]> git.saurik.com Git - apple/xnu.git/blob - libkern/c++/OSUnserializeXML.y
abcc2048489190abe296f4b0915d9274dfd01cad
[apple/xnu.git] / libkern / c++ / OSUnserializeXML.y
1 /*
2 * Copyright (c) 1999-2009 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 YYSTYPE object_t *
68 #define YYPARSE_PARAM state
69 #define YYLEX_PARAM (parser_state_t *)state
70
71 // this is the internal struct used to hold objects on parser stack
72 // it represents objects both before and after they have been created
73 typedef struct object {
74 struct object *next;
75 struct object *free;
76 struct object *elements;
77 OSObject *object;
78 OSString *key; // for dictionary
79 int size;
80 void *data; // for data
81 char *string; // for string & symbol
82 long long number; // for number
83 int idref;
84 } object_t;
85
86 // this code is reentrant, this structure contains all
87 // state information for the parsing of a single buffer
88 typedef struct parser_state {
89 const char *parseBuffer; // start of text to be parsed
90 int parseBufferIndex; // current index into text
91 int lineNumber; // current line number
92 object_t *objects; // internal objects in use
93 object_t *freeObjects; // internal objects that are free
94 OSDictionary *tags; // used to remember "ID" tags
95 OSString **errorString; // parse error with line
96 OSObject *parsedObject; // resultant object of parsed text
97 } parser_state_t;
98
99 #define STATE ((parser_state_t *)state)
100
101 #undef yyerror
102 #define yyerror(s) OSUnserializeerror(STATE, (s))
103 static int OSUnserializeerror(parser_state_t *state, const char *s);
104
105 static int yylex(YYSTYPE *lvalp, parser_state_t *state);
106
107 static object_t *newObject(parser_state_t *state);
108 static void freeObject(parser_state_t *state, object_t *o);
109 static void rememberObject(parser_state_t *state, int tag, OSObject *o);
110 static object_t *retrieveObject(parser_state_t *state, int tag);
111 static void cleanupObjects(parser_state_t *state);
112
113 static object_t *buildDictionary(parser_state_t *state, object_t *o);
114 static object_t *buildArray(parser_state_t *state, object_t *o);
115 static object_t *buildSet(parser_state_t *state, object_t *o);
116 static object_t *buildString(parser_state_t *state, object_t *o);
117 static object_t *buildData(parser_state_t *state, object_t *o);
118 static object_t *buildNumber(parser_state_t *state, object_t *o);
119 static object_t *buildBoolean(parser_state_t *state, object_t *o);
120
121 extern "C" {
122 extern void *kern_os_malloc(size_t size);
123 extern void *kern_os_realloc(void * addr, size_t size);
124 extern void kern_os_free(void * addr);
125
126 } /* extern "C" */
127
128 #define malloc(s) kern_os_malloc(s)
129 #define realloc(a, s) kern_os_realloc(a, s)
130 #define free(a) kern_os_free((void *)a)
131
132 %}
133 %token ARRAY
134 %token BOOLEAN
135 %token DATA
136 %token DICTIONARY
137 %token IDREF
138 %token KEY
139 %token NUMBER
140 %token SET
141 %token STRING
142 %token SYNTAX_ERROR
143 %% /* Grammar rules and actions follow */
144
145 input: /* empty */ { yyerror("unexpected end of buffer");
