinary.h>
1414
1415 OSObject*
1416 OSUnserializeXML(const char *buffer, size_t bufferSize, OSString **errorString)
1417 {
1418 if (!buffer) {
1419 return 0;
1420 }
1421 if (bufferSize < sizeof(kOSSerializeBinarySignature)) {
1422 return 0;
1423 }
1424
1425 if (!strcmp(kOSSerializeBinarySignature, buffer)
1426 || (kOSSerializeIndexedBinarySignature == (uint8_t)buffer[0])) {
1427 return OSUnserializeBinary(buffer, bufferSize, errorString);
1428 }
1429
1430 // XML must be null terminated
1431 if (buffer[bufferSize - 1]) {
1432 return 0;
1433 }
1434
1435 return OSUnserializeXML(buffer, errorString);
1436 }
1437
1438
1439 //
1440 //
1441 //
1442 //
1443 //
1444 // DO NOT EDIT OSUnserializeXML.cpp!
1445 //
1446 // this means you!
1447 //
1448 //
1449 //
1450 //
1451 //