json 1.2.4 → 1.4.0

data/ext/json/ext/generator.h

@@ -0,0 +1,170 @@
+#ifndef _GENERATOR_H_
+#define _GENERATOR_H_
+
+#include <string.h>
+#include <assert.h>
+#include <math.h>
+
+#include "ruby.h"
+
+#if HAVE_RUBY_RE_H
+#include "ruby/re.h"
+#endif
+
+#if HAVE_RE_H
+#include "re.h"
+#endif
+
+#ifdef HAVE_RUBY_ENCODING_H
+#include "ruby/encoding.h"
+#define FORCE_UTF8(obj) rb_enc_associate((obj), rb_utf8_encoding())
+#else
+#define FORCE_UTF8(obj)
+#endif
+
+#define option_given_p(opts, key) RTEST(rb_funcall(opts, i_key_p, 1, key))
+
+#ifndef RHASH_SIZE
+#define RHASH_SIZE(hsh) (RHASH(hsh)->tbl->num_entries)
+#endif
+
+#ifndef RFLOAT_VALUE
+#define RFLOAT_VALUE(val) (RFLOAT(val)->value)
+#endif
+
+#ifndef RARRAY_PTR
+#define RARRAY_PTR(ARRAY) RARRAY(ARRAY)->ptr
+#endif
+#ifndef RARRAY_LEN
+#define RARRAY_LEN(ARRAY) RARRAY(ARRAY)->len
+#endif
+#ifndef RSTRING_PTR
+#define RSTRING_PTR(string) RSTRING(string)->ptr
+#endif
+#ifndef RSTRING_LEN
+#define RSTRING_LEN(string) RSTRING(string)->len
+#endif
+
+#define RSTRING_PAIR(string) RSTRING_PTR(string), RSTRING_LEN(string)
+
+/* fbuffer implementation */
+
+typedef struct FBufferStruct {
+    unsigned int initial_length;
+    char *ptr;
+    unsigned int len;
+    unsigned int capa;
+} FBuffer;
+
+#define FBUFFER_INITIAL_LENGTH 4096
+
+#define FBUFFER_PTR(fb) (fb->ptr)
+#define FBUFFER_LEN(fb) (fb->len)
+#define FBUFFER_CAPA(fb) (fb->capa)
+#define FBUFFER_PAIR(fb) FBUFFER_PTR(fb), FBUFFER_LEN(fb)
+
+static char *fstrndup(const char *ptr, int len);
+static FBuffer *fbuffer_alloc();
+static FBuffer *fbuffer_alloc_with_length(unsigned initial_length);
+static void fbuffer_free(FBuffer *fb);
+static void fbuffer_free_only_buffer(FBuffer *fb);
+static void fbuffer_clear(FBuffer *fb);
+static void fbuffer_append(FBuffer *fb, const char *newstr, unsigned int len);
+static void fbuffer_append_integer(FBuffer *fb, int number);
+static void fbuffer_append_char(FBuffer *fb, char newchr);
+static FBuffer *fbuffer_dup(FBuffer *fb);
+
+/* unicode defintions */
+
+#define UNI_STRICT_CONVERSION 1
+
+typedef unsigned long	UTF32;	/* at least 32 bits */
+typedef unsigned short	UTF16;	/* at least 16 bits */
+typedef unsigned char	UTF8;	/* typically 8 bits */
+
+#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
+
+static const int halfShift  = 10; /* used for shifting by 10 bits */
+
+static const UTF32 halfBase = 0x0010000UL;
+static const UTF32 halfMask = 0x3FFUL;
+
+static unsigned char isLegalUTF8(const UTF8 *source, int length);
+static void unicode_escape(char *buf, UTF16 character);
+static void unicode_escape_to_buffer(FBuffer *buffer, char buf[6], UTF16 character);
+static void convert_UTF8_to_JSON_ASCII(FBuffer *buffer, VALUE string);
+static void convert_UTF8_to_JSON(FBuffer *buffer, VALUE string);
+
+/* ruby api and some helpers */
+
+typedef struct JSON_Generator_StateStruct {
+    char *indent;
+    long indent_len;
+    char *space;
+    long space_len;
+    char *space_before;
+    long space_before_len;
+    char *object_nl;
+    long object_nl_len;
+    char *array_nl;
+    long array_nl_len;
+    FBuffer *array_delim;
+    FBuffer *object_delim;
+    FBuffer *object_delim2;
+    long max_nesting;
+    char allow_nan;
+    char ascii_only;
+} JSON_Generator_State;
+
+#define GET_STATE(self)                       \
+    JSON_Generator_State *state;              \
+    Data_Get_Struct(self, JSON_Generator_State, state)
+
+static VALUE mHash_to_json(int argc, VALUE *argv, VALUE self);
+static VALUE mArray_to_json(int argc, VALUE *argv, VALUE self);
+static VALUE mInteger_to_json(int argc, VALUE *argv, VALUE self);
+static VALUE mFloat_to_json(int argc, VALUE *argv, VALUE self);
+static VALUE mString_included_s(VALUE self, VALUE modul);
+static VALUE mString_to_json(int argc, VALUE *argv, VALUE self);
+static VALUE mString_to_json_raw_object(VALUE self);
+static VALUE mString_to_json_raw(int argc, VALUE *argv, VALUE self);
+static VALUE mString_Extend_json_create(VALUE self, VALUE o);
+static VALUE mTrueClass_to_json(int argc, VALUE *argv, VALUE self);
+static VALUE mFalseClass_to_json(int argc, VALUE *argv, VALUE self);
+static VALUE mNilClass_to_json(int argc, VALUE *argv, VALUE self);
+static VALUE mObject_to_json(int argc, VALUE *argv, VALUE self);
+static void State_free(JSON_Generator_State *state);
+static JSON_Generator_State *State_allocate();
+static VALUE cState_s_allocate(VALUE klass);
+static VALUE cState_configure(VALUE self, VALUE opts);
+static VALUE cState_to_h(VALUE self);
+static void generate_json(FBuffer *buffer, VALUE Vstate, JSON_Generator_State *state, VALUE obj, long depth);
+static VALUE cState_partial_generate(VALUE self, VALUE obj, VALUE depth);
+static VALUE cState_generate(VALUE self, VALUE obj);
+static VALUE cState_initialize(int argc, VALUE *argv, VALUE self);
+static VALUE cState_from_state_s(VALUE self, VALUE opts);
+static VALUE cState_indent(VALUE self);
+static VALUE cState_indent_set(VALUE self, VALUE indent);
+static VALUE cState_space(VALUE self);
+static VALUE cState_space_set(VALUE self, VALUE space);
+static VALUE cState_space_before(VALUE self);
+static VALUE cState_space_before_set(VALUE self, VALUE space_before);
+static VALUE cState_object_nl(VALUE self);
+static VALUE cState_object_nl_set(VALUE self, VALUE object_nl);
+static VALUE cState_array_nl(VALUE self);
+static VALUE cState_array_nl_set(VALUE self, VALUE array_nl);
+static VALUE cState_max_nesting(VALUE self);
+static VALUE cState_max_nesting_set(VALUE self, VALUE depth);
+static VALUE cState_allow_nan_p(VALUE self);
+static VALUE cState_ascii_only_p(VALUE self);
+
+#endif