146 YYERROR;
147 }
148 | object { STATE->parsedObject = $1->object;
149 $1->object = 0;
150 freeObject(STATE, $1);
151 YYACCEPT;
152 }
153 | SYNTAX_ERROR { yyerror("syntax error");
154 YYERROR;
155 }
156 ;
157
158 object: dict { $$ = buildDictionary(STATE, $1); }
159 | array { $$ = buildArray(STATE, $1); }
160 | set { $$ = buildSet(STATE, $1); }
161 | string { $$ = buildString(STATE, $1); }
162 | data { $$ = buildData(STATE, $1); }
163 | number { $$ = buildNumber(STATE, $1); }
164 | boolean { $$ = buildBoolean(STATE, $1); }
165 | idref { $$ = retrieveObject(STATE, $1->idref);
166 if ($$) {
167 $$->object->retain();
168 } else {
169 yyerror("forward reference detected");
170 YYERROR;
171 }
172 freeObject(STATE, $1);
173 }
174 ;
175
176 //------------------------------------------------------------------------------
177
178 dict: '{' '}' { $$ = $1;
179 $$->elements = NULL;
180 }
181 | '{' pairs '}' { $$ = $1;
182 $$->elements = $2;
183 }
184 | DICTIONARY
185 ;
186
187 pairs: pair
188 | pairs pair { $$ = $2;
189 $$->next = $1;
190 }
191 ;
192
193 pair: key object { $$ = $1;
194 $$->key = (OSString *)$$->object;
195 $$->object = $2->object;
196 $$->next = NULL;
197 $2->object = 0;
198 freeObject(STATE, $2);
199 }
200 ;
201
202 key: KEY { $$ = buildString(STATE, $1); }
203 ;
204
205 //------------------------------------------------------------------------------
206
207 array: '(' ')' { $$ = $1;
208 $$->elements = NULL;
209 }
210 | '(' elements ')' { $$ = $1;
211 $$->elements = $2;
212 }
213 | ARRAY
214 ;
215
216 set: '[' ']' { $$ = $1;
217 $$->elements = NULL;
218 }
219 | '[' elements ']' { $$ = $1;
220 $$->elements = $2;
221 }
222 | SET
223 ;
224
225 elements: object { $$ = $1;
226 $$->next = NULL;
227 }
228 | elements object { $$ = $2;
229 $$->next = $1;
230 }
231 ;
232
233 //------------------------------------------------------------------------------
234
235 boolean: BOOLEAN
236 ;
237
238 data: DATA
239 ;
240
241 idref: IDREF
242 ;
243
244 number: NUMBER
245 ;
246
247 string: STRING
248 ;
249
250 %%
251
252 int
253 OSUnserializeerror(parser_state_t * state, const char *s) /* Called by yyparse on errors */
254 {
255 if (state->errorString) {
256 char tempString[128];
257 snprintf(tempString, 128, "OSUnserializeXML: %s near line %d\n", s, state->lineNumber);
258 *(state->errorString) = OSString::withCString(tempString);
259 }
260
261 return 0;
262 }
263
264 #define TAG_MAX_LENGTH 32
265 #define TAG_MAX_ATTRIBUTES 32
266 #define TAG_BAD 0
267 #define TAG_START 1
268 #define TAG_END 2
269 #define TAG_EMPTY 3
270 #define TAG_IGNORE 4
271
272 #define currentChar() (state->parseBuffer[state->parseBufferIndex])
273 #define nextChar() (state->parseBuffer[++state->parseBufferIndex])
274 #define prevChar() (state->parseBuffer[state->parseBufferIndex - 1])
275
276 #define isSpace(c) ((c) == ' ' || (c) == '\t')
277 #define isAlpha(c) (((c) >= 'A' && (c) <= 'Z') || ((c) >= 'a' && (c) <= 'z'))
278 #define isDigit(c) ((c) >= '0' && (c) <= '9')
279 #define isAlphaDigit(c) ((c) >= 'a' && (c) <= 'f')
280 #define isHexDigit(c) (isDigit(c) || isAlphaDigit(c))
