// // JSONKit.m // http://github.com/johnezang/JSONKit // Licensed under the terms of the BSD License, as specified below. // /* Copyright (c) 2011, John Engelhart All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. * Neither the name of the Zang Industries nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission. THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ /* Acknowledgments: The bulk of the UTF8 / UTF32 conversion and verification comes from ConvertUTF.[hc]. It has been modified from the original sources. The original sources were obtained from http://www.unicode.org/. However, the web site no longer seems to host the files. Instead, the Unicode FAQ http://www.unicode.org/faq//utf_bom.html#gen4 points to International Components for Unicode (ICU) http://site.icu-project.org/ as an example of how to write a UTF converter. The decision to use the ConvertUTF.[ch] code was made to leverage "proven" code. Hopefully the local modifications are bug free. The code in isValidCodePoint() is derived from the ICU code in utf.h for the macros U_IS_UNICODE_NONCHAR and U_IS_UNICODE_CHAR. From the original ConvertUTF.[ch]: * Copyright 2001-2004 Unicode, Inc. * * Disclaimer * * This source code is provided as is by Unicode, Inc. No claims are * made as to fitness for any particular purpose. No warranties of any * kind are expressed or implied. The recipient agrees to determine * applicability of information provided. If this file has been * purchased on magnetic or optical media from Unicode, Inc., the * sole remedy for any claim will be exchange of defective media * within 90 days of receipt. * * Limitations on Rights to Redistribute This Code * * Unicode, Inc. hereby grants the right to freely use the information * supplied in this file in the creation of products supporting the * Unicode Standard, and to make copies of this file in any form * for internal or external distribution as long as this notice * remains attached. */ #include #include #include #include #include #include #include #include #include #import "JSONKit.h" //#include #include #include #include #include //#import #import #import #import #import #import #import #import #ifdef __OBJC_GC__ #error JSONKit does not support Objective-C Garbage Collection #endif #if (UINT_MAX < 0xffffffffU) || (INT_MIN < (-2147483647-1)) || (ULLONG_MAX < 0xffffffffffffffffULL) || (LLONG_MIN < (-0x7fffffffffffffffLL-1)) #error JSONKit requires the C 'int' and 'long long' types to be at least 32 and 64 bits respectively. #endif #ifdef JK_ENABLE_CF_TRANSFER_OWNERSHIP_CALLBACKS #warning As of JSONKit v1.4, JK_ENABLE_CF_TRANSFER_OWNERSHIP_CALLBACKS is no longer required. It is no longer a valid option. #endif // For DJB hash. #define JK_HASH_INIT (1402737925UL) // Use __builtin_clz() instead of trailingBytesForUTF8[] table lookup. #define JK_FAST_TRAILING_BYTES // JK_CACHE_SLOTS must be a power of 2. Default size is 1024 slots. #define JK_CACHE_SLOTS_BITS (10) #define JK_CACHE_SLOTS (1UL << JK_CACHE_SLOTS_BITS) // JK_CACHE_PROBES is the number of probe attempts. #define JK_CACHE_PROBES (4UL) // JK_INIT_CACHE_AGE must be (1 << AGE) - 1 #define JK_INIT_CACHE_AGE (0) // JK_TOKENBUFFER_SIZE is the default stack size for the temporary buffer used to hold "non-simple" strings (i.e., contains \ escapes) #define JK_TOKENBUFFER_SIZE (1024UL * 2UL) // JK_STACK_OBJS is the default number of spaces reserved on the stack for temporarily storing pointers to Obj-C objects before they can be transferred to a NSArray / NSDictionary. #define JK_STACK_OBJS (1024UL * 1UL) #define JK_JSONBUFFER_SIZE (1024UL * 4UL) #define JK_UTF8BUFFER_SIZE (1024UL * 16UL) #if defined (__GNUC__) && (__GNUC__ >= 4) #define JK_ATTRIBUTES(attr, ...) __attribute__((attr, ##__VA_ARGS__)) #define JK_EXPECTED(cond, expect) __builtin_expect((long)(cond), (expect)) #define JK_EXPECT_T(cond) JK_EXPECTED(cond, 1U) #define JK_EXPECT_F(cond) JK_EXPECTED(cond, 0U) #define JK_PREFETCH(ptr) __builtin_prefetch(ptr) #else // defined (__GNUC__) && (__GNUC__ >= 4) #define JK_ATTRIBUTES(attr, ...) #define JK_EXPECTED(cond, expect) (cond) #define JK_EXPECT_T(cond) (cond) #define JK_EXPECT_F(cond) (cond) #define JK_PREFETCH(ptr) #endif // defined (__GNUC__) && (__GNUC__ >= 4) #define JK_STATIC_INLINE static __inline__ JK_ATTRIBUTES(always_inline) #define JK_ALIGNED(arg) JK_ATTRIBUTES(aligned(arg)) #define JK_UNUSED_ARG JK_ATTRIBUTES(unused) #define JK_WARN_UNUSED JK_ATTRIBUTES(warn_unused_result) #define JK_WARN_UNUSED_CONST JK_ATTRIBUTES(warn_unused_result, const) #define JK_WARN_UNUSED_PURE JK_ATTRIBUTES(warn_unused_result, pure) #define JK_WARN_UNUSED_SENTINEL JK_ATTRIBUTES(warn_unused_result, sentinel) #define JK_NONNULL_ARGS(arg, ...) JK_ATTRIBUTES(nonnull(arg, ##__VA_ARGS__)) #define JK_WARN_UNUSED_NONNULL_ARGS(arg, ...) JK_ATTRIBUTES(warn_unused_result, nonnull(arg, ##__VA_ARGS__)) #define JK_WARN_UNUSED_CONST_NONNULL_ARGS(arg, ...) JK_ATTRIBUTES(warn_unused_result, const, nonnull(arg, ##__VA_ARGS__)) #define JK_WARN_UNUSED_PURE_NONNULL_ARGS(arg, ...) JK_ATTRIBUTES(warn_unused_result, pure, nonnull(arg, ##__VA_ARGS__)) #if defined (__GNUC__) && (__GNUC__ >= 4) && (__GNUC_MINOR__ >= 3) #define JK_ALLOC_SIZE_NON_NULL_ARGS_WARN_UNUSED(as, nn, ...) JK_ATTRIBUTES(warn_unused_result, nonnull(nn, ##__VA_ARGS__), alloc_size(as)) #else // defined (__GNUC__) && (__GNUC__ >= 4) && (__GNUC_MINOR__ >= 3) #define JK_ALLOC_SIZE_NON_NULL_ARGS_WARN_UNUSED(as, nn, ...) JK_ATTRIBUTES(warn_unused_result, nonnull(nn, ##__VA_ARGS__)) #endif // defined (__GNUC__) && (__GNUC__ >= 4) && (__GNUC_MINOR__ >= 3) @class JKArray, JKDictionaryEnumerator, JKDictionary; enum { JSONNumberStateStart = 0, JSONNumberStateFinished = 1, JSONNumberStateError = 2, JSONNumberStateWholeNumberStart = 3, JSONNumberStateWholeNumberMinus = 4, JSONNumberStateWholeNumberZero = 5, JSONNumberStateWholeNumber = 6, JSONNumberStatePeriod = 7, JSONNumberStateFractionalNumberStart = 8, JSONNumberStateFractionalNumber = 9, JSONNumberStateExponentStart = 10, JSONNumberStateExponentPlusMinus = 11, JSONNumberStateExponent = 12, }; enum { JSONStringStateStart = 0, JSONStringStateParsing = 1, JSONStringStateFinished = 2, JSONStringStateError = 3, JSONStringStateEscape = 4, JSONStringStateEscapedUnicode1 = 5, JSONStringStateEscapedUnicode2 = 6, JSONStringStateEscapedUnicode3 = 7, JSONStringStateEscapedUnicode4 = 8, JSONStringStateEscapedUnicodeSurrogate1 = 9, JSONStringStateEscapedUnicodeSurrogate2 = 10, JSONStringStateEscapedUnicodeSurrogate3 = 11, JSONStringStateEscapedUnicodeSurrogate4 = 12, JSONStringStateEscapedNeedEscapeForSurrogate = 13, JSONStringStateEscapedNeedEscapedUForSurrogate = 14, }; enum { JKParseAcceptValue = (1 << 0), JKParseAcceptComma = (1 << 1), JKParseAcceptEnd = (1 << 2), JKParseAcceptValueOrEnd = (JKParseAcceptValue | JKParseAcceptEnd), JKParseAcceptCommaOrEnd = (JKParseAcceptComma | JKParseAcceptEnd), }; enum { JKClassUnknown = 0, JKClassString = 1, JKClassNumber = 2, JKClassArray = 3, JKClassDictionary = 4, JKClassNull = 5, }; enum { JKManagedBufferOnStack = 1, JKManagedBufferOnHeap = 2, JKManagedBufferLocationMask = (0x3), JKManagedBufferLocationShift = (0), JKManagedBufferMustFree = (1 << 2), }; typedef JKFlags JKManagedBufferFlags; enum { JKObjectStackOnStack = 1, JKObjectStackOnHeap = 2, JKObjectStackLocationMask = (0x3), JKObjectStackLocationShift = (0), JKObjectStackMustFree = (1 << 2), }; typedef JKFlags JKObjectStackFlags; enum { JKTokenTypeInvalid = 0, JKTokenTypeNumber = 1, JKTokenTypeString = 2, JKTokenTypeObjectBegin = 3, JKTokenTypeObjectEnd = 4, JKTokenTypeArrayBegin = 5, JKTokenTypeArrayEnd = 6, JKTokenTypeSeparator = 7, JKTokenTypeComma = 8, JKTokenTypeTrue = 9, JKTokenTypeFalse = 10, JKTokenTypeNull = 11, JKTokenTypeWhiteSpace = 12, }; typedef NSUInteger JKTokenType; // These are prime numbers to assist with hash slot probing. enum { JKValueTypeNone = 0, JKValueTypeString = 5, JKValueTypeLongLong = 7, JKValueTypeUnsignedLongLong = 11, JKValueTypeDouble = 13, }; typedef NSUInteger JKValueType; enum { JKEncodeAsData = 1, JKEncodeAsString = 2, }; typedef NSUInteger JKEncodeAsType; typedef NSUInteger JKHash; typedef id (*NSNumberAllocImp)(id object, SEL selector); typedef id (*NSNumberInitWithUnsignedLongLongImp)(id object, SEL selector, unsigned long long value); typedef struct JKTokenCacheItem JKTokenCacheItem; typedef struct JKTokenCache JKTokenCache; typedef struct JKTokenValue JKTokenValue; typedef struct JKParseToken JKParseToken; typedef struct JKPtrRange JKPtrRange; typedef struct JKObjectStack JKObjectStack; typedef struct JKBuffer JKBuffer; typedef struct JKConstBuffer JKConstBuffer; typedef struct JKConstPtrRange JKConstPtrRange; typedef struct JKRange JKRange; typedef struct JKManagedBuffer JKManagedBuffer; typedef struct JKFastClassLookup JKFastClassLookup; typedef struct JKEncodeState JKEncodeState; typedef struct JKObjCImpCache JKObjCImpCache; typedef struct JKHashTableEntry JKHashTableEntry; struct JKPtrRange { unsigned char *ptr; size_t length; }; struct JKConstPtrRange { const unsigned char *ptr; size_t length; }; struct JKRange { size_t location, length; }; struct JKManagedBuffer { JKPtrRange bytes; JKManagedBufferFlags flags; size_t roundSizeUpToMultipleOf; }; struct JKObjectStack { void **objects, **keys; CFHashCode *cfHashes; size_t count, index, roundSizeUpToMultipleOf; JKObjectStackFlags flags; }; struct JKBuffer { JKPtrRange bytes; }; struct JKConstBuffer { JKConstPtrRange bytes; }; struct JKTokenValue { JKConstPtrRange ptrRange; JKValueType type; JKHash hash; union { long long longLongValue; unsigned long long unsignedLongLongValue; double doubleValue; } number; JKTokenCacheItem *cacheItem; }; struct JKParseToken { JKConstPtrRange tokenPtrRange; JKTokenType type; JKTokenValue value; JKManagedBuffer tokenBuffer; }; struct JKTokenCacheItem { void *object; JKHash hash; CFHashCode cfHash; size_t size; unsigned char *bytes; JKValueType type; }; struct JKTokenCache { JKTokenCacheItem *items; size_t count; unsigned int prng_lfsr; unsigned char age[JK_CACHE_SLOTS]; }; struct JKObjCImpCache { Class NSNumberClass; NSNumberAllocImp NSNumberAlloc; NSNumberInitWithUnsignedLongLongImp NSNumberInitWithUnsignedLongLong; }; struct JKParseState { JKParseOptionFlags parseOptionFlags; JKConstBuffer stringBuffer; size_t atIndex, lineNumber, lineStartIndex; size_t prev_atIndex, prev_lineNumber, prev_lineStartIndex; JKParseToken token; JKObjectStack objectStack; JKTokenCache cache; JKObjCImpCache objCImpCache; NSError *error; int errorIsPrev; }; struct JKFastClassLookup { void *stringClass; void *numberClass; void *arrayClass; void *dictionaryClass; void *nullClass; }; struct JKEncodeState { JKManagedBuffer utf8ConversionBuffer; JKManagedBuffer stringBuffer; size_t atIndex; JKFastClassLookup fastClassLookup; JKSerializeOptionFlags serializeOptionFlags; NSError *error; }; struct JKHashTableEntry { NSUInteger keyHash; id key, object; }; typedef uint32_t UTF32; /* at least 32 bits */ typedef uint16_t UTF16; /* at least 16 bits */ typedef uint8_t UTF8; /* typically 8 bits */ typedef enum { conversionOK, /* conversion successful */ sourceExhausted, /* partial character in source, but hit end */ targetExhausted, /* insuff. room in target for conversion */ sourceIllegal /* source sequence is illegal/malformed */ } ConversionResult; #define UNI_REPLACEMENT_CHAR (UTF32)0x0000FFFD #define UNI_MAX_BMP (UTF32)0x0000FFFF #define UNI_MAX_UTF16 (UTF32)0x0010FFFF #define UNI_MAX_UTF32 (UTF32)0x7FFFFFFF #define UNI_MAX_LEGAL_UTF32 (UTF32)0x0010FFFF #define UNI_SUR_HIGH_START (UTF32)0xD800 #define UNI_SUR_HIGH_END (UTF32)0xDBFF #define UNI_SUR_LOW_START (UTF32)0xDC00 #define UNI_SUR_LOW_END (UTF32)0xDFFF #if !defined(JK_FAST_TRAILING_BYTES) static const char trailingBytesForUTF8[256] = { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 3,3,3,3,3,3,3,3,4,4,4,4,5,5,5,5 }; #endif static const UTF32 offsetsFromUTF8[6] = { 0x00000000UL, 0x00003080UL, 0x000E2080UL, 0x03C82080UL, 0xFA082080UL, 0x82082080UL }; static const UTF8 firstByteMark[7] = { 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC }; #define JK_AT_STRING_PTR(x) (&((x)->stringBuffer.bytes.ptr[(x)->atIndex])) #define JK_END_STRING_PTR(x) (&((x)->stringBuffer.bytes.ptr[(x)->stringBuffer.bytes.length])) static JKArray *_JKArrayCreate(id *objects, NSUInteger count); static JKDictionary *_JKDictionaryCreate(id *keys, NSUInteger *keyHashes, id *objects, NSUInteger count); static JKHashTableEntry *_JKDictionaryHashEntry(JKDictionary *dictionary); static NSUInteger _JKDictionaryCapacity(JKDictionary *dictionary); static void _JKDictionaryAddObject(JKDictionary *dictionary, NSUInteger keyHash, id key, id object); static void _JSONDecoderCleanup(JSONDecoder *decoder); static void jk_managedBuffer_release(JKManagedBuffer *managedBuffer); static void jk_managedBuffer_setToStackBuffer(JKManagedBuffer *managedBuffer, unsigned char *ptr, size_t length); static unsigned char *jk_managedBuffer_resize(JKManagedBuffer *managedBuffer, size_t newSize); static void jk_objectStack_release(JKObjectStack *objectStack); static void jk_objectStack_setToStackBuffer(JKObjectStack *objectStack, void **objects, void **keys, CFHashCode *cfHashes, size_t count); static int jk_objectStack_resize(JKObjectStack *objectStack, size_t newCount); static void jk_error(JKParseState *parseState, NSString *format, ...); static int jk_parse_string(JKParseState *parseState); static int jk_parse_number(JKParseState *parseState); static size_t jk_parse_is_newline(JKParseState *parseState, const unsigned char *atCharacterPtr); JK_STATIC_INLINE int jk_parse_skip_newline(JKParseState *parseState); JK_STATIC_INLINE void jk_parse_skip_whitespace(JKParseState *parseState); static int jk_parse_next_token(JKParseState *parseState); static void jk_error_parse_accept_or3(JKParseState *parseState, int state, NSString *or1String, NSString *or2String, NSString *or3String); static void *jk_create_dictionary(JKParseState *parseState, size_t startingObjectIndex); static void *jk_parse_dictionary(JKParseState *parseState); static void *jk_parse_array(JKParseState *parseState); static void *jk_object_for_token(JKParseState *parseState); static void *jk_cachedObjects(JKParseState *parseState); JK_STATIC_INLINE void jk_cache_age(JKParseState *parseState); JK_STATIC_INLINE