data/ext/json/ext/{parser/parser.c → parser.c}

@@ -1,23 +1,76 @@
 
 #line 1 "parser.rl"
-#include "ruby.h"
-#include "unicode.h"
-#if HAVE_RE_H
-#include "re.h"
-#endif
-
-#define EVIL 0x666
+#include "parser.h"
+
+/* unicode */
+
+static const char digit_values[256] = { 
+    -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, -1, -1, -1, -1, -1, -1,
+    -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, -1,
+    -1, -1, -1, -1, -1, -1, 10, 11, 12, 13, 14, 15, -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,
+    10, 11, 12, 13, 14, 15, -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,
+    -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, -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, -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, -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
+};
+
+static UTF32 unescape_unicode(const unsigned char *p)
+{
+    char b;
+    UTF32 result = 0;
+    b = digit_values[p[0]];
+    if (b < 0) return UNI_REPLACEMENT_CHAR;
+    result = (result << 4) | b;
+    b = digit_values[p[1]];
+    result = (result << 4) | b;
+    if (b < 0) return UNI_REPLACEMENT_CHAR;
+    b = digit_values[p[2]];
+    result = (result << 4) | b;
+    if (b < 0) return UNI_REPLACEMENT_CHAR;
+    b = digit_values[p[3]];
+    result = (result << 4) | b;
+    if (b < 0) return UNI_REPLACEMENT_CHAR;
+    return result;
+}
 