281 #define isAlphaNumeric(c) (isAlpha(c) || isDigit(c) || ((c) == '-'))
282
283 static int
284 getTag(parser_state_t *state,
285 char tag[TAG_MAX_LENGTH],
286 int *attributeCount,
287 char attributes[TAG_MAX_ATTRIBUTES][TAG_MAX_LENGTH],
288 char values[TAG_MAX_ATTRIBUTES][TAG_MAX_LENGTH] )
289 {
290 int length = 0;
291 int c = currentChar();
292 int tagType = TAG_START;
293
294 *attributeCount = 0;
295
296 if (c != '<') return TAG_BAD;
297 c = nextChar(); // skip '<'
298
299
300 // <!TAG declarations >
301 // <!-- comments -->
302 if (c == '!') {
303 c = nextChar();
304 bool isComment = (c == '-') && ((c = nextChar()) != 0) && (c == '-');
305 if (!isComment && !isAlpha(c)) return TAG_BAD; // <!1, <!-A, <!eos
306
307 while (c && (c = nextChar()) != 0) {
308 if (c == '\n') state->lineNumber++;
309 if (isComment) {
310 if (c != '-') continue;
311 c = nextChar();
312 if (c != '-') continue;
313 c = nextChar();
314 }
315 if (c == '>') {
316 (void)nextChar();
317 return TAG_IGNORE;
318 }
319 if (isComment) break;
320 }
321 return TAG_BAD;
322 }
323
324 else
325
326 // <? Processing Instructions ?>
327 if (c == '?') {
328 while ((c = nextChar()) != 0) {
329 if (c == '\n') state->lineNumber++;
330 if (c != '?') continue;
331 c = nextChar();
332 if (c == '>') {
333 (void)nextChar();
334 return TAG_IGNORE;
335 }
336 }
337 return TAG_BAD;
338 }
339
340 else
341
342 // </ end tag >
343 if (c == '/') {
344 c = nextChar(); // skip '/'
345 tagType = TAG_END;
346 }
347 if (!isAlpha(c)) return TAG_BAD;
348
349 /* find end of tag while copying it */
350 while (isAlphaNumeric(c)) {
351 tag[length++] = c;
352 c = nextChar();
353 if (length >= (TAG_MAX_LENGTH - 1)) return TAG_BAD;
354 }
355
356 tag[length] = 0;
357
358 // printf("tag %s, type %d\n", tag, tagType);
359
360 // look for attributes of the form attribute = "value" ...
361 while ((c != '>') && (c != '/')) {
362 while (isSpace(c)) c = nextChar();
363
364 length = 0;
365 while (isAlphaNumeric(c)) {
366 attributes[*attributeCount][length++] = c;
367 if (length >= (TAG_MAX_LENGTH - 1)) return TAG_BAD;
368 c = nextChar();
369 }
370 attributes[*attributeCount][length] = 0;
371
372 while (isSpace(c)) c = nextChar();
373
374 if (c != '=') return TAG_BAD;
375 c = nextChar();
376
377 while (isSpace(c)) c = nextChar();
378
379 if (c != '"') return TAG_BAD;
380 c = nextChar();
381 length = 0;
382 while (c != '"') {
383 values[*attributeCount][length++] = c;
384 if (length >= (TAG_MAX_LENGTH - 1)) return TAG_BAD;
385 c = nextChar();
386 }
387 values[*attributeCount][length] = 0;
388
389 c = nextChar(); // skip closing quote
390
391 // printf(" attribute '%s' = '%s', nextchar = '%c'\n",
392 // attributes[*attributeCount], values[*attributeCount], c);
393
394 (*attributeCount)++;
395 if (*attributeCount >= TAG_MAX_ATTRIBUTES) return TAG_BAD;
396 }
397
398 if (c == '/') {
399 c = nextChar(); // skip '/'
400 tagType = TAG_EMPTY;
401 }
402 if (c != '>') return TAG_BAD;
403 c = nextChar(); // skip '>'
404
405 return tagType;
406 }
407
408 static char *
409 getString(parser_state_t *state)
410 {
