ruby/ext/json/parser/parser.c
Jean Boussier a4183781ec [ruby/json] Only use the key cache if the Hash is in an Array
Otherwise the likeliness of seeing that key again is really low, and looking up
the cache is just a waste.

Before:

```
== Parsing small hash (65 bytes)
ruby 3.4.0dev (2024-11-13T12:32:57Z fstr-update-callba.. https://github.com/ruby/json/commit/9b44b455b3) +YJIT +PRISM [arm64-darwin24]
Warming up --------------------------------------
                json   343.049k i/100ms
                  oj   213.943k i/100ms
          Oj::Parser    31.583k i/100ms
           rapidjson   303.433k i/100ms
Calculating -------------------------------------
                json      3.704M (± 1.5%) i/s  (270.01 ns/i) -     18.525M in   5.003078s
                  oj      2.200M (± 1.1%) i/s  (454.46 ns/i) -     11.125M in   5.056526s
          Oj::Parser    285.369k (± 4.8%) i/s    (3.50 μs/i) -      1.453M in   5.103866s
           rapidjson      3.216M (± 1.6%) i/s  (310.95 ns/i) -     16.082M in   5.001973s

Comparison:
                json:  3703517.4 i/s
           rapidjson:  3215983.0 i/s - 1.15x  slower
                  oj:  2200417.1 i/s - 1.68x  slower
          Oj::Parser:   285369.1 i/s - 12.98x  slower

== Parsing test from oj (258 bytes)
ruby 3.4.0dev (2024-11-13T12:32:57Z fstr-update-callba.. https://github.com/ruby/json/commit/9b44b455b3) +YJIT +PRISM [arm64-darwin24]
Warming up --------------------------------------
                json    54.539k i/100ms
                  oj    41.473k i/100ms
          Oj::Parser    24.064k i/100ms
           rapidjson    51.466k i/100ms
Calculating -------------------------------------
                json    549.386k (± 1.6%) i/s    (1.82 μs/i) -      2.781M in   5.064316s
                  oj    417.003k (± 1.3%) i/s    (2.40 μs/i) -      2.115M in   5.073047s
          Oj::Parser    226.500k (± 4.7%) i/s    (4.42 μs/i) -      1.131M in   5.005466s
           rapidjson    526.124k (± 1.0%) i/s    (1.90 μs/i) -      2.676M in   5.087176s

Comparison:
                json:   549385.6 i/s
           rapidjson:   526124.3 i/s - 1.04x  slower
                  oj:   417003.4 i/s - 1.32x  slower
          Oj::Parser:   226500.4 i/s - 2.43x  slower
```

After:

```
== Parsing small hash (65 bytes)
ruby 3.4.0dev (2024-11-13T12:32:57Z fstr-update-callba.. https://github.com/ruby/json/commit/9b44b455b3) +YJIT +PRISM [arm64-darwin24]
Warming up --------------------------------------
                json   361.394k i/100ms
                  oj   217.203k i/100ms
          Oj::Parser    28.855k i/100ms
           rapidjson   303.404k i/100ms
Calculating -------------------------------------
                json      3.859M (± 2.9%) i/s  (259.13 ns/i) -     19.515M in   5.061302s
                  oj      2.191M (± 1.6%) i/s  (456.49 ns/i) -     11.077M in   5.058043s
          Oj::Parser    315.132k (± 7.1%) i/s    (3.17 μs/i) -      1.587M in   5.065707s
           rapidjson      3.156M (± 4.0%) i/s  (316.88 ns/i) -     15.777M in   5.008949s

Comparison:
                json:  3859046.5 i/s
           rapidjson:  3155778.5 i/s - 1.22x  slower
                  oj:  2190616.0 i/s - 1.76x  slower
          Oj::Parser:   315132.4 i/s - 12.25x  slower

== Parsing test from oj (258 bytes)
ruby 3.4.0dev (2024-11-13T12:32:57Z fstr-update-callba.. https://github.com/ruby/json/commit/9b44b455b3) +YJIT +PRISM [arm64-darwin24]
Warming up --------------------------------------
                json    55.682k i/100ms
                  oj    40.343k i/100ms
          Oj::Parser    25.119k i/100ms
           rapidjson    51.500k i/100ms
Calculating -------------------------------------
                json    555.808k (± 1.4%) i/s    (1.80 μs/i) -      2.784M in   5.010092s
                  oj    412.283k (± 1.7%) i/s    (2.43 μs/i) -      2.098M in   5.089900s
          Oj::Parser    279.306k (±13.3%) i/s    (3.58 μs/i) -      1.356M in   5.022079s
           rapidjson    517.177k (± 2.7%) i/s    (1.93 μs/i) -      2.626M in   5.082352s

Comparison:
                json:   555808.3 i/s
           rapidjson:   517177.1 i/s - 1.07x  slower
                  oj:   412283.2 i/s - 1.35x  slower
          Oj::Parser:   279306.5 i/s - 1.99x  slower
```