-#define option_given_p(opts, key) RTEST(rb_funcall((opts), i_key_p, 1, (key)))
+static int convert_UTF32_to_UTF8(char *buf, UTF32 ch) 
+{
+    int len = 1;
+    if (ch <= 0x7F) {
+        buf[0] = (char) ch;
+    } else if (ch <= 0x07FF) {
+        buf[0] = (char) ((ch >> 6) | 0xC0);
+        buf[1] = (char) ((ch & 0x3F) | 0x80);
+        len++;
+    } else if (ch <= 0xFFFF) {
+        buf[0] = (char) ((ch >> 12) | 0xE0);
+        buf[1] = (char) (((ch >> 6) & 0x3F) | 0x80);
+        buf[2] = (char) ((ch & 0x3F) | 0x80);
+        len += 2;
+    } else if (ch <= 0x1fffff) {
+        buf[0] =(char) ((ch >> 18) | 0xF0);
+        buf[1] =(char) (((ch >> 12) & 0x3F) | 0x80);
+        buf[2] =(char) (((ch >> 6) & 0x3F) | 0x80);
+        buf[3] =(char) ((ch & 0x3F) | 0x80);
+        len += 3;
+    } else {
+        buf[0] = '?';
+    }
+    return len;
+}
 
 #ifdef HAVE_RUBY_ENCODING_H
-#include "ruby/encoding.h"
-#define FORCE_UTF8(obj) rb_enc_associate((obj), rb_utf8_encoding())
-static VALUE mEncoding_ASCII_8BIT, mEncoding_UTF_8, mEncoding_UTF_16BE,
-    mEncoding_UTF_16LE, mEncoding_UTF_32BE, mEncoding_UTF_32LE;
+static VALUE CEncoding_ASCII_8BIT, CEncoding_UTF_8, CEncoding_UTF_16BE,
+    CEncoding_UTF_16LE, CEncoding_UTF_32BE, CEncoding_UTF_32LE;
 static ID i_encoding, i_encode, i_encode_bang, i_force_encoding;
 #else
-#define FORCE_UTF8(obj)
 static ID i_iconv;
 #endif
 
@@ -28,40 +81,12 @@ static ID i_json_creatable_p, i_json_create, i_create_id, i_create_additions,
           i_chr, i_max_nesting, i_allow_nan, i_symbolize_names, i_object_class,
           i_array_class, i_key_p, i_deep_const_get;
 
-#define MinusInfinity "-Infinity"
-
-typedef struct JSON_ParserStruct {
-    VALUE Vsource;
-    char *source;
-    long len;
-    char *memo;
-    VALUE create_id;
-    int max_nesting;
-    int current_nesting;
-    int allow_nan;
-    int parsing_name;
-    int symbolize_names;
-    VALUE object_class;
-    VALUE array_class;
-} JSON_Parser;
-
-static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result);
-static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result);
-static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result);
-static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result);
-static char *JSON_parse_integer(JSON_Parser *json, char *p, char *pe, VALUE *result);
-static char *JSON_parse_float(JSON_Parser *json, char *p, char *pe, VALUE *result);
-
-#define GET_STRUCT                          \
-    JSON_Parser *json;                      \
-    Data_Get_Struct(self, JSON_Parser, json);
 
+#line 108 "parser.rl"
 
-#line 83 "parser.rl"
 
 
-
-#line 65 "parser.c"
+#line 90 "parser.c"
 static const int JSON_object_start = 1;
 static const int JSON_object_first_final = 27;
 static const int JSON_object_error = 0;
@@ -69,7 +94,7 @@ static const int JSON_object_error = 0;
 static const int JSON_object_en_main = 1;
 
 
-#line 118 "parser.rl"
+#line 143 "parser.rl"
 
 
 static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result)
@@ -85,14 +110,14 @@ static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *resu
     *result = NIL_P(object_class) ? rb_hash_new() : rb_class_new_instance(0, 0, object_class);
 
     
-#line 89 "parser.c"
+#line 114 "parser.c"
 	{
 	cs = JSON_object_start;
 	}
 
-#line 133 "parser.rl"
+#line 158 "parser.rl"
     
-#line 96 "parser.c"
+#line 121 "parser.c"
 	{
 	if ( p == pe )
 		goto _test_eof;
@@ -120,7 +145,7 @@ case 2:
 		goto st2;
 	goto st0;
 tr2:
-#line 102 "parser.rl"
+#line 127 "parser.rl"
 	{
         json->parsing_name = 1;
         char *np = JSON_parse_string(json, p, pe, &last_name);
@@ -132,7 +157,7 @@ st3:
 	if ( ++p == pe )
 		goto _test_eof3;
 case 3:
-#line 136 "parser.c"
+#line 161 "parser.c"
 	switch( (*p) ) {
 		case 13: goto st3;
 		case 32: goto st3;
@@ -199,7 +224,7 @@ case 8:
 		goto st8;
 	goto st0;
 tr11:
-#line 91 "parser.rl"
+#line 116 "parser.rl"
 	{
         VALUE v = Qnil;
         char *np = JSON_parse_value(json, p, pe, &v); 
@@ -215,7 +240,7 @@ st9:
 	if ( ++p == pe )
 		goto _test_eof9;
 case 9:
-#line 219 "parser.c"
+#line 244 "parser.c"
 	switch( (*p) ) {
 		case 13: goto st9;
 		case 32: goto st9;
@@ -304,14 +329,14 @@ case 18:
 		goto st9;
 	goto st18;
 tr4:
-#line 109 "parser.rl"
+#line 134 "parser.rl"
 	{ p--; {p++; cs = 27; goto _out;} }
 	goto st27;
 st27:
 	if ( ++p == pe )
 		goto _test_eof27;
 case 27:
-#line 315 "parser.c"
+#line 340 "parser.c"
 	goto st0;
 st19:
 	if ( ++p == pe )
@@ -409,7 +434,7 @@ case 26:
 	_out: {}
 	}
 