411 int c = currentChar();
412 int start, length, i, j;
413 char * tempString;
414
415 start = state->parseBufferIndex;
416 /* find end of string */
417
418 while (c != 0) {
419 if (c == '\n') state->lineNumber++;
420 if (c == '<') {
421 break;
422 }
423 c = nextChar();
424 }
425
426 if (c != '<') return 0;
427
428 length = state->parseBufferIndex - start;
429
430 /* copy to null terminated buffer */
431 tempString = (char *)malloc(length + 1);
432 if (tempString == 0) {
433 printf("OSUnserializeXML: can't alloc temp memory\n");
434 goto error;
435 }
436
437 // copy out string in tempString
438 // "&amp;" -> '&', "&lt;" -> '<', "&gt;" -> '>'
439
440 i = j = 0;
441 while (i < length) {
442 c = state->parseBuffer[start + i++];
443 if (c != '&') {
444 tempString[j++] = c;
445 } else {
446 if ((i+3) > length) goto error;
447 c = state->parseBuffer[start + i++];
448 if (c == 'l') {
449 if (state->parseBuffer[start + i++] != 't') goto error;
450 if (state->parseBuffer[start + i++] != ';') goto error;
451 tempString[j++] = '<';
452 continue;
453 }
454 if (c == 'g') {
455 if (state->parseBuffer[start + i++] != 't') goto error;
456 if (state->parseBuffer[start + i++] != ';') goto error;
457 tempString[j++] = '>';
458 continue;
459 }
460 if ((i+3) > length) goto error;
461 if (c == 'a') {
462 if (state->parseBuffer[start + i++] != 'm') goto error;
463 if (state->parseBuffer[start + i++] != 'p') goto error;
464 if (state->parseBuffer[start + i++] != ';') goto error;
465 tempString[j++] = '&';
466 continue;
467 }
468 goto error;
469 }
470 }
471 tempString[j] = 0;
472
473 // printf("string %s\n", tempString);
474
475 return tempString;
476
477 error:
478 if (tempString) free(tempString);
479 return 0;
480 }
481
482 static long long
483 getNumber(parser_state_t *state)
484 {
485 unsigned long long n = 0;
486 int base = 10;
487 bool negate = false;
488 int c = currentChar();
489
490 if (c == '0') {
491 c = nextChar();
492 if (c == 'x') {
493 base = 16;
494 c = nextChar();
495 }
496 }
497 if (base == 10) {
498 if (c == '-') {
499 negate = true;
500 c = nextChar();
501 }
502 while(isDigit(c)) {
503 n = (n * base + c - '0');
504 c = nextChar();
505 }
506 if (negate) {
507 n = (unsigned long long)((long long)n * (long long)-1);
508 }
509 } else {
510 while(isHexDigit(c)) {
511 if (isDigit(c)) {
512 n = (n * base + c - '0');
513 } else {
514 n = (n * base + 0xa + c - 'a');
515 }
516 c = nextChar();
517 }
518 }
519 // printf("number 0x%x\n", (unsigned long)n);
520 return n;
521 }
522
523 // taken from CFXMLParsing/CFPropertyList.c
524
525 static const signed char __CFPLDataDecodeTable[128] = {
526 /* 000 */ -1, -1, -1, -1, -1, -1, -1, -1,
527 /* 010 */ -1, -1, -1, -1, -1, -1, -1, -1,
528 /* 020 */ -1, -1, -1, -1, -1, -1, -1, -1,
529 /* 030 */ -1, -1, -1, -1, -1, -1, -1, -1,
530 /* ' ' */ -1, -1, -1, -1, -1, -1, -1, -1,
531 /* '(' */ -1, -1, -1, 62, -1, -1, -1, 63,
532 /* '0' */ 52, 53, 54, 55, 56, 57, 58, 59,
533 /* '8' */ 60, 61, -1, -1, -1, 0, -1, -1,
534 /* '@' */ -1, 0, 1, 2, 3, 4, 5, 6,
535 /* 'H' */ 7, 8, 9, 10, 11, 12, 13, 14,
536 /* 'P' */ 15, 16, 17, 18, 19, 20, 21, 22,