void jk_set_parsed_token(JKParseState *parseState, const unsigned char *ptr, size_t length, JKTokenType type, size_t advanceBy); static void jk_encode_error(JKEncodeState *encodeState, NSString *format, ...); static int jk_encode_printf(JKEncodeState *encodeState, const char *format, ...); static int jk_encode_write(JKEncodeState *encodeState, const char *format); static int jk_encode_write1(JKEncodeState *encodeState, const char *format); static int jk_encode_writen(JKEncodeState *encodeState, const char *format, size_t length); static int jk_encode_add_atom_to_buffer(JKEncodeState *encodeState, void *objectPtr); static id jk_encode(void *object, JKSerializeOptionFlags optionFlags, JKEncodeAsType encodeAs, NSError **error); JK_STATIC_INLINE size_t jk_min(size_t a, size_t b); JK_STATIC_INLINE size_t jk_max(size_t a, size_t b); JK_STATIC_INLINE JKHash calculateHash(JKHash currentHash, unsigned char c); #pragma mark - @interface JKArray : NSArray { id *objects; NSUInteger count; } @end @implementation JKArray static Class _JKArrayClass = NULL; static size_t _JKArrayInstanceSize = 0UL; + (void)load { NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; // Though technically not required, the run time enviornment at +load time may be less than ideal. _JKArrayClass = [JKArray class]; _JKArrayInstanceSize = class_getInstanceSize(_JKArrayClass); if(_JKArrayInstanceSize < 16UL) { _JKArrayInstanceSize = 16UL; } [pool release]; pool = NULL; } + (id)allocWithZone:(NSZone *)zone { #pragma unused(zone) [NSException raise:NSInvalidArgumentException format:@"*** - [%@ %@]: The %@ class is private to JSONKit and should not be used in this fashion.", NSStringFromClass([self class]), NSStringFromSelector(_cmd), NSStringFromClass([self class])]; return(NULL); } static JKArray *_JKArrayCreate(id *objects, NSUInteger count) { NSCParameterAssert((objects != NULL) && (_JKArrayClass != NULL) && (_JKArrayInstanceSize > 0UL)); JKArray *array = NULL; if(JK_EXPECT_T((array = (JKArray *)calloc(1UL, _JKArrayInstanceSize)) != NULL)) { // Directly allocate the JKArray instance via calloc. array->isa = _JKArrayClass; array->count = count; if(JK_EXPECT_F((array->objects = (id *)malloc(sizeof(id) * array->count)) == NULL)) { [array autorelease]; return(NULL); } memcpy(array->objects, objects, array->count * sizeof(id)); } return(array); } - (void)dealloc { if(JK_EXPECT_T(objects != NULL)) { NSUInteger atObject = 0UL; for(atObject = 0UL; atObject < count; atObject++) { if(JK_EXPECT_T(objects[atObject] != NULL)) { CFRelease(objects[atObject]); objects[atObject] = NULL; } } free(objects); objects = NULL; } [super dealloc]; } - (NSUInteger)count { return(count); } - (void)getObjects:(id *)objectsPtr range:(NSRange)range { NSParameterAssert(objects != NULL); if((range.location > count) || (NSMaxRange(range) > count)) { [NSException raise:NSRangeException format:@"*** -[%@ %@]: index (%lu) beyond bounds (%lu)", NSStringFromClass([self class]), NSStringFromSelector(_cmd), NSMaxRange(range), count]; } memcpy(objectsPtr, objects + range.location, range.length * sizeof(id)); } - (id)objectAtIndex:(NSUInteger)objectIndex { if(objectIndex >= count) { [NSException raise:NSRangeException format:@"*** -[%@ %@]: index (%lu) beyond bounds (%lu)", NSStringFromClass([self class]), NSStringFromSelector(_cmd), objectIndex, count]; } NSParameterAssert((objects != NULL) && (objects[objectIndex] != NULL)); return(objects[objectIndex]); } - (NSUInteger)countByEnumeratingWithState:(NSFastEnumerationState *)state objects:(id *)stackbuf count:(NSUInteger)len { NSParameterAssert(objects != NULL); if(JK_EXPECT_F(state->state == 0UL)) { state->mutationsPtr = &state->extra[0]; state->itemsPtr = stackbuf; } if(JK_EXPECT_F(state->state >= count)) { return(0UL); } NSUInteger enumeratedCount = 0UL; while(JK_EXPECT_T(enumeratedCount < len) && JK_EXPECT_T(state->state < count)) { NSParameterAssert(objects[state->state] != NULL); stackbuf[enumeratedCount++] = objects[state->state++]; } return(enumeratedCount); } - (id)copyWithZone:(NSZone *)zone { #pragma unused(zone) return([self retain]); } - (id)mutableCopyWithZone:(NSZone *)zone { NSParameterAssert(objects != NULL); return([[NSMutableArray allocWithZone:zone] initWithObjects:objects count:count]); } @end #pragma mark - @interface JKDictionaryEnumerator : NSEnumerator { id collection; NSUInteger nextObject; } - (id)initWithJKDictionary:(JKDictionary *)initDictionary; - (NSArray *)allObjects; - (id)nextObject; @end @implementation JKDictionaryEnumerator - (id)initWithJKDictionary:(JKDictionary *)initDictionary { NSParameterAssert(initDictionary != NULL); if((self = [super init]) == NULL) { return(NULL); } if((collection = (id)CFRetain(initDictionary)) == NULL) { [self autorelease]; return(NULL); } return(self); } - (void)dealloc { if(collection != NULL) { CFRelease(collection); collection = NULL; } [super dealloc]; } - (NSArray *)allObjects { NSParameterAssert(collection != NULL); NSUInteger count = [collection count], atObject = 0UL; id objects[count]; while((objects[atObject] = [self nextObject]) != NULL) { NSParameterAssert(atObject < count); atObject++; } return([NSArray arrayWithObjects:objects count:atObject]); } - (id)nextObject { NSParameterAssert((collection != NULL) && (_JKDictionaryHashEntry(collection) != NULL)); JKHashTableEntry *entry = _JKDictionaryHashEntry(collection); NSUInteger capacity = _JKDictionaryCapacity(collection); id returnObject = NULL; if(entry != NULL) { while((nextObject < capacity) && ((returnObject = entry[nextObject++].key) == NULL)) { /* ... */ } } return(returnObject); } @end #pragma mark - @interface JKDictionary : NSDictionary { NSUInteger count, capacity; JKHashTableEntry *entry; } @end @implementation JKDictionary static Class _JKDictionaryClass = NULL; static size_t _JKDictionaryInstanceSize = 0UL; + (void)load { NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; // Though technically not required, the run time enviornment at +load time may be less than ideal. _JKDictionaryClass = [JKDictionary class]; _JKDictionaryInstanceSize = class_getInstanceSize(_JKDictionaryClass); if(_JKDictionaryInstanceSize < 16UL) { _JKDictionaryInstanceSize = 16UL; } [pool release]; pool = NULL; } + (id)allocWithZone:(NSZone *)zone { #pragma unused(zone) [NSException raise:NSInvalidArgumentException format:@"*** - [%@ %@]: The %@ class is private to JSONKit and should not be used in this fashion.", NSStringFromClass([self class]), NSStringFromSelector(_cmd), NSStringFromClass([self class])]; return(NULL); } // These values are taken from Core Foundation CF-550 CFBasicHash.m. As a bonus, they allign very well with our JKHashTableEntry struct too. static const NSUInteger jk_dictionaryCapacities[] = { 0UL, 3UL, 7UL, 13UL, 23UL, 41UL, 71UL, 127UL, 191UL, 251UL, 383UL, 631UL, 1087UL, 1723UL, 2803UL, 4523UL, 7351UL, 11959UL, 19447UL, 31231UL, 50683UL, 81919UL, 132607UL, 214519UL, 346607UL, 561109UL, 907759UL, 1468927UL, 2376191UL, 3845119UL, 6221311UL, 10066421UL, 16287743UL, 26354171UL, 42641881UL, 68996069UL, 111638519UL, 180634607UL, 292272623UL, 472907251UL }; static JKDictionary *_JKDictionaryCreate(id *keys, NSUInteger *keyHashes, id *objects, NSUInteger count) { NSCParameterAssert((keys != NULL) && (keyHashes != NULL) && (objects != NULL) && (_JKDictionaryClass != NULL) && (_JKDictionaryInstanceSize > 0UL)); JKDictionary *dictionary = NULL; if(JK_EXPECT_T((dictionary = (JKDictionary *)calloc(1UL, _JKDictionaryInstanceSize)) != NULL)) { // Directly allocate the JKArray instance via calloc. dictionary->isa = _JKDictionaryClass; NSUInteger bottom = 0UL, top = sizeof(jk_dictionaryCapacities) / sizeof(NSUInteger), mid = 0UL, tableSize = lround(floor((count) * 1.33)); while(top > bottom) { mid = (top + bottom) / 2UL; if(jk_dictionaryCapacities[mid] < tableSize) { bottom = mid + 1UL; } else { top = mid; } } dictionary->capacity = jk_dictionaryCapacities[bottom]; dictionary->count = 0UL; if(JK_EXPECT_F((dictionary->entry = (JKHashTableEntry *)calloc(1UL, sizeof(JKHashTableEntry) * dictionary->capacity)) == NULL)) { [dictionary autorelease]; return(NULL); } NSUInteger idx = 0UL; for(idx = 0UL; idx < count; idx++) { _JKDictionaryAddObject(dictionary, keyHashes[idx], keys[idx], objects[idx]); } } return(dictionary); } - (void)dealloc { if(JK_EXPECT_T(entry != NULL)) { NSUInteger atEntry = 0UL; for(atEntry = 0UL; atEntry < capacity; atEntry++) { if(JK_EXPECT_T(entry[atEntry].key != NULL)) { CFRelease(entry[atEntry].key); entry[atEntry].key = NULL; } if(JK_EXPECT_T(entry[atEntry].object != NULL)) { CFRelease(entry[atEntry].object); entry[atEntry].object = NULL; } } free(entry); entry = NULL; } [super dealloc]; } static JKHashTableEntry *_JKDictionaryHashEntry(JKDictionary *dictionary) { NSCParameterAssert(dictionary != NULL); return(dictionary->entry); } static NSUInteger _JKDictionaryCapacity(JKDictionary *dictionary) { NSCParameterAssert(dictionary != NULL); return(dictionary->capacity); } static void _JKDictionaryAddObject(JKDictionary *dictionary, NSUInteger keyHash, id key, id object) { NSCParameterAssert((dictionary != NULL) && (key != NULL) && (object != NULL) && (dictionary->count < dictionary->capacity) && (dictionary->entry != NULL)); NSUInteger keyEntry = keyHash % dictionary->capacity, idx = 0UL; for(idx = 0UL; idx < dictionary->capacity; idx++) { NSUInteger entryIdx = (keyEntry + idx) % dictionary->capacity; JKHashTableEntry *atEntry = &dictionary->entry[entryIdx]; if(JK_EXPECT_F(atEntry->keyHash == keyHash) && JK_EXPECT_T(atEntry->key != NULL) && (JK_EXPECT_F(key == atEntry->key) || JK_EXPECT_F(CFEqual(atEntry->key, key)))) { NSCParameterAssert((atEntry->object != NULL) && (dictionary->count > 0UL)); CFRelease(atEntry->key); atEntry->key = NULL; CFRelease(atEntry->object); atEntry->object = NULL; atEntry->keyHash = 0UL; dictionary->count--; } if(JK_EXPECT_T(atEntry->key == NULL)) { atEntry->key = key; atEntry->object = object; atEntry->keyHash = keyHash; dictionary->count++; return; } } // We should never get here. If we do, we -release the key / object because it's our responsibility. CFRelease(key); CFRelease(object); } - (NSUInteger)count { return(count); } - (id)objectForKey:(id)aKey { NSParameterAssert((entry != NULL) && (count <= capacity)); NSUInteger keyHash = CFHash(aKey), keyEntry = (keyHash % capacity), idx = 0UL; id returnObject = NULL; for(idx = 0UL; idx < capacity; idx++) { JKHashTableEntry *atEntry = &entry[(keyEntry + idx) % capacity]; if(JK_EXPECT_T(atEntry->keyHash == keyHash) && JK_EXPECT_T(atEntry->key != NULL) && ((atEntry->key == aKey) || CFEqual(atEntry->key, aKey))) { NSParameterAssert(atEntry->object != NULL); returnObject = atEntry->object; break; } } return(returnObject); } - (void)getObjects:(id *)objects andKeys:(id *)keys { NSParameterAssert((entry != NULL) && (count <= capacity)); NSUInteger idx = 0UL, atEntry = 0UL; for(atEntry = 0UL; atEntry < capacity; atEntry++) { if(JK_EXPECT_T(entry[atEntry].key != NULL)) { NSParameterAssert((entry[atEntry].object != NULL) && (idx < count)); if(JK_EXPECT_T(keys != NULL)) { keys[idx] = entry[atEntry].key; } if(JK_EXPECT_T(objects != NULL)) { objects[idx] = entry[atEntry].object; } idx++; } } } - (NSUInteger)countByEnumeratingWithState:(NSFastEnumerationState *)state objects:(id *)stackbuf count:(NSUInteger)len { NSParameterAssert((entry != NULL) && (count <= capacity)); if(JK_EXPECT_F(state->state == 0UL)) { state->mutationsPtr = &state->extra[0]; state->itemsPtr = stackbuf; } if(JK_EXPECT_F(state->state >= capacity)) { return(0UL); } NSUInteger enumeratedCount = 0UL; while(JK_EXPECT_T(enumeratedCount < len) && JK_EXPECT_T(state->state < capacity)) { if(JK_EXPECT_T(entry[state->state].key != NULL)) { stackbuf[enumeratedCount++] = entry[state->state].key; } state->state++; } return(enumeratedCount); } - (NSEnumerator *)keyEnumerator { return([[[JKDictionaryEnumerator alloc] initWithJKDictionary:self] autorelease]); } - (id)copyWithZone:(NSZone *)zone { #pragma unused(zone) return([self retain]); } - (id)mutableCopyWithZone:(NSZone *)zone { NSParameterAssert((entry != NULL) && (count <= capacity)); id stackKeys[count], stackObjects[count]; NSUInteger idx = 0UL, stackIdx = 0UL; for(idx = 0UL; idx < capacity; idx++) { if(entry[idx].key != NULL) { NSParameterAssert(stackIdx < count); stackKeys[stackIdx] = entry[idx].key; stackObjects[stackIdx++] = entry[idx].object; } } NSParameterAssert(stackIdx == count); return([[NSMutableDictionary allocWithZone:zone] initWithObjects:stackObjects forKeys:stackKeys count:stackIdx]); } @end #pragma mark - JK_STATIC_INLINE size_t jk_min(size_t a, size_t b) { return((a < b) ? a : b); } JK_STATIC_INLINE size_t jk_max(size_t a, size_t b) { return((a > b) ? a : b); } JK_STATIC_INLINE JKHash calculateHash(JKHash currentHash, unsigned char c) { return(((currentHash << 5) + currentHash) + c); } static void jk_error(JKParseState *parseState, NSString *format, ...) { NSCParameterAssert((parseState != NULL) && (format != NULL)); va_list varArgsList; va_start(varArgsList, format); NSString *formatString = [[[NSString alloc] initWithFormat:format arguments:varArgsList] autorelease]; va_end(varArgsList); #if 0 const unsigned char *lineStart = parseState->stringBuffer.bytes.ptr + parseState->lineStartIndex; const unsigned char *lineEnd = lineStart; const unsigned char *atCharacterPtr = NULL; for(atCharacterPtr = lineStart; atCharacterPtr < JK_END_STRING_PTR(parseState); atCharacterPtr++) { lineEnd = atCharacterPtr; if(jk_parse_is_newline(parseState, atCharacterPtr)) { break; } } NSString *lineString = @"", *carretString = @""; if(lineStart < JK_END_STRING_PTR(parseState)) { lineString = [[[NSString alloc] initWithBytes:lineStart length:(lineEnd - lineStart) encoding:NSUTF8StringEncoding] autorelease]; carretString = [NSString stringWithFormat:@"%*.