-#line 134 "parser.rl"
+#line 159 "parser.rl"
 
     if (cs >= JSON_object_first_final) {
         if (RTEST(json->create_id)) {
@@ -428,7 +453,7 @@ case 26:
 }
 
 
-#line 432 "parser.c"
+#line 457 "parser.c"
 static const int JSON_value_start = 1;
 static const int JSON_value_first_final = 21;
 static const int JSON_value_error = 0;
@@ -436,7 +461,7 @@ static const int JSON_value_error = 0;
 static const int JSON_value_en_main = 1;
 
 
-#line 232 "parser.rl"
+#line 257 "parser.rl"
 
 
 static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result)
@@ -444,14 +469,14 @@ static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *resul
     int cs = EVIL;
 
     
-#line 448 "parser.c"
+#line 473 "parser.c"
 	{
 	cs = JSON_value_start;
 	}
 
-#line 239 "parser.rl"
+#line 264 "parser.rl"
     
-#line 455 "parser.c"
+#line 480 "parser.c"
 	{
 	if ( p == pe )
 		goto _test_eof;
@@ -476,14 +501,14 @@ st0:
 cs = 0;
 	goto _out;
 tr0:
-#line 180 "parser.rl"
+#line 205 "parser.rl"
 	{
         char *np = JSON_parse_string(json, p, pe, result);
         if (np == NULL) { p--; {p++; cs = 21; goto _out;} } else {p = (( np))-1;}
     }
 	goto st21;
 tr2:
-#line 185 "parser.rl"
+#line 210 "parser.rl"
 	{
         char *np;
         if(pe > p + 9 && !strncmp(MinusInfinity, p, 9)) {
@@ -503,7 +528,7 @@ tr2:
     }
 	goto st21;
 tr5:
-#line 203 "parser.rl"
+#line 228 "parser.rl"
 	{ 
         char *np;
         json->current_nesting++;
@@ -513,7 +538,7 @@ tr5:
     }
 	goto st21;
 tr9:
-#line 211 "parser.rl"
+#line 236 "parser.rl"
 	{ 
         char *np;
         json->current_nesting++;
@@ -523,7 +548,7 @@ tr9:
     }
 	goto st21;
 tr16:
-#line 173 "parser.rl"
+#line 198 "parser.rl"
 	{
         if (json->allow_nan) {
             *result = CInfinity;
@@ -533,7 +558,7 @@ tr16:
     }
 	goto st21;
 tr18:
-#line 166 "parser.rl"
+#line 191 "parser.rl"
 	{
         if (json->allow_nan) {
             *result = CNaN;
@@ -543,19 +568,19 @@ tr18:
     }
 	goto st21;
 tr22:
-#line 160 "parser.rl"
+#line 185 "parser.rl"
 	{
         *result = Qfalse;
     }
 	goto st21;
 tr25:
-#line 157 "parser.rl"
+#line 182 "parser.rl"
 	{
         *result = Qnil;
     }
 	goto st21;
 tr28:
-#line 163 "parser.rl"
+#line 188 "parser.rl"
 	{
         *result = Qtrue;
     }
@@ -564,9 +589,9 @@ st21:
 	if ( ++p == pe )
 		goto _test_eof21;
 case 21:
-#line 219 "parser.rl"
+#line 244 "parser.rl"
 	{ p--; {p++; cs = 21; goto _out;} }
-#line 570 "parser.c"
+#line 595 "parser.c"
 	goto st0;
 st2:
 	if ( ++p == pe )
@@ -727,7 +752,7 @@ case 20:
 	_out: {}
 	}
 
-#line 240 "parser.rl"
+#line 265 "parser.rl"
 
     if (cs >= JSON_value_first_final) {
         return p;
@@ -737,7 +762,7 @@ case 20:
 }
 