537 /* 'X' */ 23, 24, 25, -1, -1, -1, -1, -1,
538 /* '`' */ -1, 26, 27, 28, 29, 30, 31, 32,
539 /* 'h' */ 33, 34, 35, 36, 37, 38, 39, 40,
540 /* 'p' */ 41, 42, 43, 44, 45, 46, 47, 48,
541 /* 'x' */ 49, 50, 51, -1, -1, -1, -1, -1
542 };
543
544 #define DATA_ALLOC_SIZE 4096
545
546 static void *
547 getCFEncodedData(parser_state_t *state, unsigned int *size)
548 {
549 int numeq = 0, acc = 0, cntr = 0;
550 int tmpbufpos = 0, tmpbuflen = 0;
551 unsigned char *tmpbuf = (unsigned char *)malloc(DATA_ALLOC_SIZE);
552
553 int c = currentChar();
554 *size = 0;
555
556 while (c != '<') {
557 c &= 0x7f;
558 if (c == 0) {
559 free(tmpbuf);
560 return 0;
561 }
562 if (c == '=') numeq++; else numeq = 0;
563 if (c == '\n') state->lineNumber++;
564 if (__CFPLDataDecodeTable[c] < 0) {
565 c = nextChar();
566 continue;
567 }
568 cntr++;
569 acc <<= 6;
570 acc += __CFPLDataDecodeTable[c];
571 if (0 == (cntr & 0x3)) {
572 if (tmpbuflen <= tmpbufpos + 2) {
573 tmpbuflen += DATA_ALLOC_SIZE;
574 tmpbuf = (unsigned char *)realloc(tmpbuf, tmpbuflen);
575 }
576 tmpbuf[tmpbufpos++] = (acc >> 16) & 0xff;
577 if (numeq < 2)
578 tmpbuf[tmpbufpos++] = (acc >> 8) & 0xff;
579 if (numeq < 1)
580 tmpbuf[tmpbufpos++] = acc & 0xff;
581 }
582 c = nextChar();
583 }
584 *size = tmpbufpos;
585 if (*size == 0) {
586 free(tmpbuf);
587 return 0;
588 }
589 return tmpbuf;
590 }
591
592 static void *
593 getHexData(parser_state_t *state, unsigned int *size)
594 {
595 int c;
596 unsigned char *d, *start, *lastStart;
597
598 start = lastStart = d = (unsigned char *)malloc(DATA_ALLOC_SIZE);
599 c = currentChar();
600
601 while (c != '<') {
602
603 if (isSpace(c)) while ((c = nextChar()) != 0 && isSpace(c)) {};
604 if (c == '\n') {
605 state->lineNumber++;
606 c = nextChar();
607 continue;
608 }
609
610 // get high nibble
611 if (isDigit(c)) {
612 *d = (c - '0') << 4;
613 } else if (isAlphaDigit(c)) {
614 *d = (0xa + (c - 'a')) << 4;
615 } else {
616 goto error;
617 }
618
619 // get low nibble
620 c = nextChar();
621 if (isDigit(c)) {
622 *d |= c - '0';
623 } else if (isAlphaDigit(c)) {
624 *d |= 0xa + (c - 'a');
625 } else {
626 goto error;
627 }
628
629 d++;
630 if ((d - lastStart) >= DATA_ALLOC_SIZE) {
631 int oldsize = d - start;
632 start = (unsigned char *)realloc(start, oldsize + DATA_ALLOC_SIZE);
633 d = lastStart = start + oldsize;
634 }
635 c = nextChar();
636 }
637
638 *size = d - start;
639 return start;
640
641 error:
642
643 *size = 0;
644 free(start);
645 return 0;
646 }
647
648 static int
649 yylex(YYSTYPE *lvalp, parser_state_t *state)
650 {
651 int c, i;
652 int tagType;
653 char tag[TAG_MAX_LENGTH];
654 int attributeCount;
655 char attributes[TAG_MAX_ATTRIBUTES][TAG_MAX_LENGTH];
656 char values[TAG_MAX_ATTRIBUTES][TAG_MAX_LENGTH];
657 object_t *object;
658
659 top:
660 c = currentChar();
661
662 /* skip white space */
663 if (isSpace(c)) while ((c = nextChar()) != 0 && isSpace(c)) {};
664
665 /* keep track of line number, don't return \n's */
666 if (c == '\n') {
667 STATE->lineNumber++;
668 (void)nextChar();
669 goto top;
670 }
671
672 // end of the buffer?