https://github.com/ruby/json/commit/00c45ddc9f
2024-11-14 11:21:39 +09:00

3239 lines
73 KiB
C

/* This file is automatically generated from parser.rl by using ragel */
#line 1 "parser.rl"
#include "ruby.h"
#include "../fbuffer/fbuffer.h"
static VALUE mJSON, mExt, cParser, eNestingError, Encoding_UTF_8;
static VALUE CNaN, CInfinity, CMinusInfinity;
static ID i_json_creatable_p, i_json_create, i_create_id,
i_chr, i_deep_const_get, i_match, i_aset, i_aref,
i_leftshift, i_new, i_try_convert, i_uminus, i_encode;
static VALUE sym_max_nesting, sym_allow_nan, sym_allow_trailing_comma, sym_symbolize_names, sym_freeze,
sym_create_additions, sym_create_id, sym_object_class, sym_array_class,
sym_decimal_class, sym_match_string;
static int binary_encindex;
static int utf8_encindex;
#ifdef HAVE_RB_CATEGORY_WARN
# define json_deprecated(message) rb_category_warn(RB_WARN_CATEGORY_DEPRECATED, message)
#else
# define json_deprecated(message) rb_warn(message)
#endif
static const char deprecated_create_additions_warning[] =
"JSON.load implicit support for `create_additions: true` is deprecated "
"and will be removed in 3.0, use JSON.unsafe_load or explicitly "
"pass `create_additions: true`";
#ifndef HAVE_RB_GC_MARK_LOCATIONS
// For TruffleRuby
void rb_gc_mark_locations(const VALUE *start, const VALUE *end)
{
VALUE *value = start;
while (value < end) {
rb_gc_mark(*value);
value++;
}
}
#endif
#ifndef HAVE_RB_HASH_BULK_INSERT
// For TruffleRuby
void rb_hash_bulk_insert(long count, const VALUE *pairs, VALUE hash)
{
long index = 0;
while (index < count) {
VALUE name = pairs[index++];
VALUE value = pairs[index++];
rb_hash_aset(hash, name, value);
}
RB_GC_GUARD(hash);
}
#endif
/* name cache */
#include <string.h>
#include <ctype.h>
// Object names are likely to be repeated, and are frozen.
// As such we can re-use them if we keep a cache of the ones we've seen so far,
// and save much more expensive lookups into the global fstring table.
// This cache implementation is deliberately simple, as we're optimizing for compactness,
// to be able to fit safely on the stack.
// As such, binary search into a sorted array gives a good tradeoff between compactness and
// performance.
#define JSON_RVALUE_CACHE_CAPA 63
typedef struct rvalue_cache_struct {
int length;
VALUE entries[JSON_RVALUE_CACHE_CAPA];
} rvalue_cache;
static rb_encoding *enc_utf8;
#define JSON_RVALUE_CACHE_MAX_ENTRY_LENGTH 55
static inline VALUE build_interned_string(const char *str, const long length)
{
# ifdef HAVE_RB_ENC_INTERNED_STR
return rb_enc_interned_str(str, length, enc_utf8);
# else
VALUE rstring = rb_utf8_str_new(str, length);
return rb_funcall(rb_str_freeze(rstring), i_uminus, 0);
# endif
}
static inline VALUE build_symbol(const char *str, const long length)
{
return rb_str_intern(build_interned_string(str, length));
}
static void rvalue_cache_insert_at(rvalue_cache *cache, int index, VALUE rstring)
{
MEMMOVE(&cache->entries[index + 1], &cache->entries[index], VALUE, cache->length - index);
cache->length++;
cache->entries[index] = rstring;
}
static inline int rstring_cache_cmp(const char *str, const long length, VALUE rstring)
{
long rstring_length = RSTRING_LEN(rstring);
if (length == rstring_length) {
return memcmp(str, RSTRING_PTR(rstring), length);
} else {
return (int)(length - rstring_length);
}
}
static VALUE rstring_cache_fetch(rvalue_cache *cache, const char *str, const long length)
{
if (RB_UNLIKELY(length > JSON_RVALUE_CACHE_MAX_ENTRY_LENGTH)) {
// Common names aren't likely to be very long. So we just don't
// cache names above an arbitrary threshold.
return Qfalse;
}
if (RB_UNLIKELY(!isalpha(str[0]))) {
// Simple heuristic, if the first character isn't a letter,