 
-#line 741 "parser.c"
+#line 766 "parser.c"
 static const int JSON_integer_start = 1;
 static const int JSON_integer_first_final = 5;
 static const int JSON_integer_error = 0;
@@ -745,7 +770,7 @@ static const int JSON_integer_error = 0;
 static const int JSON_integer_en_main = 1;
 
 
-#line 256 "parser.rl"
+#line 281 "parser.rl"
 
 
 static char *JSON_parse_integer(JSON_Parser *json, char *p, char *pe, VALUE *result)
@@ -753,15 +778,15 @@ static char *JSON_parse_integer(JSON_Parser *json, char *p, char *pe, VALUE *res
     int cs = EVIL;
 
     
-#line 757 "parser.c"
+#line 782 "parser.c"
 	{
 	cs = JSON_integer_start;
 	}
 
-#line 263 "parser.rl"
+#line 288 "parser.rl"
     json->memo = p;
     
-#line 765 "parser.c"
+#line 790 "parser.c"
 	{
 	if ( p == pe )
 		goto _test_eof;
@@ -795,14 +820,14 @@ case 3:
 		goto st0;
 	goto tr4;
 tr4:
-#line 253 "parser.rl"
+#line 278 "parser.rl"
 	{ p--; {p++; cs = 5; goto _out;} }
 	goto st5;
 st5:
 	if ( ++p == pe )
 		goto _test_eof5;
 case 5:
-#line 806 "parser.c"
+#line 831 "parser.c"
 	goto st0;
 st4:
 	if ( ++p == pe )
@@ -821,7 +846,7 @@ case 4:
 	_out: {}
 	}
 
-#line 265 "parser.rl"
+#line 290 "parser.rl"
 
     if (cs >= JSON_integer_first_final) {
         long len = p - json->memo;
@@ -833,7 +858,7 @@ case 4:
 }
 
 
-#line 837 "parser.c"
+#line 862 "parser.c"
 static const int JSON_float_start = 1;
 static const int JSON_float_first_final = 10;
 static const int JSON_float_error = 0;
@@ -841,7 +866,7 @@ static const int JSON_float_error = 0;
 static const int JSON_float_en_main = 1;
 
 
-#line 287 "parser.rl"
+#line 312 "parser.rl"
 
 
 static char *JSON_parse_float(JSON_Parser *json, char *p, char *pe, VALUE *result)
@@ -849,15 +874,15 @@ static char *JSON_parse_float(JSON_Parser *json, char *p, char *pe, VALUE *resul
     int cs = EVIL;
 
     
-#line 853 "parser.c"
+#line 878 "parser.c"
 	{
 	cs = JSON_float_start;
 	}
 
-#line 294 "parser.rl"
+#line 319 "parser.rl"
     json->memo = p;
     
-#line 861 "parser.c"
+#line 886 "parser.c"
 	{
 	if ( p == pe )
 		goto _test_eof;
@@ -915,14 +940,14 @@ case 5:
 		goto st0;
 	goto tr7;
 tr7:
-#line 281 "parser.rl"
+#line 306 "parser.rl"
 	{ p--; {p++; cs = 10; goto _out;} }
 	goto st10;
 st10:
 	if ( ++p == pe )
 		goto _test_eof10;
 case 10:
-#line 926 "parser.c"
+#line 951 "parser.c"
 	goto st0;
 st6:
 	if ( ++p == pe )
@@ -983,7 +1008,7 @@ case 9:
 	_out: {}
 	}
 
-#line 296 "parser.rl"
+#line 321 "parser.rl"
 
     if (cs >= JSON_float_first_final) {
         long len = p - json->memo;
@@ -996,7 +1021,7 @@ case 9:
 
 
 
-#line 1000 "parser.c"
+#line 1025 "parser.c"
 static const int JSON_array_start = 1;
 static const int JSON_array_first_final = 17;
 static const int JSON_array_error = 0;
@@ -1004,7 +1029,7 @@ static const int JSON_array_error = 0;
 static const int JSON_array_en_main = 1;
 
 
-#line 332 "parser.rl"
+#line 357 "parser.rl"
 
 
 static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result)
@@ -1018,14 +1043,14 @@ static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *resul
     *result = NIL_P(array_class) ? rb_ary_new() : rb_class_new_instance(0, 0, array_class);
 
     
-#line 1022 "parser.c"
+#line 1047 "parser.c"
 	{
 	cs = JSON_array_start;
 	}
 
-#line 345 "parser.rl"
+#line 370 "parser.rl"
     