673 if (!c) return 0;
674
675 tagType = getTag(STATE, tag, &attributeCount, attributes, values);
676 if (tagType == TAG_BAD) return SYNTAX_ERROR;
677 if (tagType == TAG_IGNORE) goto top;
678
679 // handle allocation and check for "ID" and "IDREF" tags up front
680 *lvalp = object = newObject(STATE);
681 object->idref = -1;
682 for (i=0; i < attributeCount; i++) {
683 if (attributes[i][0] == 'I' && attributes[i][1] == 'D') {
684 // check for idref's, note: we ignore the tag, for
685 // this to work correctly, all idrefs must be unique
686 // across the whole serialization
687 if (attributes[i][2] == 'R' && attributes[i][3] == 'E' &&
688 attributes[i][4] == 'F' && !attributes[i][5]) {
689 if (tagType != TAG_EMPTY) return SYNTAX_ERROR;
690 object->idref = strtol(values[i], NULL, 0);
691 return IDREF;
692 }
693 // check for id's
694 if (!attributes[i][2]) {
695 object->idref = strtol(values[i], NULL, 0);
696 } else {
697 return SYNTAX_ERROR;
698 }
699 }
700 }
701
702 switch (*tag) {
703 case 'a':
704 if (!strcmp(tag, "array")) {
705 if (tagType == TAG_EMPTY) {
706 object->elements = NULL;
707 return ARRAY;
708 }
709 return (tagType == TAG_START) ? '(' : ')';
710 }
711 break;
712 case 'd':
713 if (!strcmp(tag, "dict")) {
714 if (tagType == TAG_EMPTY) {
715 object->elements = NULL;
716 return DICTIONARY;
717 }
718 return (tagType == TAG_START) ? '{' : '}';
719 }
720 if (!strcmp(tag, "data")) {
721 unsigned int size;
722 if (tagType == TAG_EMPTY) {
723 object->data = NULL;
724 object->size = 0;
725 return DATA;
726 }
727
728 bool isHexFormat = false;
729 for (i=0; i < attributeCount; i++) {
730 if (!strcmp(attributes[i], "format") && !strcmp(values[i], "hex")) {
731 isHexFormat = true;
732 break;
733 }
734 }
735 // CF encoded is the default form
736 if (isHexFormat) {
737 object->data = getHexData(STATE, &size);
738 } else {
739 object->data = getCFEncodedData(STATE, &size);
740 }
741 object->size = size;
742 if ((getTag(STATE, tag, &attributeCount, attributes, values) != TAG_END) || strcmp(tag, "data")) {
743 return SYNTAX_ERROR;
744 }
745 return DATA;
746 }
747 break;
748 case 'f':
749 if (!strcmp(tag, "false")) {
750 if (tagType == TAG_EMPTY) {
751 object->number = 0;
752 return BOOLEAN;
753 }
754 }
755 break;
756 case 'i':
757 if (!strcmp(tag, "integer")) {
758 object->size = 64; // default
759 for (i=0; i < attributeCount; i++) {
760 if (!strcmp(attributes[i], "size")) {
761 object->size = strtoul(values[i], NULL, 0);
762 }
763 }
764 if (tagType == TAG_EMPTY) {
765 object->number = 0;
766 return NUMBER;
767 }
768 object->number = getNumber(STATE);
769 if ((getTag(STATE, tag, &attributeCount, attributes, values) != TAG_END) || strcmp(tag, "integer")) {
770 return SYNTAX_ERROR;
771 }
772 return NUMBER;
773 }
774 break;
775 case 'k':
776 if (!strcmp(tag, "key")) {
777 if (tagType == TAG_EMPTY) return SYNTAX_ERROR;
778 object->string = getString(STATE);
779 if (!object->string) {
780 return SYNTAX_ERROR;
781 }
782 if ((getTag(STATE, tag, &attributeCount, attributes, values) != TAG_END)
783 || strcmp(tag, "key")) {
784 return SYNTAX_ERROR;
785 }
786 return KEY;
787 }
788 break;
789 case 'p':
790 if (!strcmp(tag, "plist")) {
791 freeObject(STATE, object);
792 goto top;
793 }
794 break;
795 case 's':
796 if (!strcmp(tag, "string")) {
797 if (tagType == TAG_EMPTY) {
798 object->string = (char *)malloc(1);
799 object->string[0] = 0;
800 return STRING;
801 }
802 object->string = getString(STATE);
803 if (!object->string) {
804 return SYNTAX_ERROR;
805 }
806 if ((getTag(STATE, tag, &attributeCount, attributes, values) != TAG_END)
807 || strcmp(tag, "string")) {
808 return SYNTAX_ERROR;