*s^", (int)(parseState->atIndex - parseState->lineStartIndex), (int)(parseState->atIndex - parseState->lineStartIndex), " "]; } #endif if(parseState->error == NULL) { parseState->error = [NSError errorWithDomain:@"JKErrorDomain" code:-1L userInfo: [NSDictionary dictionaryWithObjectsAndKeys: formatString, NSLocalizedDescriptionKey, [NSNumber numberWithUnsignedLong:parseState->atIndex], @"JKAtIndexKey", [NSNumber numberWithUnsignedLong:parseState->lineNumber], @"JKLineNumberKey", //lineString, @"JKErrorLine0Key", //carretString, @"JKErrorLine1Key", NULL]]; } } #pragma mark - #pragma mark Buffer and Object Stack management functions static void jk_managedBuffer_release(JKManagedBuffer *managedBuffer) { if((managedBuffer->flags & JKManagedBufferMustFree)) { if(managedBuffer->bytes.ptr != NULL) { free(managedBuffer->bytes.ptr); managedBuffer->bytes.ptr = NULL; } managedBuffer->flags &= ~JKManagedBufferMustFree; } managedBuffer->bytes.ptr = NULL; managedBuffer->bytes.length = 0UL; managedBuffer->flags &= ~JKManagedBufferLocationMask; } static void jk_managedBuffer_setToStackBuffer(JKManagedBuffer *managedBuffer, unsigned char *ptr, size_t length) { jk_managedBuffer_release(managedBuffer); managedBuffer->bytes.ptr = ptr; managedBuffer->bytes.length = length; managedBuffer->flags = (managedBuffer->flags & ~JKManagedBufferLocationMask) | JKManagedBufferOnStack; } static unsigned char *jk_managedBuffer_resize(JKManagedBuffer *managedBuffer, size_t newSize) { size_t roundedUpNewSize = newSize; if(managedBuffer->roundSizeUpToMultipleOf > 0UL) { roundedUpNewSize = newSize + ((managedBuffer->roundSizeUpToMultipleOf - (newSize % managedBuffer->roundSizeUpToMultipleOf)) % managedBuffer->roundSizeUpToMultipleOf); } if((roundedUpNewSize != managedBuffer->bytes.length) && (roundedUpNewSize > managedBuffer->bytes.length)) { if((managedBuffer->flags & JKManagedBufferLocationMask) == JKManagedBufferOnStack) { NSCParameterAssert((managedBuffer->flags & JKManagedBufferMustFree) == 0); unsigned char *newBuffer = NULL, *oldBuffer = managedBuffer->bytes.ptr; if((newBuffer = (unsigned char *)malloc(roundedUpNewSize)) == NULL) { return(NULL); } memcpy(newBuffer, oldBuffer, jk_min(managedBuffer->bytes.length, roundedUpNewSize)); managedBuffer->flags = (managedBuffer->flags & ~JKManagedBufferLocationMask) | (JKManagedBufferOnHeap | JKManagedBufferMustFree); managedBuffer->bytes.ptr = newBuffer; managedBuffer->bytes.length = roundedUpNewSize; } else { NSCParameterAssert(((managedBuffer->flags & JKManagedBufferMustFree) != 0) && ((managedBuffer->flags & JKManagedBufferLocationMask) == JKManagedBufferOnHeap)); if((managedBuffer->bytes.ptr = (unsigned char *)reallocf(managedBuffer->bytes.ptr, roundedUpNewSize)) == NULL) { return(NULL); } managedBuffer->bytes.length = roundedUpNewSize; } } return(managedBuffer->bytes.ptr); } static void jk_objectStack_release(JKObjectStack *objectStack) { NSCParameterAssert(objectStack != NULL); NSCParameterAssert(objectStack->index <= objectStack->count); size_t atIndex = 0UL; for(atIndex = 0UL; atIndex < objectStack->index; atIndex++) { if(objectStack->objects[atIndex] != NULL) { CFRelease(objectStack->objects[atIndex]); objectStack->objects[atIndex] = NULL; } if(objectStack->keys[atIndex] != NULL) { CFRelease(objectStack->keys[atIndex]); objectStack->keys[atIndex] = NULL; } } objectStack->index = 0UL; if(objectStack->flags & JKObjectStackMustFree) { NSCParameterAssert((objectStack->flags & JKObjectStackLocationMask) == JKObjectStackOnHeap); if(objectStack->objects != NULL) { free(objectStack->objects); objectStack->objects = NULL; } if(objectStack->keys != NULL) { free(objectStack->keys); objectStack->keys = NULL; } if(objectStack->cfHashes != NULL) { free(objectStack->cfHashes); objectStack->cfHashes = NULL; } objectStack->flags &= ~JKObjectStackMustFree; } objectStack->objects = NULL; objectStack->keys = NULL; objectStack->cfHashes = NULL; objectStack->count = 0UL; objectStack->flags &= ~JKObjectStackLocationMask; } static void jk_objectStack_setToStackBuffer(JKObjectStack *objectStack, void **objects, void **keys, CFHashCode *cfHashes, size_t count) { NSCParameterAssert((objectStack != NULL) && (objects != NULL) && (keys != NULL) && (cfHashes != NULL) && (count > 0UL)); jk_objectStack_release(objectStack); objectStack->objects = objects; objectStack->keys = keys; objectStack->cfHashes = cfHashes; objectStack->count = count; objectStack->flags = (objectStack->flags & ~JKObjectStackLocationMask) | JKObjectStackOnStack; #ifndef NS_BLOCK_ASSERTIONS size_t idx; for(idx = 0UL; idx < objectStack->count; idx++) { objectStack->objects[idx] = NULL; objectStack->keys[idx] = NULL; objectStack->cfHashes[idx] = 0UL; } #endif } static int jk_objectStack_resize(JKObjectStack *objectStack, size_t newCount) { size_t roundedUpNewCount = newCount; int returnCode = 0; void **newObjects = NULL, **newKeys = NULL; CFHashCode *newCFHashes = NULL; if(objectStack->roundSizeUpToMultipleOf > 0UL) { roundedUpNewCount = newCount + ((objectStack->roundSizeUpToMultipleOf - (newCount % objectStack->roundSizeUpToMultipleOf)) % objectStack->roundSizeUpToMultipleOf); } if((roundedUpNewCount != objectStack->count) && (roundedUpNewCount > objectStack->count)) { if((objectStack->flags & JKObjectStackLocationMask) == JKObjectStackOnStack) { NSCParameterAssert((objectStack->flags & JKObjectStackMustFree) == 0); if((newObjects = (void ** )calloc(1UL, roundedUpNewCount * sizeof(void * ))) == NULL) { returnCode = 1; goto errorExit; } memcpy(newObjects, objectStack->objects, jk_min(objectStack->count, roundedUpNewCount) * sizeof(void *)); if((newKeys = (void ** )calloc(1UL, roundedUpNewCount * sizeof(void * ))) == NULL) { returnCode = 1; goto errorExit; } memcpy(newKeys, objectStack->keys, jk_min(objectStack->count, roundedUpNewCount) * sizeof(void *)); if((newCFHashes = (CFHashCode *)calloc(1UL, roundedUpNewCount * sizeof(CFHashCode))) == NULL) { returnCode = 1; goto errorExit; } memcpy(newCFHashes, objectStack->cfHashes, jk_min(objectStack->count, roundedUpNewCount) * sizeof(CFHashCode)); objectStack->flags = (objectStack->flags & ~JKObjectStackLocationMask) | (JKObjectStackOnHeap | JKObjectStackMustFree); objectStack->objects = newObjects; newObjects = NULL; objectStack->keys = newKeys; newKeys = NULL; objectStack->cfHashes = newCFHashes; newCFHashes = NULL; objectStack->count = roundedUpNewCount; } else { NSCParameterAssert(((objectStack->flags & JKObjectStackMustFree) != 0) && ((objectStack->flags & JKObjectStackLocationMask) == JKObjectStackOnHeap)); if((newObjects = (void ** )realloc(objectStack->objects, roundedUpNewCount * sizeof(void * ))) != NULL) { objectStack->objects = newObjects; newObjects = NULL; } else { returnCode = 1; goto errorExit; } if((newKeys = (void ** )realloc(objectStack->keys, roundedUpNewCount * sizeof(void * ))) != NULL) { objectStack->keys = newKeys; newKeys = NULL; } else { returnCode = 1; goto errorExit; } if((newCFHashes = (CFHashCode *)realloc(objectStack->cfHashes, roundedUpNewCount * sizeof(CFHashCode))) != NULL) { objectStack->cfHashes = newCFHashes; newCFHashes = NULL; } else { returnCode = 1; goto errorExit; } #ifndef NS_BLOCK_ASSERTIONS size_t idx; for(idx = objectStack->count; idx < roundedUpNewCount; idx++) { objectStack->objects[idx] = NULL; objectStack->keys[idx] = NULL; objectStack->cfHashes[idx] = 0UL; } #endif objectStack->count = roundedUpNewCount; } } errorExit: if(newObjects != NULL) { free(newObjects); newObjects = NULL; } if(newKeys != NULL) { free(newKeys); newKeys = NULL; } if(newCFHashes != NULL) { free(newCFHashes); newCFHashes = NULL; } return(returnCode); } //////////// #pragma mark - #pragma mark Unicode related functions JK_STATIC_INLINE ConversionResult isValidCodePoint(UTF32 *u32CodePoint) { ConversionResult result = conversionOK; UTF32 ch = *u32CodePoint; if(JK_EXPECT_F(ch >= UNI_SUR_HIGH_START) && (JK_EXPECT_T(ch <= UNI_SUR_LOW_END))) { result = sourceIllegal; ch = UNI_REPLACEMENT_CHAR; goto finished; } if(JK_EXPECT_F(ch >= 0xFDD0U) && (JK_EXPECT_F(ch <= 0xFDEFU) || JK_EXPECT_F((ch & 0xFFFEU) == 0xFFFEU)) && JK_EXPECT_T(ch <= 0x10FFFFU)) { result = sourceIllegal; ch = UNI_REPLACEMENT_CHAR; goto finished; } if(JK_EXPECT_F(ch == 0U)) { result = sourceIllegal; ch = UNI_REPLACEMENT_CHAR; goto finished; } finished: *u32CodePoint = ch; return(result); } static int isLegalUTF8(const UTF8 *source, size_t length) { const UTF8 *srcptr = source + length; UTF8 a; switch(length) { default: return(0); // Everything else falls through when "true"... case 4: if(JK_EXPECT_F(((a = (*--srcptr)) < 0x80) || (a > 0xBF))) { return(0); } case 3: if(JK_EXPECT_F(((a = (*--srcptr)) < 0x80) || (a > 0xBF))) { return(0); } case 2: if(JK_EXPECT_F( (a = (*--srcptr)) > 0xBF )) { return(0); } switch(*source) { // no fall-through in this inner switch case 0xE0: if(JK_EXPECT_F(a < 0xA0)) { return(0); } break; case 0xED: if(JK_EXPECT_F(a > 0x9F)) { return(0); } break; case 0xF0: if(JK_EXPECT_F(a < 0x90)) { return(0); } break; case 0xF4: if(JK_EXPECT_F(a > 0x8F)) { return(0); } break; default: if(JK_EXPECT_F(a < 0x80)) { return(0); } } case 1: if(JK_EXPECT_F((JK_EXPECT_T(*source < 0xC2)) && JK_EXPECT_F(*source >= 0x80))) { return(0); } } if(JK_EXPECT_F(*source > 0xF4)) { return(0); } return(1); } static ConversionResult ConvertSingleCodePointInUTF8(const UTF8 *sourceStart, const UTF8 *sourceEnd, UTF8 const **nextUTF8, UTF32 *convertedUTF32) { ConversionResult result = conversionOK; const UTF8 *source = sourceStart; UTF32 ch = 0UL; #if !defined(JK_FAST_TRAILING_BYTES) unsigned short extraBytesToRead = trailingBytesForUTF8[*source]; #else unsigned short extraBytesToRead = __builtin_clz(((*source)^0xff) << 25); #endif if(JK_EXPECT_F((source + extraBytesToRead + 1) > sourceEnd) || JK_EXPECT_F(!isLegalUTF8(source, extraBytesToRead + 1))) { source++; while((source < sourceEnd) && (((*source) & 0xc0) == 0x80) && ((source - sourceStart) < (extraBytesToRead + 1))) { source++; } NSCParameterAssert(source <= sourceEnd); result = ((source < sourceEnd) && (((*source) & 0xc0) != 0x80)) ? sourceIllegal : ((sourceStart + extraBytesToRead + 1) > sourceEnd) ? sourceExhausted : sourceIllegal; ch = UNI_REPLACEMENT_CHAR; goto finished; } switch(extraBytesToRead) { // The cases all fall through. case 5: ch += *source++; ch <<= 6; case 4: ch += *source++; ch <<= 6; case 3: ch += *source++; ch <<= 6; case 2: ch += *source++; ch <<= 6; case 1: ch += *source++; ch <<= 6; case 0: ch += *source++; } ch -= offsetsFromUTF8[extraBytesToRead]; result = isValidCodePoint(&ch); finished: *nextUTF8 = source; *convertedUTF32 = ch; return(result); } static ConversionResult ConvertUTF32toUTF8 (UTF32 u32CodePoint, UTF8 **targetStart, UTF8 *targetEnd) { const UTF32 byteMask = 0xBF, byteMark = 0x80; ConversionResult result = conversionOK; UTF8 *target = *targetStart; UTF32 ch = u32CodePoint; unsigned short bytesToWrite = 0; result = isValidCodePoint(&ch); // Figure out how many bytes the result will require. Turn any illegally large UTF32 things (> Plane 17) into replacement chars. if(ch < (UTF32)0x80) { bytesToWrite = 1; } else if(ch < (UTF32)0x800) { bytesToWrite = 2; } else if(ch < (UTF32)0x10000) { bytesToWrite = 3; } else if(ch <= UNI_MAX_LEGAL_UTF32) { bytesToWrite = 4; } else { bytesToWrite = 3; ch = UNI_REPLACEMENT_CHAR; result = sourceIllegal; } target += bytesToWrite; if (target > targetEnd) { target -= bytesToWrite; result = targetExhausted; goto finished; } switch (bytesToWrite) { // note: everything falls through. case 4: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6; case 3: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6; case 2: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6; case 1: *--target = (UTF8) (ch | firstByteMark[bytesToWrite]); } target += bytesToWrite; finished: *targetStart = target; return(result); } JK_STATIC_INLINE int jk_string_add_unicodeCodePoint(JKParseState *parseState, uint32_t unicodeCodePoint, size_t *tokenBufferIdx, JKHash *stringHash) { UTF8 *u8s = &parseState->token.tokenBuffer.bytes.ptr[*tokenBufferIdx]; ConversionResult result; if((result = ConvertUTF32toUTF8(unicodeCodePoint, &u8s, (parseState->token.tokenBuffer.bytes.ptr + parseState->token.tokenBuffer.bytes.length))) != conversionOK) { if(result == targetExhausted) { return(1); } } size_t utf8len = u8s - &parseState->token.tokenBuffer.bytes.ptr[*tokenBufferIdx], nextIdx = (*tokenBufferIdx) + utf8len; while(*tokenBufferIdx < nextIdx) { *stringHash = calculateHash(*stringHash, parseState->token.tokenBuffer.bytes.ptr[(*tokenBufferIdx)++]); } return(0); } //////////// #pragma mark - #pragma mark Decoding / parsing / deserializing functions static int jk_parse_string(JKParseState *parseState) { NSCParameterAssert((parseState != NULL) && (JK_AT_STRING_PTR(parseState) <= JK_END_STRING_PTR(parseState))); const unsigned char *stringStart = JK_AT_STRING_PTR(parseState) + 1; const unsigned char *endOfBuffer = JK_END_STRING_PTR(parseState); const unsigned char *atStringCharacter = stringStart; unsigned char *tokenBuffer = parseState->token.tokenBuffer.bytes.ptr; size_t tokenStartIndex = parseState->atIndex; size_t tokenBufferIdx = 0UL; int onlySimpleString = 1, stringState = JSONStringStateStart; uint16_t escapedUnicode1 = 0U, escapedUnicode2 = 0U; uint32_t escapedUnicodeCodePoint = 0U; JKHash stringHash = JK_HASH_INIT; while(1) { unsigned long currentChar; if(JK_EXPECT_F(atStringCharacter == endOfBuffer)) { /* XXX Add error message */ stringState = JSONStringStateError; goto finishedParsing; } if(JK_EXPECT_F((currentChar = *atStringCharacter++) >= 0x80UL)) { const unsigned char *nextValidCharacter = NULL; UTF32 u32ch = 0U; ConversionResult result; if(JK_EXPECT_F((result = ConvertSingleCodePointInUTF8(atStringCharacter - 1, endOfBuffer, (UTF8 const **)&nextValidCharacter, &u32ch)) != conversionOK)) { goto switchToSlowPath; } stringHash = calculateHash(stringHash, currentChar); while(atStringCharacter < nextValidCharacter) { stringHash = calculateHash(stringHash, *atStringCharacter++); } continue; } else { if(JK_EXPECT_F(currentChar == (unsigned long)'"')) { stringState = JSONStringStateFinished; goto finishedParsing; } if(JK_EXPECT_F(currentChar == (unsigned long)'\\')) { switchToSlowPath: onlySimpleString = 0; stringState = JSONStringStateParsing; tokenBufferIdx = (atStringCharacter - stringStart) - 1L; if(JK_EXPECT_F((tokenBufferIdx + 16UL) > parseState->token.tokenBuffer.bytes.length)) { if((tokenBuffer = jk_managedBuffer_resize(&parseState->token.tokenBuffer, tokenBufferIdx + 1024UL)) == NULL) { jk_error(parseState, @"Internal error: Unable to resize temporary buffer. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); stringState = JSONStringStateError; goto finishedParsing; } } memcpy(tokenBuffer, stringStart, tokenBufferIdx); goto slowMatch; } if(JK_EXPECT_F(currentChar < 0x20UL)) { jk_error(parseState, @"Invalid character < 0x20 found in string: 0x%2.2x.", currentChar); stringState = JSONStringStateError; goto finishedParsing; } stringHash = calculateHash(stringHash, currentChar); } } slowMatch: for(atStringCharacter = (stringStart + ((atStringCharacter - stringStart) - 1L)); (atStringCharacter < endOfBuffer) && (tokenBufferIdx < parseState->token.tokenBuffer.bytes.length); atStringCharacter++) { if((tokenBufferIdx + 16UL) > parseState->token.tokenBuffer.bytes.length) { if((tokenBuffer = jk_managedBuffer_resize(&parseState->token.tokenBuffer, tokenBufferIdx + 1024UL)) == NULL) { jk_error(parseState, @"Internal error: Unable to resize temporary buffer. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); stringState = JSONStringStateError; goto finishedParsing; } } NSCParameterAssert(tokenBufferIdx < parseState->token.tokenBuffer.bytes.length); unsigned long currentChar = (*atStringCharacter), escapedChar; if(JK_EXPECT_T(stringState == JSONStringStateParsing)) { if(JK_EXPECT_T(currentChar >= 0x20UL)) { if(JK_EXPECT_T(currentChar < (unsigned long)0x80)) { // Not a UTF8 sequence if(JK_EXPECT_F(currentChar == (unsigned long)'\\')) { stringState = JSONStringStateEscape; continue; } if(JK_EXPECT_F(currentChar == (unsigned long)'"')) { stringState = JSONStringStateFinished; atStringCharacter++; goto finishedParsing; } stringHash = calculateHash(stringHash, currentChar); tokenBuffer[tokenBufferIdx++] = currentChar; continue; } else { // UTF8 sequence const unsigned char *nextValidCharacter = NULL; UTF32 u32ch = 0U; ConversionResult result; if(JK_EXPECT_F((result = ConvertSingleCodePointInUTF8(atStringCharacter, endOfBuffer, (UTF8 const **)&nextValidCharacter, &u32ch)) != conversionOK)) { if((result == sourceIllegal) && ((parseState->parseOptionFlags & JKParseOptionLooseUnicode) == 0)) { jk_error(parseState, @"Illegal UTF8 sequence found in \"\" string."); stringState = JSONStringStateError; goto finishedParsing; } if(result == sourceExhausted) { jk_error(parseState, @"End of buffer reached while parsing UTF8 in \"\" string."); stringState = JSONStringStateError; goto finishedParsing; } if(jk_string_add_unicodeCodePoint(parseState, u32ch, &tokenBufferIdx, &stringHash)) { jk_error(parseState, @"Internal error: Unable to add UTF8 sequence to internal string buffer. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); stringState = JSONStringStateError; goto finishedParsing; } atStringCharacter = nextValidCharacter - 1; continue; } else { while(atStringCharacter < nextValidCharacter) { tokenBuffer[tokenBufferIdx++] = *atStringCharacter; stringHash = calculateHash(stringHash, *atStringCharacter++); } atStringCharacter--; continue; } } } else { // currentChar < 0x20 jk_error(parseState, @"Invalid character < 0x20 found in string: 0x%2.2x.", currentChar); stringState = JSONStringStateError; goto finishedParsing; } } else { // stringState != JSONStringStateParsing int isSurrogate = 1; switch(stringState) { case JSONStringStateEscape: switch(currentChar) { case 'u': escapedUnicode1 = 0U; escapedUnicode2 = 0U; escapedUnicodeCodePoint = 0U; stringState = JSONStringStateEscapedUnicode1; break; case 'b': escapedChar = '\b'; goto parsedEscapedChar; case 'f': escapedChar = '\f'; goto parsedEscapedChar; case 'n': escapedChar = '\n'; goto parsedEscapedChar; case 'r': escapedChar = '\r'; goto parsedEscapedChar; case 't': escapedChar = '\t'; goto parsedEscapedChar; case '\\': escapedChar = '\\'; goto parsedEscapedChar; case '/': escapedChar = '/'; goto parsedEscapedChar; case '"': escapedChar = '"'; goto parsedEscapedChar; parsedEscapedChar: stringState = JSONStringStateParsing; stringHash = calculateHash(stringHash, escapedChar); tokenBuffer[tokenBufferIdx++] = escapedChar; break; default: jk_error(parseState, @"Invalid escape sequence found in \"\" string."); stringState = JSONStringStateError; goto finishedParsing; break; } break; case JSONStringStateEscapedUnicode1: case JSONStringStateEscapedUnicode2: case JSONStringStateEscapedUnicode3: case JSONStringStateEscapedUnicode4: isSurrogate = 0; case JSONStringStateEscapedUnicodeSurrogate1: case JSONStringStateEscapedUnicodeSurrogate2: case JSONStringStateEscapedUnicodeSurrogate3: case JSONStringStateEscapedUnicodeSurrogate4: { uint16_t hexValue = 0U; switch(currentChar) { case '0' ... '9': hexValue = currentChar - '0'; goto parsedHex; case 'a' ... 'f': hexValue = (currentChar - 'a') + 10U; goto parsedHex; case 'A' ... 'F': hexValue = (currentChar - 'A') + 10U; goto parsedHex; parsedHex: if(!isSurrogate) { escapedUnicode1 = (escapedUnicode1 << 4) | hexValue; } else { escapedUnicode2 = (escapedUnicode2 << 4) | hexValue; } if(stringState == JSONStringStateEscapedUnicode4) { if(((escapedUnicode1 >= 0xD800U) && (escapedUnicode1 < 0xE000U))) { if((escapedUnicode1 >= 0xD800U) && (escapedUnicode1 < 0xDC00U)) { stringState = JSONStringStateEscapedNeedEscapeForSurrogate; } else if((escapedUnicode1 >= 0xDC00U) && (escapedUnicode1 < 0xE000U)) { if((parseState->parseOptionFlags & JKParseOptionLooseUnicode)) { escapedUnicodeCodePoint = UNI_REPLACEMENT_CHAR; } else { jk_error(parseState, @"Illegal \\u Unicode escape sequence."); stringState = JSONStringStateError; goto finishedParsing; } } } else { escapedUnicodeCodePoint = escapedUnicode1; } } if(stringState == JSONStringStateEscapedUnicodeSurrogate4) { if((escapedUnicode2 < 0xdc00) || (escapedUnicode2 > 0xdfff)) { if((parseState->parseOptionFlags & JKParseOptionLooseUnicode)) { escapedUnicodeCodePoint = UNI_REPLACEMENT_CHAR; } else { jk_error(parseState, @"Illegal \\u Unicode escape sequence."); stringState = JSONStringStateError; goto finishedParsing; } } else { escapedUnicodeCodePoint = ((escapedUnicode1 - 0xd800) * 0x400) + (escapedUnicode2 - 0xdc00) + 0x10000; } } if((stringState == JSONStringStateEscapedUnicode4) || (stringState == JSONStringStateEscapedUnicodeSurrogate4)) { if((isValidCodePoint(&escapedUnicodeCodePoint) == sourceIllegal) && ((parseState->parseOptionFlags & JKParseOptionLooseUnicode) == 0)) { jk_error(parseState, @"Illegal \\u Unicode escape sequence."); stringState = JSONStringStateError; goto finishedParsing; } stringState = JSONStringStateParsing; if(jk_string_add_unicodeCodePoint(parseState, escapedUnicodeCodePoint, &tokenBufferIdx, &stringHash)) { jk_error(parseState, @"Internal error: Unable to add UTF8 sequence to internal string buffer. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); stringState = JSONStringStateError; goto finishedParsing; } } else if((stringState >= JSONStringStateEscapedUnicode1) && (stringState <= JSONStringStateEscapedUnicodeSurrogate4)) { stringState++; } break; default: jk_error(parseState, @"Unexpected character found in \\u Unicode escape sequence. Found '%c', expected [0-9a-fA-F].", currentChar); stringState = JSONStringStateError; goto finishedParsing; break; } } break; case JSONStringStateEscapedNeedEscapeForSurrogate: if((currentChar == '\\')) { stringState = JSONStringStateEscapedNeedEscapedUForSurrogate; } else { if((parseState->parseOptionFlags & JKParseOptionLooseUnicode) == 0) { jk_error(parseState, @"Required a second \\u Unicode escape sequence following a surrogate \\u Unicode escape sequence."); stringState = JSONStringStateError; goto finishedParsing; } else { stringState = JSONStringStateParsing; atStringCharacter--; if(jk_string_add_unicodeCodePoint(parseState, UNI_REPLACEMENT_CHAR, &tokenBufferIdx, &stringHash)) { jk_error(parseState, @"Internal error: Unable to add UTF8 sequence to internal string buffer. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); stringState = JSONStringStateError; goto finishedParsing; } } } break; case JSONStringStateEscapedNeedEscapedUForSurrogate: if(currentChar == 'u') { stringState = JSONStringStateEscapedUnicodeSurrogate1; } else { if((parseState->parseOptionFlags & JKParseOptionLooseUnicode) == 0) { jk_error(parseState, @"Required a second \\u Unicode escape sequence following a surrogate \\u Unicode escape sequence."); stringState = JSONStringStateError; goto finishedParsing; } else { stringState = JSONStringStateParsing; atStringCharacter -= 2; if(jk_string_add_unicodeCodePoint(parseState, UNI_REPLACEMENT_CHAR, &tokenBufferIdx, &stringHash)) { jk_error(parseState, @"Internal error: Unable to add UTF8 sequence to internal string buffer. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); stringState = JSONStringStateError; goto finishedParsing; } } } break; default: jk_error(parseState, @"Internal error: Unknown stringState. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); stringState = JSONStringStateError; goto finishedParsing; break; } } } finishedParsing: if(JK_EXPECT_T(stringState == JSONStringStateFinished)) { NSCParameterAssert((parseState->stringBuffer.bytes.ptr + tokenStartIndex) < atStringCharacter); parseState->token.tokenPtrRange.ptr = parseState->stringBuffer.bytes.ptr + tokenStartIndex; parseState->token.tokenPtrRange.length = (atStringCharacter - parseState->token.tokenPtrRange.ptr); if(JK_EXPECT_T(onlySimpleString)) { NSCParameterAssert(((parseState->token.tokenPtrRange.ptr + 1) < endOfBuffer) && (parseState->token.tokenPtrRange.length >= 2UL) && (((parseState->token.tokenPtrRange.ptr + 1) + (parseState->token.tokenPtrRange.length - 2)) < endOfBuffer)); parseState->token.value.ptrRange.ptr = parseState->token.tokenPtrRange.ptr + 1; parseState->token.value.ptrRange.length = parseState->token.tokenPtrRange.length - 2UL; } else { parseState->token.value.ptrRange.ptr = parseState->token.tokenBuffer.bytes.ptr; parseState->token.value.ptrRange.length = tokenBufferIdx; } parseState->token.value.hash = stringHash; parseState->token.value.type = JKValueTypeString; parseState->atIndex = (atStringCharacter - parseState->stringBuffer.bytes.ptr); } if(JK_EXPECT_F(stringState != JSONStringStateFinished)) { jk_error(parseState, @"Invalid string."); } return(JK_EXPECT_T(stringState == JSONStringStateFinished) ? 0 : 1); } static int jk_parse_number(JKParseState *parseState) { NSCParameterAssert((parseState != NULL) && (JK_AT_STRING_PTR(parseState) <= JK_END_STRING_PTR(parseState))); const unsigned char *numberStart = JK_AT_STRING_PTR(parseState); const unsigned char *endOfBuffer = JK_END_STRING_PTR(parseState); const unsigned char *atNumberCharacter = NULL; int numberState = JSONNumberStateWholeNumberStart, isFloatingPoint = 0, isNegative = 0, backup = 0; size_t startingIndex = parseState->atIndex; for(atNumberCharacter = numberStart; (JK_EXPECT_T(atNumberCharacter < endOfBuffer)) && (JK_EXPECT_T(!(JK_EXPECT_F(numberState == JSONNumberStateFinished) || JK_EXPECT_F(numberState == JSONNumberStateError)))); atNumberCharacter++) { unsigned long currentChar = (unsigned long)(*atNumberCharacter), lowerCaseCC = currentChar | 0x20UL; switch(numberState) { case JSONNumberStateWholeNumberStart: if (currentChar == '-') { numberState = JSONNumberStateWholeNumberMinus; isNegative = 1; break; } case JSONNumberStateWholeNumberMinus: if (currentChar == '0') { numberState = JSONNumberStateWholeNumberZero; break; } else if( (currentChar >= '1') && (currentChar <= '9')) { numberState = JSONNumberStateWholeNumber; break; } else { /* XXX Add error message */ numberState = JSONNumberStateError; break; } case JSONNumberStateExponentStart: if( (currentChar == '+') || (currentChar == '-')) { numberState = JSONNumberStateExponentPlusMinus; break; } case JSONNumberStateFractionalNumberStart: case JSONNumberStateExponentPlusMinus:if(!((currentChar >= '0') && (currentChar <= '9'))) { /* XXX Add error message */ numberState = JSONNumberStateError; break; } else { if(numberState == JSONNumberStateFractionalNumberStart) { numberState = JSONNumberStateFractionalNumber; } else { numberState = JSONNumberStateExponent; } break; } case JSONNumberStateWholeNumberZero: case JSONNumberStateWholeNumber: if (currentChar == '.') { numberState = JSONNumberStateFractionalNumberStart; isFloatingPoint = 1; break; } case JSONNumberStateFractionalNumber: if (lowerCaseCC == 'e') { numberState = JSONNumberStateExponentStart; isFloatingPoint = 1; break; } case JSONNumberStateExponent: if(!((currentChar >= '0') && (currentChar <= '9')) || (numberState == JSONNumberStateWholeNumberZero)) { numberState = JSONNumberStateFinished; backup = 1; break; } break; default: /* XXX Add error message */ numberState = JSONNumberStateError; break; } } parseState->token.tokenPtrRange.ptr = parseState->stringBuffer.bytes.ptr + startingIndex; parseState->token.tokenPtrRange.length = (atNumberCharacter - parseState->token.tokenPtrRange.ptr) - backup; parseState->atIndex = (parseState->token.tokenPtrRange.ptr + parseState->token.tokenPtrRange.length) - parseState->stringBuffer.bytes.ptr; if(JK_EXPECT_T(numberState == JSONNumberStateFinished)) { unsigned char numberTempBuf[parseState->token.tokenPtrRange.length + 4UL]; unsigned char *endOfNumber = NULL; memcpy(numberTempBuf, parseState->token.tokenPtrRange.ptr, parseState->token.tokenPtrRange.length); numberTempBuf[parseState->token.tokenPtrRange.length] = 0; errno = 0; // Treat "-0" as a floating point number, which is capable of representing negative zeros. if(JK_EXPECT_F(parseState->token.tokenPtrRange.length == 2UL) && JK_EXPECT_F(numberTempBuf[1] == '0') && JK_EXPECT_F(isNegative)) { isFloatingPoint = 1; } if(isFloatingPoint) { parseState->token.value.number.doubleValue = strtod((const char *)numberTempBuf, (char **)&endOfNumber); parseState->token.value.type = JKValueTypeDouble; parseState->token.value.ptrRange.ptr = (const unsigned char *)&parseState->token.value.number.doubleValue; parseState->token.value.ptrRange.length = sizeof(double); parseState->token.value.hash = (JK_HASH_INIT + parseState->token.value.type); } else { if(isNegative) { parseState->token.value.number.longLongValue = strtoll((const char *)numberTempBuf, (char **)&endOfNumber, 10); parseState->token.value.type = JKValueTypeLongLong; parseState->token.value.ptrRange.ptr = (const unsigned char *)&parseState->token.value.number.longLongValue; parseState->token.value.ptrRange.length = sizeof(long long); parseState->token.value.hash = (JK_HASH_INIT + parseState->token.value.type) + (JKHash)parseState->token.value.number.longLongValue; } else { parseState->token.value.number.unsignedLongLongValue = strtoull((const char *)numberTempBuf, (char **)&endOfNumber, 10); parseState->token.value.type = JKValueTypeUnsignedLongLong; parseState->token.value.ptrRange.ptr = (const unsigned char *)&parseState->token.value.number.unsignedLongLongValue; parseState->token.value.ptrRange.length = sizeof(unsigned long long); parseState->token.value.hash = (JK_HASH_INIT + parseState->token.value.type) + (JKHash)parseState->token.value.number.unsignedLongLongValue; } } if(JK_EXPECT_F(errno != 0)) { numberState = JSONNumberStateError; if(errno == ERANGE) { switch(parseState->token.value.type) { case JKValueTypeDouble: jk_error(parseState, @"The value '%s' could not be represented as a 'double' due to %s.", numberTempBuf, (parseState->token.value.number.doubleValue == 0.0) ? "underflow" : "overflow"); break; case JKValueTypeLongLong: jk_error(parseState, @"The value '%s' exceeded the minimum value that could be represented: %lld.", numberTempBuf, parseState->token.value.number.longLongValue); break; case JKValueTypeUnsignedLongLong: jk_error(parseState, @"The value '%s' exceeded the maximum value that could be represented: %llu.", numberTempBuf, parseState->token.value.number.unsignedLongLongValue); break; default: jk_error(parseState, @"Internal error: Unknown token value type. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); break; } } } if(JK_EXPECT_F(endOfNumber != &numberTempBuf[parseState->token.tokenPtrRange.length]) && JK_EXPECT_F(numberState != JSONNumberStateError)) { numberState = JSONNumberStateError; jk_error(parseState, @"The conversion function did not consume all of the number tokens characters."); } size_t hashIndex = 0UL; for(hashIndex = 0UL; hashIndex < parseState->token.value.ptrRange.length; hashIndex++) { parseState->token.value.hash = calculateHash(parseState->token.value.hash, parseState->token.value.ptrRange.ptr[hashIndex]); } } if(JK_EXPECT_F(numberState != JSONNumberStateFinished)) { jk_error(parseState, @"Invalid number."); } return(JK_EXPECT_T((numberState == JSONNumberStateFinished)) ? 0 : 1); } JK_STATIC_INLINE void jk_set_parsed_token(JKParseState *parseState, const unsigned char *ptr, size_t length, JKTokenType type, size_t advanceBy) { parseState->token.tokenPtrRange.ptr = ptr; parseState->token.tokenPtrRange.length = length; parseState->token.type = type; parseState->atIndex += advanceBy; } static size_t jk_parse_is_newline(JKParseState *parseState, const unsigned char *atCharacterPtr) { NSCParameterAssert((parseState != NULL) && (atCharacterPtr != NULL) && (atCharacterPtr >= parseState->stringBuffer.bytes.ptr) && (atCharacterPtr < JK_END_STRING_PTR(parseState))); const unsigned char *endOfStringPtr = JK_END_STRING_PTR(parseState); if(JK_EXPECT_F(atCharacterPtr >= endOfStringPtr)) { return(0UL); } if(JK_EXPECT_F((*(atCharacterPtr + 0)) == '\n')) { return(1UL); } if(JK_EXPECT_F((*(atCharacterPtr + 0)) == '\r')) { if((JK_EXPECT_T((atCharacterPtr + 1) < endOfStringPtr)) && ((*(atCharacterPtr + 1)) == '\n')) { return(2UL); } return(1UL); } if(parseState->parseOptionFlags & JKParseOptionUnicodeNewlines) { if((JK_EXPECT_F((*(atCharacterPtr + 0)) == 0xc2)) && (((atCharacterPtr + 1) < endOfStringPtr) && ((*(atCharacterPtr + 1)) == 0x85))) { return(2UL); } if((JK_EXPECT_F((*(atCharacterPtr + 0)) == 0xe2)) && (((atCharacterPtr + 2) < endOfStringPtr) && ((*(atCharacterPtr + 1)) == 0x80) && (((*(atCharacterPtr + 2)) == 0xa8) || ((*(atCharacterPtr + 2)) == 0xa9)))) { return(3UL); } } return(0UL); } JK_STATIC_INLINE int jk_parse_skip_newline(JKParseState *parseState) { size_t newlineAdvanceAtIndex = 0UL; if(JK_EXPECT_F((newlineAdvanceAtIndex = jk_parse_is_newline(parseState, JK_AT_STRING_PTR(parseState))) > 0UL)) { parseState->lineNumber++; parseState->atIndex += (newlineAdvanceAtIndex - 1UL); parseState->lineStartIndex = parseState->atIndex + 1UL; return(1); } return(0); } JK_STATIC_INLINE void jk_parse_skip_whitespace(JKParseState *parseState) { #ifndef __clang_analyzer__ NSCParameterAssert((parseState != NULL) && (JK_AT_STRING_PTR(parseState) <= JK_END_STRING_PTR(parseState))); const unsigned char *atCharacterPtr = NULL; const unsigned char *endOfStringPtr = JK_END_STRING_PTR(parseState); for(atCharacterPtr = JK_AT_STRING_PTR(parseState); (JK_EXPECT_T((atCharacterPtr = JK_AT_STRING_PTR(parseState)) < endOfStringPtr)); parseState->atIndex++) { if(((*(atCharacterPtr + 0)) == ' ') || ((*(atCharacterPtr + 0)) == '\t')) { continue; } if(jk_parse_skip_newline(parseState)) { continue; } if(parseState->parseOptionFlags & JKParseOptionComments) { if((JK_EXPECT_F((*(atCharacterPtr + 0)) == '/')) && (JK_EXPECT_T((atCharacterPtr + 1) < endOfStringPtr))) { if((*(atCharacterPtr + 1)) == '/') { parseState->atIndex++; for(atCharacterPtr = JK_AT_STRING_PTR(parseState); (JK_EXPECT_T((atCharacterPtr = JK_AT_STRING_PTR(parseState)) < endOfStringPtr)); parseState->atIndex++) { if(jk_parse_skip_newline(parseState)) { break; } } continue; } if((*(atCharacterPtr + 1)) == '*') { parseState->atIndex++; for(atCharacterPtr = JK_AT_STRING_PTR(parseState); (JK_EXPECT_T((atCharacterPtr = JK_AT_STRING_PTR(parseState)) < endOfStringPtr)); parseState->atIndex++) { if(jk_parse_skip_newline(parseState)) { continue; } if(((*(atCharacterPtr + 0)) == '*') && (((atCharacterPtr + 1) < endOfStringPtr) && ((*(atCharacterPtr + 1)) == '/'))) { parseState->atIndex++; break; } } continue; } } } break; } #endif } static int jk_parse_next_token(JKParseState *parseState) { NSCParameterAssert((parseState != NULL) && (JK_AT_STRING_PTR(parseState) <= JK_END_STRING_PTR(parseState))); const unsigned char *atCharacterPtr = NULL; const unsigned char *endOfStringPtr = JK_END_STRING_PTR(parseState); unsigned char currentCharacter = 0U; int stopParsing = 0; parseState->prev_atIndex = parseState->atIndex; parseState->prev_lineNumber = parseState->lineNumber; parseState->prev_lineStartIndex = parseState->lineStartIndex; jk_parse_skip_whitespace(parseState); if((JK_AT_STRING_PTR(parseState) == endOfStringPtr)) { stopParsing = 1; } if((JK_EXPECT_T(stopParsing == 0)) && (JK_EXPECT_T((atCharacterPtr = JK_AT_STRING_PTR(parseState)) < endOfStringPtr))) { currentCharacter = *atCharacterPtr; switch(currentCharacter) { case '{': jk_set_parsed_token(parseState, atCharacterPtr, 1UL, JKTokenTypeObjectBegin, 1UL); break; case '}': jk_set_parsed_token(parseState, atCharacterPtr, 1UL, JKTokenTypeObjectEnd, 1UL); break; case '[': jk_set_parsed_token(parseState, atCharacterPtr, 1UL, JKTokenTypeArrayBegin, 1UL); break; case ']': jk_set_parsed_token(parseState, atCharacterPtr, 1UL, JKTokenTypeArrayEnd, 1UL); break; case ',': jk_set_parsed_token(parseState, atCharacterPtr, 1UL, JKTokenTypeComma, 1UL); break; case ':': jk_set_parsed_token(parseState, atCharacterPtr, 1UL, JKTokenTypeSeparator, 1UL); break; case 't': if(!((JK_EXPECT_T((atCharacterPtr + 4UL) < endOfStringPtr)) && (JK_EXPECT_T(atCharacterPtr[1] == 'r')) && (JK_EXPECT_T(atCharacterPtr[2] == 'u')) && (JK_EXPECT_T(atCharacterPtr[3] == 'e')))) { stopParsing = 1; /* XXX Add error message */ } else { jk_set_parsed_token(parseState, atCharacterPtr, 4UL, JKTokenTypeTrue, 4UL); } break; case 'f': if(!((JK_EXPECT_T((atCharacterPtr + 5UL) < endOfStringPtr)) && (JK_EXPECT_T(atCharacterPtr[1] == 'a')) && (JK_EXPECT_T(atCharacterPtr[2] == 'l')) && (JK_EXPECT_T(atCharacterPtr[3] == 's')) && (JK_EXPECT_T(atCharacterPtr[4] == 'e')))) { stopParsing = 1; /* XXX Add error message */ } else { jk_set_parsed_token(parseState, atCharacterPtr, 5UL, JKTokenTypeFalse, 5UL); } break; case 'n': if(!((JK_EXPECT_T((atCharacterPtr + 4UL) < endOfStringPtr)) && (JK_EXPECT_T(atCharacterPtr[1] == 'u')) && (JK_EXPECT_T(atCharacterPtr[2] == 'l')) && (JK_EXPECT_T(atCharacterPtr[3] == 'l')))) { stopParsing = 1; /* XXX Add error message */ } else { jk_set_parsed_token(parseState, atCharacterPtr, 4UL, JKTokenTypeNull, 4UL); } break; case '"': if(JK_EXPECT_T((stopParsing = jk_parse_string(parseState)) == 0)) { jk_set_parsed_token(parseState, parseState->token.tokenPtrRange.ptr, parseState->token.tokenPtrRange.length, JKTokenTypeString, 0UL); } break; case '-': // fall-thru case '0' ... '9': if(JK_EXPECT_T((stopParsing = jk_parse_number(parseState)) == 0)) { jk_set_parsed_token(parseState, parseState->token.tokenPtrRange.ptr, parseState->token.tokenPtrRange.length, JKTokenTypeNumber, 0UL); } break; default: stopParsing = 1; /* XXX Add error message */ break; } } if(JK_EXPECT_F(stopParsing)) { jk_error(parseState, @"Unexpected token, wanted '{', '}', '[', ']', ',', ':', 'true', 'false', 'null', '\"STRING\"', 'NUMBER'."); } return(stopParsing); } static void jk_error_parse_accept_or3(JKParseState *parseState, int state, NSString *or1String, NSString *or2String, NSString *or3String) { NSString *acceptStrings[16]; int acceptIdx = 0; if(state & JKParseAcceptValue) { acceptStrings[acceptIdx++] = or1String; } if(state & JKParseAcceptComma) { acceptStrings[acceptIdx++] = or2String; } if(state & JKParseAcceptEnd) { acceptStrings[acceptIdx++] = or3String; } if(acceptIdx == 1) { jk_error(parseState, @"Expected %@, not '%*.*s'", acceptStrings[0], (int)parseState->token.tokenPtrRange.length, (int)parseState->token.tokenPtrRange.length, parseState->token.tokenPtrRange.ptr); } else if(acceptIdx == 2) { jk_error(parseState, @"Expected %@ or %@, not '%*.*s'", acceptStrings[0], acceptStrings[1], (int)parseState->token.tokenPtrRange.length, (int)parseState->token.tokenPtrRange.length, parseState->token.tokenPtrRange.ptr); } else if(acceptIdx == 3) { jk_error(parseState, @"Expected %@, %@, or %@, not '%*.*s", acceptStrings[0], acceptStrings[1], acceptStrings[2], (int)parseState->token.tokenPtrRange.length, (int)parseState->token.tokenPtrRange.length, parseState->token.tokenPtrRange.ptr); } } static void *jk_parse_array(JKParseState *parseState) { size_t startingObjectIndex = parseState->objectStack.index; int arrayState = JKParseAcceptValueOrEnd, stopParsing = 0; void *parsedArray = NULL; while(JK_EXPECT_T((JK_EXPECT_T(stopParsing == 0)) && (JK_EXPECT_T(parseState->atIndex < parseState->stringBuffer.bytes.length)))) { if(JK_EXPECT_F(parseState->objectStack.index > (parseState->objectStack.count - 4UL))) { if(jk_objectStack_resize(&parseState->objectStack, parseState->objectStack.count + 128UL)) { jk_error(parseState, @"Internal error: [array] objectsIndex > %zu, resize failed? %@ line %#ld", (parseState->objectStack.count - 4UL), [NSString stringWithUTF8String:__FILE__], (long)__LINE__); break; } } if(JK_EXPECT_T((stopParsing = jk_parse_next_token(parseState)) == 0)) { void *object = NULL; #ifndef NS_BLOCK_ASSERTIONS parseState->objectStack.objects[parseState->objectStack.index] = NULL; parseState->objectStack.keys [parseState->objectStack.index] = NULL; #endif switch(parseState->token.type) { case JKTokenTypeNumber: case JKTokenTypeString: case JKTokenTypeTrue: case JKTokenTypeFalse: case JKTokenTypeNull: case JKTokenTypeArrayBegin: case JKTokenTypeObjectBegin: if(JK_EXPECT_F((arrayState & JKParseAcceptValue) == 0)) { parseState->errorIsPrev = 1; jk_error(parseState, @"Unexpected value."); stopParsing = 1; break; } if(JK_EXPECT_F((object = jk_object_for_token(parseState)) == NULL)) { jk_error(parseState, @"Internal error: Object == NULL"); stopParsing = 1; break; } else { parseState->objectStack.objects[parseState->objectStack.index++] = object; arrayState = JKParseAcceptCommaOrEnd; } break; case JKTokenTypeArrayEnd: if(JK_EXPECT_T(arrayState & JKParseAcceptEnd)) { NSCParameterAssert(parseState->objectStack.index >= startingObjectIndex); parsedArray = (void *)_JKArrayCreate((id *)&parseState->objectStack.objects[startingObjectIndex], (parseState->objectStack.index - startingObjectIndex)); } else { parseState->errorIsPrev = 1; jk_error(parseState, @"Unexpected ']'."); } stopParsing = 1; break; case JKTokenTypeComma: if(JK_EXPECT_T(arrayState & JKParseAcceptComma)) { arrayState = JKParseAcceptValue; } else { parseState->errorIsPrev = 1; jk_error(parseState, @"Unexpected ','."); stopParsing = 1; } break; default: parseState->errorIsPrev = 1; jk_error_parse_accept_or3(parseState, arrayState, @"a value", @"a comma", @"a ']'"); stopParsing = 1; break; } } } if(JK_EXPECT_F(parsedArray == NULL)) { size_t idx = 0UL; for(idx = startingObjectIndex; idx < parseState->objectStack.index; idx++) { if(parseState->objectStack.objects[idx] != NULL) { CFRelease(parseState->objectStack.objects[idx]); parseState->objectStack.objects[idx] = NULL; } } } #if !defined(NS_BLOCK_ASSERTIONS) else { size_t idx = 0UL; for(idx = startingObjectIndex; idx < parseState->objectStack.index; idx++) { parseState->objectStack.objects[idx] = NULL; parseState->objectStack.keys[idx] = NULL; } } #endif parseState->objectStack.index = startingObjectIndex; return(parsedArray); } static void *jk_create_dictionary(JKParseState *parseState, size_t startingObjectIndex) { void *parsedDictionary = NULL; parseState->objectStack.index--; parsedDictionary = _JKDictionaryCreate((id *)&parseState->objectStack.keys[startingObjectIndex], (NSUInteger *)&parseState->objectStack.cfHashes[startingObjectIndex], (id *)&parseState->objectStack.objects[startingObjectIndex], (parseState->objectStack.index - startingObjectIndex)); return(parsedDictionary); } static void *jk_parse_dictionary(JKParseState *parseState) { size_t startingObjectIndex = parseState->objectStack.index; int dictState = JKParseAcceptValueOrEnd, stopParsing = 0; void *parsedDictionary = NULL; while(JK_EXPECT_T((JK_EXPECT_T(stopParsing == 0)) && (JK_EXPECT_T(parseState->atIndex < parseState->stringBuffer.bytes.length)))) { if(JK_EXPECT_F(parseState->objectStack.index > (parseState->objectStack.count - 4UL))) { if(jk_objectStack_resize(&parseState->objectStack, parseState->objectStack.count + 128UL)) { jk_error(parseState, @"Internal error: [dictionary] objectsIndex > %zu, resize failed? %@ line #%ld", (parseState->objectStack.count - 4UL), [NSString stringWithUTF8String:__FILE__], (long)__LINE__); break; } } size_t objectStackIndex = parseState->objectStack.index++; parseState->objectStack.keys[objectStackIndex] = NULL; parseState->objectStack.objects[objectStackIndex] = NULL; void *key = NULL, *object = NULL; if(JK_EXPECT_T((JK_EXPECT_T(stopParsing == 0)) && (JK_EXPECT_T((stopParsing = jk_parse_next_token(parseState)) == 0)))) { switch(parseState->token.type) { case JKTokenTypeString: if(JK_EXPECT_F((dictState & JKParseAcceptValue) == 0)) { parseState->errorIsPrev = 1; jk_error(parseState, @"Unexpected string."); stopParsing = 1; break; } if(JK_EXPECT_F((key = jk_object_for_token(parseState)) == NULL)) { jk_error(parseState, @"Internal error: Key == NULL."); stopParsing = 1; break; } else { parseState->objectStack.keys[objectStackIndex] = key; if(JK_EXPECT_T(parseState->token.value.cacheItem != NULL)) { if((parseState->token.value.cacheItem->cfHash == 0UL)) { parseState->token.value.cacheItem->cfHash = CFHash(key); } parseState->objectStack.cfHashes[objectStackIndex] = parseState->token.value.cacheItem->cfHash; } else { parseState->objectStack.cfHashes[objectStackIndex] = CFHash(key); } } break; case JKTokenTypeObjectEnd: if((JK_EXPECT_T(dictState & JKParseAcceptEnd))) { NSCParameterAssert(parseState->objectStack.index >= startingObjectIndex); parsedDictionary = jk_create_dictionary(parseState, startingObjectIndex); } else { parseState->errorIsPrev = 1; jk_error(parseState, @"Unexpected '}'."); } stopParsing = 1; break; case JKTokenTypeComma: if((JK_EXPECT_T(dictState & JKParseAcceptComma))) { dictState = JKParseAcceptValue; parseState->objectStack.index--; continue; } else { parseState->errorIsPrev = 1; jk_error(parseState, @"Unexpected ','."); stopParsing = 1; } break; default: parseState->errorIsPrev = 1; jk_error_parse_accept_or3(parseState, dictState, @"a \"STRING\"", @"a comma", @"a '}'"); stopParsing = 1; break; } } if(JK_EXPECT_T(stopParsing == 0)) { if(JK_EXPECT_T((stopParsing = jk_parse_next_token(parseState)) == 0)) { if(JK_EXPECT_F(parseState->token.type != JKTokenTypeSeparator)) { parseState->errorIsPrev = 1; jk_error(parseState, @"Expected ':'."); stopParsing = 1; } } } if((JK_EXPECT_T(stopParsing == 0)) && (JK_EXPECT_T((stopParsing = jk_parse_next_token(parseState)) == 0))) { switch(parseState->token.type) { case JKTokenTypeNumber: case JKTokenTypeString: case JKTokenTypeTrue: case JKTokenTypeFalse: case JKTokenTypeNull: case JKTokenTypeArrayBegin: case JKTokenTypeObjectBegin: if(JK_EXPECT_F((dictState & JKParseAcceptValue) == 0)) { parseState->errorIsPrev = 1; jk_error(parseState, @"Unexpected value."); stopParsing = 1; break; } if(JK_EXPECT_F((object = jk_object_for_token(parseState)) == NULL)) { jk_error(parseState, @"Internal error: Object == NULL."); stopParsing = 1; break; } else { parseState->objectStack.objects[objectStackIndex] = object; dictState = JKParseAcceptCommaOrEnd; } break; default: parseState->errorIsPrev = 1; jk_error_parse_accept_or3(parseState, dictState, @"a value", @"a comma", @"a '}'"); stopParsing = 1; break; } } } if(JK_EXPECT_F(parsedDictionary == NULL)) { size_t idx = 0UL; for(idx = startingObjectIndex; idx < parseState->objectStack.index; idx++) { if(parseState->objectStack.keys[idx] != NULL) { CFRelease(parseState->objectStack.keys[idx]); parseState->objectStack.keys[idx] = NULL; } if(parseState->objectStack.objects[idx] != NULL) { CFRelease(parseState->objectStack.objects[idx]); parseState->objectStack.objects[idx] = NULL; } } } #if !defined(NS_BLOCK_ASSERTIONS) else { size_t idx = 0UL; for(idx = startingObjectIndex; idx < parseState->objectStack.index; idx++) { parseState->objectStack.objects[idx] = NULL; parseState->objectStack.keys[idx] = NULL; } } #endif parseState->objectStack.index = startingObjectIndex; return(parsedDictionary); } static id json_parse_it(JKParseState *parseState) { id parsedObject = NULL; int stopParsing = 0; while((JK_EXPECT_T(stopParsing == 0)) && (JK_EXPECT_T(parseState->atIndex < parseState->stringBuffer.bytes.length))) { if((JK_EXPECT_T(stopParsing == 0)) && (JK_EXPECT_T((stopParsing = jk_parse_next_token(parseState)) == 0))) { switch(parseState->token.type) { case JKTokenTypeArrayBegin: case JKTokenTypeObjectBegin: parsedObject = [(id)jk_object_for_token(parseState) autorelease]; stopParsing = 1; break; default: jk_error(parseState, @"Expected either '[' or '{'."); stopParsing = 1; break; } } } NSCParameterAssert((parseState->objectStack.index == 0) && (JK_AT_STRING_PTR(parseState) <= JK_END_STRING_PTR(parseState))); if((parsedObject == NULL) && (JK_AT_STRING_PTR(parseState) == JK_END_STRING_PTR(parseState))) { jk_error(parseState, @"Reached the end of the buffer."); } if(parsedObject == NULL) { jk_error(parseState, @"Unable to parse JSON."); } if((parsedObject != NULL) && (JK_AT_STRING_PTR(parseState) < JK_END_STRING_PTR(parseState))) { jk_parse_skip_whitespace(parseState); if((parsedObject != NULL) && ((parseState->parseOptionFlags & JKParseOptionPermitTextAfterValidJSON) == 0) && (JK_AT_STRING_PTR(parseState) < JK_END_STRING_PTR(parseState))) { jk_error(parseState, @"A valid JSON object was parsed but there were additional non-white-space characters remaining."); parsedObject = NULL; } } return(parsedObject); } //////////// #pragma mark - #pragma mark Object cache // This uses a Galois Linear Feedback Shift Register (LFSR) PRNG to pick which item in the cache to age. It has a period of (2^32)-1. // NOTE: A LFSR *MUST* be initialized to a non-zero value. JK_STATIC_INLINE void jk_cache_age(JKParseState *parseState) { parseState->cache.prng_lfsr = (parseState->cache.prng_lfsr >> 1) ^ ((0U - (parseState->cache.prng_lfsr & 1U)) & 0x80200003U); parseState->cache.age[parseState->cache.prng_lfsr & (parseState->cache.count - 1UL)] >>= 1; } // The object cache is nothing more than a hash table with open addressing collision resolution that is bounded by JK_CACHE_PROBES attempts. // // The hash table is a linear C array of JKTokenCacheItem. The terms "item" and "bucket" are synonymous with the index in to the cache array, i.e. cache.items[bucket]. // // Items in the cache have an age associated with them. The age is the number of rightmost 1 bits, i.e. 0000 = 0, 0001 = 1, 0011 = 2, 0111 = 3, 1111 = 4. // This allows us to use left and right shifts to add or subtract from an items age. Add = (age << 1) | 1. Subtract = age >> 0. Subtract is synonymous with "age" (i.e., age an item). // The reason for this is it allows us to perform saturated adds and subtractions, which is branchless. // The primitive C type MUST be unsigned. It is currently a "char", which allows (at a minimum and in practice) 8 bits. // // A "useable bucket" is a bucket that is not in use (never populated), or has an age == 0. // // When an item is found in the cache, it's age is incremented. // If a useable bucket hasn't been found, the current item (bucket) is aged along with two random items. // // If a value is not found in the cache, and no useable bucket has been found, that value is not added to the cache. static void *jk_cachedObjects(JKParseState *parseState) { unsigned long bucket = parseState->token.value.hash & (parseState->cache.count - 1UL), setBucket = 0UL, useableBucket = 0UL, x = 0UL; void *parsedAtom = NULL; if(JK_EXPECT_F(parseState->token.value.ptrRange.length == 0UL) && JK_EXPECT_T(parseState->token.value.type == JKValueTypeString)) { return(@""); } for(x = 0UL; x < JK_CACHE_PROBES; x++) { if(JK_EXPECT_F(parseState->cache.items[bucket].object == NULL)) { setBucket = 1UL; useableBucket = bucket; break; } if((JK_EXPECT_T(parseState->cache.items[bucket].hash == parseState->token.value.hash)) && (JK_EXPECT_T(parseState->cache.items[bucket].size == parseState->token.value.ptrRange.length)) && (JK_EXPECT_T(parseState->cache.items[bucket].type == parseState->token.value.type)) && (JK_EXPECT_T(parseState->cache.items[bucket].bytes != NULL)) && (JK_EXPECT_T(strncmp((const char *)parseState->cache.items[bucket].bytes, (const char *)parseState->token.value.ptrRange.ptr, parseState->token.value.ptrRange.length) == 0U))) { parseState->cache.age[bucket] = (parseState->cache.age[bucket] << 1) | 1U; parseState->token.value.cacheItem = &parseState->cache.items[bucket]; NSCParameterAssert(parseState->cache.items[bucket].object != NULL); return((void *)CFRetain(parseState->cache.items[bucket].object)); } else { if(JK_EXPECT_F(setBucket == 0UL) && JK_EXPECT_F(parseState->cache.age[bucket] == 0U)) { setBucket = 1UL; useableBucket = bucket; } if(JK_EXPECT_F(setBucket == 0UL)) { parseState->cache.age[bucket] >>= 1; jk_cache_age(parseState); jk_cache_age(parseState); } // This is the open addressing function. The values length and type are used as a form of "double hashing" to distribute values with the same effective value hash across different object cache buckets. // The values type is a prime number that is relatively coprime to the other primes in the set of value types and the number of hash table buckets. bucket = (parseState->token.value.hash + (parseState->token.value.ptrRange.length * (x + 1UL)) + (parseState->token.value.type * (x + 1UL)) + (3UL * (x + 1UL))) & (parseState->cache.count - 1UL); } } switch(parseState->token.value.type) { case JKValueTypeString: parsedAtom = (void *)CFStringCreateWithBytes(NULL, parseState->token.value.ptrRange.ptr, parseState->token.value.ptrRange.length, kCFStringEncodingUTF8, 0); break; case JKValueTypeLongLong: parsedAtom = (void *)CFNumberCreate(NULL, kCFNumberLongLongType, &parseState->token.value.number.longLongValue); break; case JKValueTypeUnsignedLongLong: if(parseState->token.value.number.unsignedLongLongValue <= LLONG_MAX) { parsedAtom = (void *)CFNumberCreate(NULL, kCFNumberLongLongType, &parseState->token.value.number.unsignedLongLongValue); } else { parsedAtom = (void *)parseState->objCImpCache.NSNumberInitWithUnsignedLongLong(parseState->objCImpCache.NSNumberAlloc(parseState->objCImpCache.NSNumberClass, @selector(alloc)), @selector(initWithUnsignedLongLong:), parseState->token.value.number.unsignedLongLongValue); } break; case JKValueTypeDouble: parsedAtom = (void *)CFNumberCreate(NULL, kCFNumberDoubleType, &parseState->token.value.number.doubleValue); break; default: jk_error(parseState, @"Internal error: Unknown token value type. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); break; } if(JK_EXPECT_T(setBucket) && (JK_EXPECT_T(parsedAtom != NULL))) { bucket = useableBucket; if(JK_EXPECT_T((parseState->cache.items[bucket].object != NULL))) { CFRelease(parseState->cache.items[bucket].object); parseState->cache.items[bucket].object = NULL; } if(JK_EXPECT_T((parseState->cache.items[bucket].bytes = (unsigned char *)reallocf(parseState->cache.items[bucket].bytes, parseState->token.value.ptrRange.length)) != NULL)) { memcpy(parseState->cache.items[bucket].bytes, parseState->token.value.ptrRange.ptr, parseState->token.value.ptrRange.length); parseState->cache.items[bucket].object = (void *)CFRetain(parsedAtom); parseState->cache.items[bucket].hash = parseState->token.value.hash; parseState->cache.items[bucket].cfHash = 0UL; parseState->cache.items[bucket].size = parseState->token.value.ptrRange.length; parseState->cache.items[bucket].type = parseState->token.value.type; parseState->token.value.cacheItem = &parseState->cache.items[bucket]; parseState->cache.age[bucket] = JK_INIT_CACHE_AGE; } else { // The realloc failed, so clear the appropriate fields. parseState->cache.items[bucket].hash = 0UL; parseState->cache.items[bucket].cfHash = 0UL; parseState->cache.items[bucket].size = 0UL; parseState->cache.items[bucket].type = 0UL; } } return(parsedAtom); } static void *jk_object_for_token(JKParseState *parseState) { void *parsedAtom = NULL; parseState->token.value.cacheItem = NULL; switch(parseState->token.type) { case JKTokenTypeString: parsedAtom = jk_cachedObjects(parseState); break; case JKTokenTypeNumber: parsedAtom = jk_cachedObjects(parseState); break; case JKTokenTypeObjectBegin: parsedAtom = jk_parse_dictionary(parseState); break; case JKTokenTypeArrayBegin: parsedAtom = jk_parse_array(parseState); break; case JKTokenTypeTrue: parsedAtom = (void *)kCFBooleanTrue; break; case JKTokenTypeFalse: parsedAtom = (void *)kCFBooleanFalse; break; case JKTokenTypeNull: parsedAtom = (void *)kCFNull; break; default: jk_error(parseState, @"Internal error: Unknown token type. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); break; } return(parsedAtom); } #pragma mark - @implementation JSONDecoder static Class _jk_NSNumberClass; static NSNumberAllocImp _jk_NSNumberAllocImp; static NSNumberInitWithUnsignedLongLongImp _jk_NSNumberInitWithUnsignedLongLongImp; + (void)load { NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; // Though technically not required, the run time enviornment at +load time may be less than ideal. _jk_NSNumberClass = [NSNumber class]; _jk_NSNumberAllocImp = (NSNumberAllocImp)[NSNumber methodForSelector:@selector(alloc)]; // Hacktacular. Need to do it this way due to the nature of class clusters. id temp_NSNumber = [NSNumber alloc]; _jk_NSNumberInitWithUnsignedLongLongImp = (NSNumberInitWithUnsignedLongLongImp)[temp_NSNumber methodForSelector:@selector(initWithUnsignedLongLong:)]; [[temp_NSNumber init] release]; temp_NSNumber = NULL; [pool release]; pool = NULL; } + (id)decoder { return([self decoderWithParseOptions:JKParseOptionStrict]); } + (id)decoderWithParseOptions:(JKParseOptionFlags)parseOptionFlags { return([[[self alloc] initWithParseOptions:parseOptionFlags] autorelease]); } - (id)init { return([self initWithParseOptions:JKParseOptionStrict]); } - (id)initWithParseOptions:(JKParseOptionFlags)parseOptionFlags { if((self = [super init]) == NULL) { return(NULL); } if(parseOptionFlags & ~JKParseOptionValidFlags) { [self autorelease]; [NSException raise:NSInvalidArgumentException format:@"Invalid parse options."]; } if((parseState = (JKParseState *)calloc(1UL, sizeof(JKParseState))) == NULL) { goto errorExit; } parseState->parseOptionFlags = parseOptionFlags; parseState->token.tokenBuffer.roundSizeUpToMultipleOf = 4096UL; parseState->objectStack.roundSizeUpToMultipleOf = 2048UL; parseState->objCImpCache.NSNumberClass = _jk_NSNumberClass; parseState->objCImpCache.NSNumberAlloc = _jk_NSNumberAllocImp; parseState->objCImpCache.NSNumberInitWithUnsignedLongLong = _jk_NSNumberInitWithUnsignedLongLongImp; parseState->cache.prng_lfsr = 1U; parseState->cache.count = JK_CACHE_SLOTS; if((parseState->cache.items = (JKTokenCacheItem *)calloc(1UL, sizeof(JKTokenCacheItem) * parseState->cache.count)) == NULL) { goto errorExit; } return(self); errorExit: if(self) { [self autorelease]; self = NULL; } return(NULL); } // This is here primarily to support the NSString and NSData convienence functions so the autoreleased JSONDecoder