-#line 1029 "parser.c"
+#line 1054 "parser.c"
 	{
 	if ( p == pe )
 		goto _test_eof;
@@ -1064,7 +1089,7 @@ case 2:
 		goto st2;
 	goto st0;
 tr2:
-#line 313 "parser.rl"
+#line 338 "parser.rl"
 	{
         VALUE v = Qnil;
         char *np = JSON_parse_value(json, p, pe, &v); 
@@ -1080,7 +1105,7 @@ st3:
 	if ( ++p == pe )
 		goto _test_eof3;
 case 3:
-#line 1084 "parser.c"
+#line 1109 "parser.c"
 	switch( (*p) ) {
 		case 13: goto st3;
 		case 32: goto st3;
@@ -1180,14 +1205,14 @@ case 12:
 		goto st3;
 	goto st12;
 tr4:
-#line 324 "parser.rl"
+#line 349 "parser.rl"
 	{ p--; {p++; cs = 17; goto _out;} }
 	goto st17;
 st17:
 	if ( ++p == pe )
 		goto _test_eof17;
 case 17:
-#line 1191 "parser.c"
+#line 1216 "parser.c"
 	goto st0;
 st13:
 	if ( ++p == pe )
@@ -1243,73 +1268,88 @@ case 16:
 	_out: {}
 	}
 
-#line 346 "parser.rl"
+#line 371 "parser.rl"
 
     if(cs >= JSON_array_first_final) {
         return p + 1;
     } else {
         rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p);
+        return NULL;
     }
 }
 
-static VALUE json_string_unescape(char *p, char *pe)
+static VALUE json_string_unescape(VALUE result, char *string, char *stringEnd)
 {
-    VALUE result = rb_str_buf_new(pe - p + 1);
-
-    while (p < pe) {
-        if (*p == '\\') {
-            p++;
-            if (p >= pe) return Qnil; /* raise an exception later, \ at end */
-            switch (*p) {
+    char *p = string, *pe = string, *unescape;
+    int unescape_len;
+
+    while (pe < stringEnd) {
+        if (*pe == '\\') {
+            unescape = (char *) "?";
+            unescape_len = 1;
+            if (pe > p) rb_str_buf_cat(result, p, pe - p);
+            switch (*++pe) {
+                case 'n':
+                    unescape = (char *) "\n";
+                    break;
+                case 'r':
+                    unescape = (char *) "\r";
+                    break;
+                case 't':
+                    unescape = (char *) "\t";
+                    break;
                 case '"':
+                    unescape = (char *) "\"";
+                    break;
                 case '\\':
-                    rb_str_buf_cat(result, p, 1);
-                    p++;
+                    unescape = (char *) "\\";
                     break;
                 case 'b':
-                    rb_str_buf_cat2(result, "\b");
-                    p++;
+                    unescape = (char *) "\b";
                     break;
                 case 'f':
-                    rb_str_buf_cat2(result, "\f");
-                    p++;
-                    break;
-                case 'n':
-                    rb_str_buf_cat2(result, "\n");
-                    p++;
-                    break;
-                case 'r':
-                    rb_str_buf_cat2(result, "\r");
-                    p++;
-                    break;
-                case 't':
-                    rb_str_buf_cat2(result, "\t");
-                    p++;
+                    unescape = (char *) "\f";
                     break;
                 case 'u':
-                    if (p > pe - 4) { 
+                    if (pe > stringEnd - 4) { 
                         return Qnil;
                     } else {
-                        p = JSON_convert_UTF16_to_UTF8(result, p, pe, strictConversion);
+                        char buf[4];
+                        UTF32 ch = unescape_unicode((unsigned char *) ++pe);
+                        pe += 3;
+                        if (UNI_SUR_HIGH_START == (ch & 0xFC00)) {
+                            pe++;
+                            if (pe > stringEnd - 6) return Qnil;
+                            if (pe[0] == '\\' && pe[1] == 'u') {
+                                UTF32 sur = unescape_unicode((unsigned char *) pe + 2);
+                                ch = (((ch & 0x3F) << 10) | ((((ch >> 6) & 0xF) + 1) << 16)
+                                        | (sur & 0x3FF));
+                                pe += 5;
+                            } else {
+                                unescape = (char *) "?";
+                                break;
+                            }
+                        }
+                        unescape_len = convert_UTF32_to_UTF8(buf, ch);
+                        unescape = buf;
                     }
                     break;
                 default:
-                    rb_str_buf_cat(result, p, 1);
-                    p++;
-                    break;
+                    p = pe;
+                    continue;
             }
+            rb_str_buf_cat(result, unescape, unescape_len);
+            p = ++pe;
         } else {
-            char *q = p;
-            while (*q != '\\' && q < pe) q++;
-            rb_str_buf_cat(result, p, q - p);
-            p = q;
+            pe++;
         }
     }
+    rb_str_buf_cat(result, p, pe - p);
     return result;
 }
 