809 }
810 return STRING;
811 }
812 if (!strcmp(tag, "set")) {
813 if (tagType == TAG_EMPTY) {
814 object->elements = NULL;
815 return SET;;
816 }
817 if (tagType == TAG_START) {
818 return '[';
819 } else {
820 return ']';
821 }
822 }
823 break;
824 case 't':
825 if (!strcmp(tag, "true")) {
826 if (tagType == TAG_EMPTY) {
827 object->number = 1;
828 return BOOLEAN;
829 }
830 }
831 break;
832 }
833
834 return SYNTAX_ERROR;
835 }
836
837 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
838 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
839 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
840
841 // "java" like allocation, if this code hits a syntax error in the
842 // the middle of the parsed string we just bail with pointers hanging
843 // all over place, this code helps keeps it all together
844
845 //static int object_count = 0;
846
847 object_t *
848 newObject(parser_state_t *state)
849 {
850 object_t *o;
851
852 if (state->freeObjects) {
853 o = state->freeObjects;
854 state->freeObjects = state->freeObjects->next;
855 } else {
856 o = (object_t *)malloc(sizeof(object_t));
857 // object_count++;
858 bzero(o, sizeof(object_t));
859 o->free = state->objects;
860 state->objects = o;
861 }
862
863 return o;
864 }
865
866 void
867 freeObject(parser_state_t * state, object_t *o)
868 {
869 o->next = state->freeObjects;
870 state->freeObjects = o;
871 }
872
873 void
874 cleanupObjects(parser_state_t *state)
875 {
876 object_t *t, *o = state->objects;
877
878 while (o) {
879 if (o->object) {
880 // printf("OSUnserializeXML: releasing object o=%x object=%x\n", (int)o, (int)o->object);
881 o->object->release();
882 }
883 if (o->data) {
884 // printf("OSUnserializeXML: freeing object o=%x data=%x\n", (int)o, (int)o->data);
885 free(o->data);
886 }
887 if (o->key) {
888 // printf("OSUnserializeXML: releasing object o=%x key=%x\n", (int)o, (int)o->key);
889 o->key->release();
890 }
891 if (o->string) {
892 // printf("OSUnserializeXML: freeing object o=%x string=%x\n", (int)o, (int)o->string);
893 free(o->string);
894 }
895
896 t = o;
897 o = o->free;
898 free(t);
899 // object_count--;
900 }
901 // printf("object_count = %d\n", object_count);
902 }
903
904 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
905 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
906 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
907
908 static void
909 rememberObject(parser_state_t *state, int tag, OSObject *o)
910 {
911 char key[16];
912 snprintf(key, 16, "%u", tag);
913
914 // printf("remember key %s\n", key);
915
916 state->tags->setObject(key, o);
917 }
918
919 static object_t *
920 retrieveObject(parser_state_t *state, int tag)
921 {
922 OSObject *ref;
923 object_t *o;
924 char key[16];
925 snprintf(key, 16, "%u", tag);
926
927 // printf("retrieve key '%s'\n", key);
928
929 ref = state->tags->getObject(key);
930 if (!ref) return 0;
931
932 o = newObject(state);
933 o->object = ref;
934 return o;
935 }
936
937 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
938 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
939 // !@$&)(^Q$&*^!$(*!@$_(^%_(*Q#$(_*&!$_(*&!$_(*&!#$(*!@&^!@#%!_!#
940
941 object_t *
942 buildDictionary(parser_state_t *state, object_t * header)
943 {
944 object_t *o, *t;
945 int count = 0;
946 OSDictionary *dict;
947
948 // get count and reverse order
949 o = header->elements;
950 header->elements = 0;
951 while (o) {
952 count++;
953 t = o;
954 o = o->next;
955
956 t->next = header->elements;
957 header->elements = t;
958 }
959
960 dict = OSDictionary::withCapacity(count);
961 if (header->idref >= 0) rememberObject(state, header->idref, dict);