can release most of its resources before the pool pops. static void _JSONDecoderCleanup(JSONDecoder *decoder) { if((decoder != NULL) && (decoder->parseState != NULL)) { jk_managedBuffer_release(&decoder->parseState->token.tokenBuffer); jk_objectStack_release(&decoder->parseState->objectStack); [decoder clearCache]; if(decoder->parseState->cache.items != NULL) { free(decoder->parseState->cache.items); decoder->parseState->cache.items = NULL; } free(decoder->parseState); decoder->parseState = NULL; } } - (void)dealloc { _JSONDecoderCleanup(self); [super dealloc]; } - (void)clearCache { if(JK_EXPECT_T(parseState != NULL) && JK_EXPECT_T(parseState->cache.items != NULL)) { size_t idx = 0UL; for(idx = 0UL; idx < parseState->cache.count; idx++) { if(JK_EXPECT_T(parseState->cache.items[idx].object != NULL)) { CFRelease(parseState->cache.items[idx].object); parseState->cache.items[idx].object = NULL; } if(JK_EXPECT_T(parseState->cache.items[idx].bytes != NULL)) { free(parseState->cache.items[idx].bytes); parseState->cache.items[idx].bytes = NULL; } #ifndef NS_BLOCK_ASSERTIONS memset(&parseState->cache.items[idx], 0, sizeof(JKTokenCacheItem)); #endif } } } - (id)parseUTF8String:(const unsigned char *)string length:(size_t)length { return([self parseUTF8String:string length:length error:NULL]); } // This needs to be completely rewritten. - (id)parseUTF8String:(const unsigned char *)string length:(size_t)length error:(NSError **)error { if(parseState == NULL) { [NSException raise:NSInternalInconsistencyException format:@"parseState is NULL."]; } if(string == NULL) { [NSException raise:NSInvalidArgumentException format:@"The string argument is NULL."]; } if((error != NULL) && (*error != NULL)) { *error = NULL; } parseState->stringBuffer.bytes.ptr = string; parseState->stringBuffer.bytes.length = length; parseState->atIndex = 0UL; parseState->lineNumber = 1UL; parseState->lineStartIndex = 0UL; parseState->prev_atIndex = 0UL; parseState->prev_lineNumber = 1UL; parseState->prev_lineStartIndex = 0UL; parseState->error = NULL; parseState->errorIsPrev = 0; unsigned char stackTokenBuffer[JK_TOKENBUFFER_SIZE] JK_ALIGNED(64); jk_managedBuffer_setToStackBuffer(&parseState->token.tokenBuffer, stackTokenBuffer, sizeof(stackTokenBuffer)); void *stackObjects [JK_STACK_OBJS] JK_ALIGNED(64); void *stackKeys [JK_STACK_OBJS] JK_ALIGNED(64); CFHashCode stackCFHashes[JK_STACK_OBJS] JK_ALIGNED(64); jk_objectStack_setToStackBuffer(&parseState->objectStack, stackObjects, stackKeys, stackCFHashes, JK_STACK_OBJS); id parsedJSON = json_parse_it(parseState); if((error != NULL) && (parseState->error != NULL)) { *error = parseState->error; } jk_managedBuffer_release(&parseState->token.tokenBuffer); jk_objectStack_release(&parseState->objectStack); parseState->stringBuffer.bytes.ptr = NULL; parseState->stringBuffer.bytes.length = 0UL; parseState->atIndex = 0UL; parseState->lineNumber = 1UL; parseState->lineStartIndex = 0UL; parseState->prev_atIndex = 0UL; parseState->prev_lineNumber = 1UL; parseState->prev_lineStartIndex = 0UL; parseState->error = NULL; parseState->errorIsPrev = 0; return(parsedJSON); } - (id)parseJSONData:(NSData *)jsonData { return([self parseJSONData:jsonData error:NULL]); } - (id)parseJSONData:(NSData *)jsonData error:(NSError **)error { if(jsonData == NULL) { [NSException raise:NSInvalidArgumentException format:@"The jsonData argument is NULL."]; } return([self parseUTF8String:(const unsigned char *)[jsonData bytes] length:[jsonData length] error:error]); } @end /* The NSString and NSData convience methods need a little bit of explination. Prior to JSONKit v1.4, the NSString -objectFromJSONStringWithParseOptions:error: method looked like const unsigned char *utf8String = (const unsigned char *)[self UTF8String]; if(utf8String == NULL) { return(NULL); } size_t utf8Length = strlen((const char *)utf8String); return([[JSONDecoder decoderWithParseOptions:parseOptionFlags] parseUTF8String:utf8String length:utf8Length error:error]); This changed with v1.4 to a more complicated method. The reason for this is to keep the amount of memory that is allocated, but not yet freed because it is dependent on the autorelease pool to pop before it can be reclaimed. In the simpler v1.3 code, this included all the bytes used to store the -UTF8String along with the JSONDecoder and all its overhead. Now we use an autoreleased CFMutableData that is sized to the UTF8 length of the NSString in question and is used to hold the UTF8 conversion of said string. Once parsed, the CFMutableData has its length set to 0. This should, hopefully, allow the CFMutableData to realloc and/or free the buffer. Another change made was a slight modification to JSONDecoder so that most of the cleanup work that was done in -dealloc was moved to a private, internal function. These convience routines keep the pointer to the autoreleased JSONDecoder and calls _JSONDecoderCleanup() to early release the decoders resources since we already know that particular decoder is not going to be used again. If everything goes smoothly, this will most likely result in perhaps a few hundred bytes that are allocated but waiting for the autorelease pool to pop. This is compared to the thousands and easily hundreds of thousands of bytes that would have been in autorelease limbo. It's more complicated for us, but a win for the user. Autorelease objects are used in case things don't go smoothly. By having them autoreleased, we effectively guarentee that our requirement to -release the object is always met, not matter what goes wrong. The downside is having a an object or two in autorelease limbo, but we've done our best to minimize that impact, so it all balances out. */ @implementation NSString (JSONKit) - (id)objectFromJSONString { return([self objectFromJSONStringWithParseOptions:JKParseOptionStrict error:NULL]); } - (id)objectFromJSONStringWithParseOptions:(JKParseOptionFlags)parseOptionFlags { return([self objectFromJSONStringWithParseOptions:parseOptionFlags error:NULL]); } - (id)objectFromJSONStringWithParseOptions:(JKParseOptionFlags)parseOptionFlags error:(NSError **)error { id returnObject = NULL; CFMutableDataRef mutableData = NULL; JSONDecoder *decoder = NULL; CFIndex stringLength = CFStringGetLength((CFStringRef)self); NSUInteger stringUTF8Length = [self lengthOfBytesUsingEncoding:NSUTF8StringEncoding]; if((mutableData = (CFMutableDataRef)[(id)CFDataCreateMutable(NULL, (NSUInteger)stringUTF8Length) autorelease]) != NULL) { UInt8 *utf8String = CFDataGetMutableBytePtr(mutableData); CFIndex usedBytes = 0L, convertedCount = 0L; convertedCount = CFStringGetBytes((CFStringRef)self, CFRangeMake(0L, stringLength), kCFStringEncodingUTF8, '?', NO, utf8String, (NSUInteger)stringUTF8Length, &usedBytes); if(JK_EXPECT_F(convertedCount != stringLength) || JK_EXPECT_F(usedBytes < 0L)) { if(error != NULL) { *error = [NSError errorWithDomain:@"JKErrorDomain" code:-1L userInfo:[NSDictionary dictionaryWithObject:@"An error occured converting the contents of a NSString to UTF8." forKey:NSLocalizedDescriptionKey]]; } goto exitNow; } returnObject = [(decoder = [JSONDecoder decoderWithParseOptions:parseOptionFlags]) parseUTF8String:(const unsigned char *)utf8String length:(size_t)usedBytes error:error]; } exitNow: if(mutableData != NULL) { CFDataSetLength(mutableData, 0L); } if(decoder != NULL) { _JSONDecoderCleanup(decoder); } return(returnObject); } @end @implementation NSData (JSONKit) - (id)objectFromJSONData { return([self objectFromJSONDataWithParseOptions:JKParseOptionStrict error:NULL]); } - (id)objectFromJSONDataWithParseOptions:(JKParseOptionFlags)parseOptionFlags { return([self objectFromJSONDataWithParseOptions:parseOptionFlags error:NULL]); } - (id)objectFromJSONDataWithParseOptions:(JKParseOptionFlags)parseOptionFlags error:(NSError **)error { JSONDecoder *decoder = NULL; id returnObject = [(decoder = [JSONDecoder decoderWithParseOptions:parseOptionFlags]) parseJSONData:self error:error]; if(decoder != NULL) { _JSONDecoderCleanup(decoder); } return(returnObject); } @end //////////// #pragma mark - #pragma mark Encoding / deserializing functions static void jk_encode_error(JKEncodeState *encodeState, NSString *format, ...) { NSCParameterAssert((encodeState != NULL) && (format != NULL)); va_list varArgsList; va_start(varArgsList, format); NSString *formatString = [[[NSString alloc] initWithFormat:format arguments:varArgsList] autorelease]; va_end(varArgsList); if(encodeState->error == NULL) { encodeState->error = [NSError errorWithDomain:@"JKErrorDomain" code:-1L userInfo: [NSDictionary dictionaryWithObjectsAndKeys: formatString, NSLocalizedDescriptionKey, NULL]]; } } static int jk_encode_printf(JKEncodeState *encodeState, const char *format, ...) { va_list varArgsList; va_start(varArgsList, format); va_end(varArgsList); if( encodeState->stringBuffer.bytes.length < encodeState->atIndex) { jk_encode_error(encodeState, @"Internal inconsistency error: atIndex > buffer length. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); return(1); } if((encodeState->stringBuffer.bytes.length - encodeState->atIndex) < 1024L) { if(jk_managedBuffer_resize(&encodeState->stringBuffer, encodeState->atIndex + 4096UL) == NULL) { jk_encode_error(encodeState, @"Unable to resize temporary buffer."); return(1); } } char *atPtr = (char *)encodeState->stringBuffer.bytes.ptr + encodeState->atIndex; ssize_t remaining = encodeState->stringBuffer.bytes.length - encodeState->atIndex; int printfAdded = vsnprintf(atPtr, remaining, format, varArgsList); while(printfAdded > remaining) { if(jk_managedBuffer_resize(&encodeState->stringBuffer, encodeState->stringBuffer.bytes.length + (printfAdded * 2UL) + 1024UL) == NULL) { jk_encode_error(encodeState, @"Unable to resize temporary buffer."); return(1); } remaining = encodeState->stringBuffer.bytes.length - encodeState->atIndex; printfAdded = vsnprintf(atPtr, remaining, format, varArgsList); } encodeState->atIndex += printfAdded; return(0); } static int jk_encode_write(JKEncodeState *encodeState, const char *format) { #ifndef NS_BLOCK_ASSERTIONS if(JK_EXPECT_F(encodeState->stringBuffer.bytes.length < encodeState->atIndex)) { jk_encode_error(encodeState, @"Internal inconsistency error: atIndex > buffer length. %@ line #%ld", [NSString stringWithUTF8String:__FILE__], (long)__LINE__); return(1); } #endif if(JK_EXPECT_F((encodeState->stringBuffer.bytes.length - encodeState->atIndex) < 1024L)) { if(jk_managedBuffer_resize(&encodeState->stringBuffer, encodeState->atIndex + 4096UL) == NULL) { jk_encode_error(encodeState, @"Unable to resize temporary buffer."); return(1); } } char *atPtr = (char *)encodeState->stringBuffer.bytes.ptr + encodeState->atIndex; ssize_t remaining = encodeState->stringBuffer.bytes.length - encodeState->atIndex; ssize_t idx = 0L, added = 0L; for(added = 0L, idx = 0L; format[added] != 0; added++) { if(JK_EXPECT_T(idx < remaining)) { atPtr[idx++] = format[added]; } } if(JK_EXPECT_F(added > remaining)) { if(JK_EXPECT_F(jk_managedBuffer_resize(&encodeState->stringBuffer, encodeState->atIndex + added + 1024UL) == NULL)) { jk_encode_error(encodeState, @"Unable to resize temporary buffer."); return(1); } for(added = 0L, idx = 0L; format[added] != 0; added++) { if(JK_EXPECT_T(idx < remaining)) { atPtr[idx++] = format[added]; } } } atPtr[idx] = 0; encodeState->atIndex += added; return(0); } static int jk_encode_write1(JKEncodeState *encodeState, const char *format) { if((encodeState->atIndex + 4UL) < encodeState->stringBuffer.bytes.length) { encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = format[0]; } else { if(JK_EXPECT_F(jk_encode_write(encodeState, format))) { return(1); } } return(0); } static int jk_encode_writen(JKEncodeState *encodeState, const char *format, size_t length) { if(JK_EXPECT_F((encodeState->stringBuffer.bytes.length - encodeState->atIndex) < (length + 4UL))) { if(jk_managedBuffer_resize(&encodeState->stringBuffer, encodeState->atIndex + 4096UL + length) == NULL) { jk_encode_error(encodeState, @"Unable to resize temporary buffer."); return(1); } } memcpy(encodeState->stringBuffer.bytes.ptr + encodeState->atIndex, format, length); encodeState->atIndex += length; return(0); } static int jk_encode_add_atom_to_buffer(JKEncodeState *encodeState, void *objectPtr) { NSCParameterAssert((encodeState != NULL) && (objectPtr != NULL)); NSCParameterAssert(encodeState->atIndex < encodeState->stringBuffer.bytes.length); id object = (id)objectPtr; int isClass = JKClassUnknown; if(object->isa == encodeState->fastClassLookup.stringClass) { isClass = JKClassString; } else if(object->isa == encodeState->fastClassLookup.numberClass) { isClass = JKClassNumber; } else if(object->isa == encodeState->fastClassLookup.dictionaryClass) { isClass = JKClassDictionary; } else if(object->isa == encodeState->fastClassLookup.arrayClass) { isClass = JKClassArray; } else if(object->isa == encodeState->fastClassLookup.nullClass) { isClass = JKClassNull; } else { if([object isKindOfClass:[NSString class]]) { encodeState->fastClassLookup.stringClass = object->isa; isClass = JKClassString; } else if([object isKindOfClass:[NSNumber class]]) { encodeState->fastClassLookup.numberClass = object->isa; isClass = JKClassNumber; } else if([object isKindOfClass:[NSDictionary class]]) { encodeState->fastClassLookup.dictionaryClass = object->isa; isClass = JKClassDictionary; } else if([object isKindOfClass:[NSArray class]]) { encodeState->fastClassLookup.arrayClass = object->isa; isClass = JKClassArray; } else if([object isKindOfClass:[NSNull class]]) { encodeState->fastClassLookup.nullClass = object->isa; isClass = JKClassNull; } else { jk_encode_error(encodeState, @"Unable to serialize object class %@.", NSStringFromClass([object class])); return(1); } } switch(isClass) { case JKClassString: { { const unsigned char *cStringPtr = (const unsigned char *)CFStringGetCStringPtr((CFStringRef)object, kCFStringEncodingMacRoman); if(cStringPtr != NULL) { size_t utf8Idx = 0UL, startingAtIndex = encodeState->atIndex; const unsigned char *utf8String = cStringPtr; CFIndex stringLength = CFStringGetLength((CFStringRef)object); if(JK_EXPECT_F(((encodeState->atIndex + (stringLength * 2UL) + 256UL) > encodeState->stringBuffer.bytes.length)) && JK_EXPECT_F((jk_managedBuffer_resize(&encodeState->stringBuffer, encodeState->atIndex + (stringLength * 2UL) + 1024UL) == NULL))) { jk_encode_error(encodeState, @"Unable to resize temporary buffer."); return(1); } encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\"'; for(utf8Idx = 0UL; utf8String[utf8Idx] != 0U; utf8Idx++) { NSCParameterAssert(((&encodeState->stringBuffer.bytes.ptr[encodeState->atIndex]) - encodeState->stringBuffer.bytes.ptr) < (ssize_t)encodeState->stringBuffer.bytes.length); NSCParameterAssert(encodeState->atIndex < encodeState->stringBuffer.bytes.length); if(JK_EXPECT_F(utf8String[utf8Idx] >= 0x80U)) { encodeState->atIndex = startingAtIndex; goto slowUTF8Path; } if(JK_EXPECT_F(utf8String[utf8Idx] < 0x20U)) { switch(utf8String[utf8Idx]) { case '\b': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 'b'; break; case '\f': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 'f'; break; case '\n': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 'n'; break; case '\r': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 'r'; break; case '\t': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 't'; break; default: if(jk_encode_printf(encodeState, "\\u%4.4x", utf8String[utf8Idx])) { return(1); } break; } } else { if(JK_EXPECT_F(utf8String[utf8Idx] == '\"') || JK_EXPECT_F(utf8String[utf8Idx] == '\\')) { encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; } encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = utf8String[utf8Idx]; } } NSCParameterAssert((encodeState->atIndex + 1UL) < encodeState->stringBuffer.bytes.length); encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\"'; return(0); } } slowUTF8Path: { CFIndex stringLength = CFStringGetLength((CFStringRef)object); CFIndex maxStringUTF8Length = CFStringGetMaximumSizeForEncoding(stringLength, kCFStringEncodingUTF8) + 32L; if(((size_t)maxStringUTF8Length > encodeState->utf8ConversionBuffer.bytes.length) && (jk_managedBuffer_resize(&encodeState->utf8ConversionBuffer, maxStringUTF8Length + 1024UL) == NULL)) { jk_encode_error(encodeState, @"Unable to resize temporary buffer."); return(1); } CFIndex usedBytes = 0L, convertedCount = 0L; convertedCount = CFStringGetBytes((CFStringRef)object, CFRangeMake(0L, stringLength), kCFStringEncodingUTF8, '?', NO, encodeState->utf8ConversionBuffer.bytes.ptr, encodeState->utf8ConversionBuffer.bytes.length - 16L, &usedBytes); if(JK_EXPECT_F(convertedCount != stringLength) || JK_EXPECT_F(usedBytes < 0L)) { jk_encode_error(encodeState, @"An error occured converting the contents of a NSString to UTF8."); return(1); } if(JK_EXPECT_F((encodeState->atIndex + (maxStringUTF8Length * 2UL) + 256UL) > encodeState->stringBuffer.bytes.length) && JK_EXPECT_F(jk_managedBuffer_resize(&encodeState->stringBuffer, encodeState->atIndex + (maxStringUTF8Length * 2UL) + 1024UL) == NULL)) { jk_encode_error(encodeState, @"Unable to resize temporary buffer."); return(1); } const unsigned char *utf8String = encodeState->utf8ConversionBuffer.bytes.ptr; size_t utf8Idx = 0UL; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\"'; for(utf8Idx = 0UL; utf8Idx < (size_t)usedBytes; utf8Idx++) { NSCParameterAssert(((&encodeState->stringBuffer.bytes.ptr[encodeState->atIndex]) - encodeState->stringBuffer.bytes.ptr) < (ssize_t)encodeState->stringBuffer.bytes.length); NSCParameterAssert(encodeState->atIndex < encodeState->stringBuffer.bytes.length); NSCParameterAssert((CFIndex)utf8Idx < usedBytes); if(JK_EXPECT_F(utf8String[utf8Idx] < 0x20U)) { switch(utf8String[utf8Idx]) { case '\b': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 'b'; break; case '\f': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 'f'; break; case '\n': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 'n'; break; case '\r': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 'r'; break; case '\t': encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = 't'; break; default: if(jk_encode_printf(encodeState, "\\u%4.4x", utf8String[utf8Idx])) { return(1); } break; } } else { if(JK_EXPECT_F(utf8String[utf8Idx] >= 0x80U) && (encodeState->serializeOptionFlags & JKSerializeOptionEscapeUnicode)) { const unsigned char *nextValidCharacter = NULL; UTF32 u32ch = 0U; ConversionResult result; if(JK_EXPECT_F((result = ConvertSingleCodePointInUTF8(&utf8String[utf8Idx], &utf8String[usedBytes], (UTF8 const **)&nextValidCharacter, &u32ch)) != conversionOK)) { jk_encode_error(encodeState, @"Error converting UTF8."); return(1); } else { utf8Idx = (nextValidCharacter - utf8String) - 1UL; if(u32ch <= 0xffffU) { if(jk_encode_printf(encodeState, "\\u%4.4x", u32ch)) { return(1); } } else { if(jk_encode_printf(encodeState, "\\u%4.4x\\u%4.4x", (0xd7c0U + (u32ch >> 10)), (0xdc00U + (u32ch & 0x3ffU)))) { return(1); } } } } else { if((utf8String[utf8Idx] == '\"') || (utf8String[utf8Idx] == '\\')) { encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\\'; } encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = utf8String[utf8Idx]; } } } NSCParameterAssert((encodeState->atIndex + 1UL) < encodeState->stringBuffer.bytes.length); encodeState->stringBuffer.bytes.ptr[encodeState->atIndex++] = '\"'; return(0); } } break; case JKClassNumber: { if(object == (id)kCFBooleanTrue) { return(jk_encode_writen(encodeState, "true", 4UL)); break; } else if(object == (id)kCFBooleanFalse) { return(jk_encode_writen(encodeState, "false", 5UL)); break; } const char *objCType = [object objCType]; char anum[256], *aptr = &anum[255]; int isNegative = 0; unsigned long long ullv; long long llv; switch(objCType[0]) { case 'c': case 'i': case 's': case 'l': case 'q': if(JK_EXPECT_T(CFNumberGetValue((CFNumberRef)object, kCFNumberLongLongType, &llv))) { if(llv < 0LL) { llv = -llv; isNegative = 1; } if(JK_EXPECT_F(llv < 10LL)) { *--aptr = llv + '0'; } else { while(JK_EXPECT_T(llv > 0LL)) { *--aptr = (llv % 10LL) + '0'; llv /= 10LL; NSCParameterAssert(aptr > anum); } } if(isNegative) { *--aptr = '-'; } NSCParameterAssert(aptr > anum); return(jk_encode_writen(encodeState, aptr, &anum[255] - aptr)); } else { jk_encode_error(encodeState, @"Unable to get scalar value from number object."); return(1); } case 'C': case 'I': case 'S': case 'L': case 'Q': case 'B': if(JK_EXPECT_T(CFNumberGetValue((CFNumberRef)object, kCFNumberLongLongType, &ullv))) { if(JK_EXPECT_F(ullv < 10ULL)) { *--aptr = ullv + '0'; } else { while(JK_EXPECT_T(ullv > 0ULL)) { *--aptr = (ullv % 10ULL) + '0'; ullv /= 10ULL; NSCParameterAssert(aptr > anum); } } NSCParameterAssert(aptr > anum); return(jk_encode_writen(encodeState, aptr, &anum[255] - aptr)); } else { jk_encode_error(encodeState, @"Unable to get scalar value from number object."); return(1); } case 'f': case 'd': { double dv; if(JK_EXPECT_T(CFNumberGetValue((CFNumberRef)object, kCFNumberDoubleType, &dv))) { if(JK_EXPECT_F(!isfinite(dv))) { jk_encode_error(encodeState, @"Floating point values must be finite. JSON does not support NaN or Infinity."); return(1); } return(jk_encode_printf(encodeState, "%.17g", dv)); } } break; default: jk_encode_error(encodeState, @"NSNumber conversion error, unknown type. Type: '%c' / 0x%2.2x", objCType[0], objCType[0]); return(1); break; } } break; case JKClassArray: { int printComma = 0; CFIndex arrayCount = CFArrayGetCount((CFArrayRef)object), idx = 0L; if(jk_encode_write1(encodeState, "[")) { return(1); } if(arrayCount > 1020L) { for(id arrayObject in object) { if(printComma) { if(jk_encode_write1(encodeState, ",")) { return(1); } } printComma = 1; if(jk_encode_add_atom_to_buffer(encodeState, arrayObject)) { return(1); } } } else { void *objects[1024]; CFArrayGetValues((CFArrayRef)object, CFRangeMake(0L, arrayCount), (const void **)objects); for(idx = 0L; idx < arrayCount; idx++) { if(printComma) { if(jk_encode_write1(encodeState, ",")) { return(1); } } printComma = 1; if(jk_encode_add_atom_to_buffer(encodeState, objects[idx])) { return(1); } } } if(jk_encode_write1(encodeState, "]")) { return(1); } } break; case JKClassDictionary: { int printComma = 0; CFIndex dictionaryCount = CFDictionaryGetCount((CFDictionaryRef)object), idx = 0L; if(JK_EXPECT_F(jk_encode_write1(encodeState, "{"))) { return(1); } if(JK_EXPECT_F(dictionaryCount > 1020L)) { for(id keyObject in object) { if(printComma) { if(JK_EXPECT_F(jk_encode_write1(encodeState, ","))) { return(1); } } printComma = 1; if(JK_EXPECT_F((keyObject->isa != encodeState->fastClassLookup.stringClass)) && JK_EXPECT_F(([keyObject isKindOfClass:[NSString class]] == NO))) { jk_encode_error(encodeState, @"Key must be a string object."); return(1); } if(JK_EXPECT_F(jk_encode_add_atom_to_buffer(encodeState, keyObject))) { return(1); } if(JK_EXPECT_F(jk_encode_write1(encodeState, ":"))) { return(1); } if(JK_EXPECT_F(jk_encode_add_atom_to_buffer(encodeState, (void *)CFDictionaryGetValue((CFDictionaryRef)object, keyObject)))) { return(1); } } } else { void *keys[1024], *objects[1024]; CFDictionaryGetKeysAndValues((CFDictionaryRef)object, (const void **)keys, (const void **)objects); for(idx = 0L; idx < dictionaryCount; idx++) { if(JK_EXPECT_F(printComma)) { if(JK_EXPECT_F(jk_encode_write1(encodeState, ","))) { return(1); } } printComma = 1; if(JK_EXPECT_F((((id)keys[idx])->isa != encodeState->fastClassLookup.stringClass)) && JK_EXPECT_F(([(id)keys[idx] isKindOfClass:[NSString class]] == NO))) { jk_encode_error(encodeState, @"Key must be a string object."); return(1); } if(JK_EXPECT_F(jk_encode_add_atom_to_buffer(encodeState, keys[idx]))) { return(1); } if(JK_EXPECT_F(jk_encode_write1(encodeState, ":"))) { return(1); } if(JK_EXPECT_F(jk_encode_add_atom_to_buffer(encodeState, objects[idx]))) { return(1); } } } if(JK_EXPECT_F(jk_encode_write1(encodeState, "}"))) { return(1); } } break; case JKClassNull: if(JK_EXPECT_F(jk_encode_writen(encodeState, "null", 4UL))) { return(1); } break; default: jk_encode_error(encodeState, @"Unable to serialize object class %@.", NSStringFromClass([object class])); return(1); break; } return(0); } static id jk_encode(void *object, JKSerializeOptionFlags optionFlags, JKEncodeAsType encodeAs, NSError **error) { NSCParameterAssert(object != NULL); id returnObject = NULL; if((error != NULL) && (*error != NULL)) { *error = NULL; } JKEncodeState encodeState; memset(&encodeState, 0, sizeof(JKEncodeState)); encodeState.serializeOptionFlags = optionFlags; encodeState.stringBuffer.roundSizeUpToMultipleOf = (1024UL * 32UL); encodeState.utf8ConversionBuffer.roundSizeUpToMultipleOf = 4096UL; unsigned char stackJSONBuffer[JK_JSONBUFFER_SIZE] JK_ALIGNED(64); jk_managedBuffer_setToStackBuffer(&encodeState.stringBuffer, stackJSONBuffer, sizeof(stackJSONBuffer)); unsigned char stackUTF8Buffer[JK_UTF8BUFFER_SIZE] JK_ALIGNED(64); jk_managedBuffer_setToStackBuffer(&encodeState.utf8ConversionBuffer, stackUTF8Buffer, sizeof(stackUTF8Buffer)); if(jk_encode_add_atom_to_buffer(&encodeState, object) == 0) { switch(encodeAs) { case JKEncodeAsData: { NSData *jsonData = NULL; if((encodeState.stringBuffer.flags & JKManagedBufferMustFree) == 0) { if((jsonData = [(NSData *)CFDataCreate(NULL, encodeState.stringBuffer.bytes.ptr, encodeState.atIndex) autorelease]) == NULL) { jk_encode_error(&encodeState, @"Unable to create NSData object"); } } else { if((encodeState.stringBuffer.bytes.ptr = (unsigned char *)reallocf(encodeState.stringBuffer.bytes.ptr, encodeState.atIndex)) != NULL) { if((jsonData = [(NSData *)CFDataCreateWithBytesNoCopy(NULL, encodeState.stringBuffer.bytes.ptr, encodeState.atIndex, NULL) autorelease]) == NULL) { jk_encode_error(&encodeState, @"Unable to create NSData object"); } } if((jsonData != NULL) || (encodeState.stringBuffer.bytes.ptr == NULL)) { encodeState.stringBuffer.flags &= ~JKManagedBufferMustFree; encodeState.stringBuffer.bytes.ptr = NULL; encodeState.stringBuffer.bytes.length = 0UL; } } returnObject = jsonData; } break; case JKEncodeAsString: { NSString *jsonString = NULL; if((encodeState.stringBuffer.flags & JKManagedBufferMustFree) == 0) { if((jsonString = [(NSString *)CFStringCreateWithBytes(NULL, (const UInt8 *)encodeState.stringBuffer.bytes.ptr, (CFIndex)encodeState.atIndex, kCFStringEncodingUTF8, NO) autorelease]) == NULL) { jk_encode_error(&encodeState, @"Unable to create NSString object"); } } else { if((encodeState.stringBuffer.bytes.ptr = (unsigned char *)reallocf(encodeState.stringBuffer.bytes.ptr, encodeState.atIndex + 256UL)) != NULL) { encodeState.stringBuffer.bytes.ptr[encodeState.atIndex + 1UL] = 0; if((jsonString = [(NSString *)CFStringCreateWithBytesNoCopy(NULL, (const UInt8 *)encodeState.stringBuffer.bytes.ptr, (CFIndex)encodeState.atIndex, kCFStringEncodingUTF8, NO, NULL) autorelease]) == NULL) { jk_encode_error(&encodeState, @"Unable to create NSString object"); } } if((jsonString != NULL) || (encodeState.stringBuffer.bytes.ptr == NULL)) { encodeState.stringBuffer.flags &= ~JKManagedBufferMustFree; encodeState.stringBuffer.bytes.ptr = NULL; encodeState.stringBuffer.bytes.length = 0UL; } } returnObject = jsonString; } break; default: jk_encode_error(&encodeState, @"Unknown encode as type."); break; } } if((error != NULL) && (encodeState.error != NULL)) { *error = encodeState.error; } jk_managedBuffer_release(&encodeState.stringBuffer); jk_managedBuffer_release(&encodeState.utf8ConversionBuffer); return(returnObject); } @implementation NSArray (JSONKit) - (NSData *)JSONData { return([self JSONDataWithOptions:JKSerializeOptionNone error:NULL]); } - (NSData *)JSONDataWithOptions:(JKSerializeOptionFlags)serializeOptions error:(NSError **)error { return(jk_encode(self, serializeOptions, JKEncodeAsData, error)); } - (NSString *)JSONString { return([self JSONStringWithOptions:JKSerializeOptionNone error:NULL]); } - (NSString *)JSONStringWithOptions:(JKSerializeOptionFlags)serializeOptions error:(NSError **)error { return(jk_encode(self, serializeOptions, JKEncodeAsString, error)); } @end @implementation NSDictionary (JSONKit) - (NSData *)JSONData { return([self JSONDataWithOptions:JKSerializeOptionNone error:NULL]); } - (NSData *)JSONDataWithOptions:(JKSerializeOptionFlags)serializeOptions error:(NSError **)error { return(jk_encode(self, serializeOptions, JKEncodeAsData, error)); } - (NSString *)JSONString { return([self JSONStringWithOptions:JKSerializeOptionNone error:NULL]); } - (NSString *)JSONStringWithOptions:(JKSerializeOptionFlags)serializeOptions error:(NSError **)error { return(jk_encode(self, serializeOptions, JKEncodeAsString, error)); } @end