 
-#line 1313 "parser.c"
+#line 1353 "parser.c"
 static const int JSON_string_start = 1;
 static const int JSON_string_first_final = 8;
 static const int JSON_string_error = 0;
@@ -1317,24 +1357,24 @@ static const int JSON_string_error = 0;
 static const int JSON_string_en_main = 1;
 
 
-#line 430 "parser.rl"
+#line 470 "parser.rl"
 
 
 static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result)
 {
     int cs = EVIL;
 
-    *result = rb_str_new("", 0);
+    *result = rb_str_buf_new(0);
     
-#line 1330 "parser.c"
+#line 1370 "parser.c"
 	{
 	cs = JSON_string_start;
 	}
 
-#line 438 "parser.rl"
+#line 478 "parser.rl"
     json->memo = p;
     
-#line 1338 "parser.c"
+#line 1378 "parser.c"
 	{
 	if ( p == pe )
 		goto _test_eof;
@@ -1359,25 +1399,25 @@ case 2:
 		goto st0;
 	goto st2;
 tr2:
-#line 416 "parser.rl"
+#line 456 "parser.rl"
 	{
-      *result = json_string_unescape(json->memo + 1, p);
-      if (NIL_P(*result)) {
-        p--;
-        {p++; cs = 8; goto _out;}
-      } else {
-        FORCE_UTF8(*result);
-        {p = (( p + 1))-1;}
-      }
-    }
-#line 427 "parser.rl"
+        *result = json_string_unescape(*result, json->memo + 1, p);
+        if (NIL_P(*result)) {
+			p--;
+			{p++; cs = 8; goto _out;}
+		} else {
+			FORCE_UTF8(*result);
+			{p = (( p + 1))-1;}
+		}
+	}
+#line 467 "parser.rl"
 	{ p--; {p++; cs = 8; goto _out;} }
 	goto st8;
 st8:
 	if ( ++p == pe )
 		goto _test_eof8;
 case 8:
-#line 1381 "parser.c"
+#line 1421 "parser.c"
 	goto st0;
 st3:
 	if ( ++p == pe )
@@ -1453,7 +1493,7 @@ case 7:
 	_out: {}
 	}
 
-#line 440 "parser.rl"
+#line 480 "parser.rl"
 
     if (json->symbolize_names && json->parsing_name) {
       *result = rb_str_intern(*result);
@@ -1467,7 +1507,7 @@ case 7:
 
 
 
-#line 1471 "parser.c"
+#line 1511 "parser.c"
 static const int JSON_start = 1;
 static const int JSON_first_final = 10;
 static const int JSON_error = 0;
@@ -1475,7 +1515,7 @@ static const int JSON_error = 0;
 static const int JSON_en_main = 1;
 
 
-#line 477 "parser.rl"
+#line 517 "parser.rl"
 