962
963 o = header->elements;
964 while (o) {
965 dict->setObject(o->key, o->object);
966
967 o->key->release();
968 o->object->release();
969 o->key = 0;
970 o->object = 0;
971
972 t = o;
973 o = o->next;
974 freeObject(state, t);
975 }
976 o = header;
977 o->object = dict;
978 return o;
979 };
980
981 object_t *
982 buildArray(parser_state_t *state, object_t * header)
983 {
984 object_t *o, *t;
985 int count = 0;
986 OSArray *array;
987
988 // get count and reverse order
989 o = header->elements;
990 header->elements = 0;
991 while (o) {
992 count++;
993 t = o;
994 o = o->next;
995
996 t->next = header->elements;
997 header->elements = t;
998 }
999
1000 array = OSArray::withCapacity(count);
1001 if (header->idref >= 0) rememberObject(state, header->idref, array);
1002
1003 o = header->elements;
1004 while (o) {
1005 array->setObject(o->object);
1006
1007 o->object->release();
1008 o->object = 0;
1009
1010 t = o;
1011 o = o->next;
1012 freeObject(state, t);
1013 }
1014 o = header;
1015 o->object = array;
1016 return o;
1017 };
1018
1019 object_t *
1020 buildSet(parser_state_t *state, object_t *header)
1021 {
1022 object_t *o = buildArray(state, header);
1023
1024 OSArray *array = (OSArray *)o->object;
1025 OSSet *set = OSSet::withArray(array, array->getCapacity());
1026
1027 // write over the reference created in buildArray
1028 if (header->idref >= 0) rememberObject(state, header->idref, set);
1029
1030 array->release();
1031 o->object = set;
1032 return o;
1033 };
1034
1035 object_t *
1036 buildString(parser_state_t *state, object_t *o)
1037 {
1038 OSString *string;
1039
1040 string = OSString::withCString(o->string);
1041 if (o->idref >= 0) rememberObject(state, o->idref, string);
1042
1043 free(o->string);
1044 o->string = 0;
1045 o->object = string;
1046
1047 return o;
1048 };
1049
1050 object_t *
1051 buildData(parser_state_t *state, object_t *o)
1052 {
1053 OSData *data;
1054
1055 if (o->size) {
1056 data = OSData::withBytes(o->data, o->size);
1057 } else {
1058 data = OSData::withCapacity(0);
1059 }
1060 if (o->idref >= 0) rememberObject(state, o->idref, data);
1061
1062 if (o->size) free(o->data);
1063 o->data = 0;
1064 o->object = data;
1065 return o;
1066 };
1067
1068 object_t *
1069 buildNumber(parser_state_t *state, object_t *o)
1070 {
1071 OSNumber *number = OSNumber::withNumber(o->number, o->size);
1072
1073 if (o->idref >= 0) rememberObject(state, o->idref, number);
1074
1075 o->object = number;
1076 return o;
1077 };
1078
1079 object_t *
1080 buildBoolean(parser_state_t *state __unused, object_t *o)
1081 {
1082 o->object = ((o->number == 0) ? kOSBooleanFalse : kOSBooleanTrue);
1083 o->object->retain();
1084 return o;
1085 };
1086
1087 OSObject*
1088 OSUnserializeXML(const char *buffer, OSString **errorString)
1089 {
1090 OSObject *object;
1091 parser_state_t *state = (parser_state_t *)malloc(sizeof(parser_state_t));
1092
1093 if ((!state) || (!buffer)) return 0;
1094
1095 // just in case
1096 if (errorString) *errorString = NULL;
1097
1098 state->parseBuffer = buffer;
1099 state->parseBufferIndex = 0;
1100 state->lineNumber = 1;
1101 state->objects = 0;
1102 state->freeObjects = 0;
1103 state->tags = OSDictionary::withCapacity(128);
1104 state->errorString = errorString;
1105 state->parsedObject = 0;
1106
1107 (void)yyparse((void *)state);
1108
1109 object = state->parsedObject;
1110
1111 cleanupObjects(state);
1112 state->tags->release();
1113 free(state);
1114
1115 return object;
1116 }
1117
1118
1119 //
1120 //
1121 //
1122 //
1123 //
1124 // DO NOT EDIT OSUnserializeXML.cpp!
1125 //
1126 // this means you!
1127 //
1128 //
1129 //
1130 //
1131 //