 
 /* 
@@ -1490,7 +1530,7 @@ static const int JSON_en_main = 1;
  *
  */
 
-inline static VALUE convert_encoding(VALUE source)
+static VALUE convert_encoding(VALUE source)
 {
     char *ptr = RSTRING_PTR(source);
     long len = RSTRING_LEN(source);
@@ -1500,28 +1540,28 @@ inline static VALUE convert_encoding(VALUE source)
 #ifdef HAVE_RUBY_ENCODING_H
     {
         VALUE encoding = rb_funcall(source, i_encoding, 0);
-        if (encoding == mEncoding_ASCII_8BIT) {
+        if (encoding == CEncoding_ASCII_8BIT) {
             if (len >= 4 &&  ptr[0] == 0 && ptr[1] == 0 && ptr[2] == 0) {
                 source = rb_str_dup(source);
-                rb_funcall(source, i_force_encoding, 1, mEncoding_UTF_32BE);
-                source = rb_funcall(source, i_encode_bang, 1, mEncoding_UTF_8);
+                rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32BE);
+                source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8);
             } else if (len >= 4 && ptr[0] == 0 && ptr[2] == 0) {
                 source = rb_str_dup(source);
-                rb_funcall(source, i_force_encoding, 1, mEncoding_UTF_16BE);
-                source = rb_funcall(source, i_encode_bang, 1, mEncoding_UTF_8);
+                rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16BE);
+                source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8);
             } else if (len >= 4 && ptr[1] == 0 && ptr[2] == 0 && ptr[3] == 0) {
                 source = rb_str_dup(source);
-                rb_funcall(source, i_force_encoding, 1, mEncoding_UTF_32LE);
-                source = rb_funcall(source, i_encode_bang, 1, mEncoding_UTF_8);
+                rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32LE);
+                source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8);
             } else if (len >= 4 && ptr[1] == 0 && ptr[3] == 0) {
                 source = rb_str_dup(source);
-                rb_funcall(source, i_force_encoding, 1, mEncoding_UTF_16LE);
-                source = rb_funcall(source, i_encode_bang, 1, mEncoding_UTF_8);
+                rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16LE);
+                source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8);
             } else {
-                source = rb_funcall(source, i_force_encoding, 1, mEncoding_UTF_8);
+                FORCE_UTF8(source);
             }
         } else {
-            source = rb_funcall(source, i_encode, 1, mEncoding_UTF_8);
+            source = rb_funcall(source, i_encode, 1, CEncoding_UTF_8);
         }
     }
 #else
@@ -1569,7 +1609,7 @@ static VALUE cParser_initialize(int argc, VALUE *argv, VALUE self)
     char *ptr;
     long len;
     VALUE source, opts;
-    GET_STRUCT;
+    GET_PARSER;
     rb_scan_args(argc, argv, "11", &source, &opts);
     source = convert_encoding(StringValue(source));
     ptr = RSTRING_PTR(source);
@@ -1654,19 +1694,19 @@ static VALUE cParser_parse(VALUE self)
     char *p, *pe;
     int cs = EVIL;
     VALUE result = Qnil;
-    GET_STRUCT;
+    GET_PARSER;
 
     
-#line 1661 "parser.c"
+#line 1701 "parser.c"
 	{
 	cs = JSON_start;
 	}
 
-#line 658 "parser.rl"
+#line 698 "parser.rl"
     p = json->source;
     pe = p + json->len;
     
-#line 1670 "parser.c"
+#line 1710 "parser.c"
 	{
 	if ( p == pe )
 		goto _test_eof;
@@ -1722,7 +1762,7 @@ case 5:
 		goto st1;
 	goto st5;
 tr3:
-#line 466 "parser.rl"
+#line 506 "parser.rl"
 	{
         char *np;
         json->current_nesting = 1;
@@ -1731,7 +1771,7 @@ tr3:
     }
 	goto st10;
 tr4:
-#line 459 "parser.rl"
+#line 499 "parser.rl"
 	{
         char *np;
         json->current_nesting = 1;
@@ -1743,7 +1783,7 @@ st10:
 	if ( ++p == pe )
 		goto _test_eof10;
 case 10:
-#line 1747 "parser.c"
+#line 1787 "parser.c"
 	switch( (*p) ) {
 		case 13: goto st10;
 		case 32: goto st10;
@@ -1800,16 +1840,17 @@ case 9:
 	_out: {}
 	}
 
-#line 661 "parser.rl"
+#line 701 "parser.rl"
 
     if (cs >= JSON_first_final && p == pe) {
         return result;
     } else {
         rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p);
+        return Qnil;
     }
 }
 
-inline static JSON_Parser *JSON_allocate()
+static JSON_Parser *JSON_allocate()
 {
     JSON_Parser *json = ALLOC(JSON_Parser);
     MEMZERO(json, JSON_Parser, 1);
@@ -1843,7 +1884,7 @@ static VALUE cJSON_parser_s_allocate(VALUE klass)
  */
 static VALUE cParser_source(VALUE self)
 {
-    GET_STRUCT;
+    GET_PARSER;
     return rb_str_dup(json->Vsource);
 }
 
@@ -1877,12 +1918,12 @@ void Init_parser()
     i_key_p = rb_intern("key?");
     i_deep_const_get = rb_intern("deep_const_get");
 #ifdef HAVE_RUBY_ENCODING_H
-    mEncoding_UTF_8 = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-8"));
-    mEncoding_UTF_16BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16be"));
-    mEncoding_UTF_16LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16le"));
-    mEncoding_UTF_32BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32be"));
-    mEncoding_UTF_32LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32le"));
-    mEncoding_ASCII_8BIT = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("ascii-8bit"));
+    CEncoding_UTF_8 = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-8"));
+    CEncoding_UTF_16BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16be"));
+    CEncoding_UTF_16LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16le"));
+    CEncoding_UTF_32BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32be"));
+    CEncoding_UTF_32LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32le"));
+    CEncoding_ASCII_8BIT = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("ascii-8bit"));
     i_encoding = rb_intern("encoding");
     i_encode = rb_intern("encode");
     i_encode_bang = rb_intern("encode!");