// we're much less likely to see this string again.
// We mostly want to cache strings that are likely to be repeated.
return Qfalse;
}
int low = 0;
int high = cache->length - 1;
int mid = 0;
int last_cmp = 0;
while (low <= high) {
mid = (high + low) >> 1;
VALUE entry = cache->entries[mid];
last_cmp = rstring_cache_cmp(str, length, entry);
if (last_cmp == 0) {
return entry;
} else if (last_cmp > 0) {
low = mid + 1;
} else {
high = mid - 1;
}
}
if (RB_UNLIKELY(memchr(str, '\\', length))) {
// We assume the overwhelming majority of names don't need to be escaped.
// But if they do, we have to fallback to the slow path.
return Qfalse;
}
VALUE rstring = build_interned_string(str, length);
if (cache->length < JSON_RVALUE_CACHE_CAPA) {
if (last_cmp > 0) {
mid += 1;
}
rvalue_cache_insert_at(cache, mid, rstring);
}
return rstring;
}
static VALUE rsymbol_cache_fetch(rvalue_cache *cache, const char *str, const long length)
{
if (RB_UNLIKELY(length > JSON_RVALUE_CACHE_MAX_ENTRY_LENGTH)) {
// Common names aren't likely to be very long. So we just don't
// cache names above an arbitrary threshold.
return Qfalse;
}
if (RB_UNLIKELY(!isalpha(str[0]))) {
// Simple heuristic, if the first character isn't a letter,
// we're much less likely to see this string again.
// We mostly want to cache strings that are likely to be repeated.
return Qfalse;
}
int low = 0;
int high = cache->length - 1;
int mid = 0;
int last_cmp = 0;
while (low <= high) {
mid = (high + low) >> 1;
VALUE entry = cache->entries[mid];
last_cmp = rstring_cache_cmp(str, length, rb_sym2str(entry));
if (last_cmp == 0) {
return entry;
} else if (last_cmp > 0) {
low = mid + 1;
} else {
high = mid - 1;
}
}
if (RB_UNLIKELY(memchr(str, '\\', length))) {
// We assume the overwhelming majority of names don't need to be escaped.
// But if they do, we have to fallback to the slow path.
return Qfalse;
}
VALUE rsymbol = build_symbol(str, length);
if (cache->length < JSON_RVALUE_CACHE_CAPA) {
if (last_cmp > 0) {
mid += 1;
}
rvalue_cache_insert_at(cache, mid, rsymbol);
}
return rsymbol;
}
/* rvalue stack */
#define RVALUE_STACK_INITIAL_CAPA 128
enum rvalue_stack_type {
RVALUE_STACK_HEAP_ALLOCATED = 0,
RVALUE_STACK_STACK_ALLOCATED = 1,
};
typedef struct rvalue_stack_struct {
enum rvalue_stack_type type;
long capa;
long head;
VALUE *ptr;
} rvalue_stack;
static rvalue_stack *rvalue_stack_spill(rvalue_stack *old_stack, VALUE *handle, rvalue_stack **stack_ref);
static rvalue_stack *rvalue_stack_grow(rvalue_stack *stack, VALUE *handle, rvalue_stack **stack_ref)
{
long required = stack->capa * 2;
if (stack->type == RVALUE_STACK_STACK_ALLOCATED) {
stack = rvalue_stack_spill(stack, handle, stack_ref);
} else {
REALLOC_N(stack->ptr, VALUE, required);
stack->capa = required;
}
return stack;
}
static void rvalue_stack_push(rvalue_stack *stack, VALUE value, VALUE *handle, rvalue_stack **stack_ref)
{
if (RB_UNLIKELY(stack->head >= stack->capa)) {
stack = rvalue_stack_grow(stack, handle, stack_ref);
}
stack->ptr[stack->head] = value;
stack->head++;
}
static inline VALUE *rvalue_stack_peek(rvalue_stack *stack, long count)
{
return stack->ptr + (stack->head - count);
}
static inline void rvalue_stack_pop(rvalue_stack *stack, long count)
{
stack->head -= count;
}
static void rvalue_stack_mark(void *ptr)
{
rvalue_stack *stack = (rvalue_stack *)ptr;
rb_gc_mark_locations(stack->ptr, stack->ptr + stack->head);
}
static void rvalue_stack_free(void *ptr)
{
rvalue_stack *stack = (rvalue_stack *)ptr;
if (stack) {
ruby_xfree(stack->ptr);
ruby_xfree(stack);
}
}
static size_t rvalue_stack_memsize(const void *ptr)
{
const rvalue_stack *stack = (const rvalue_stack *)ptr;
return sizeof(rvalue_stack) + sizeof(VALUE) * stack->capa;
}
static const rb_data_type_t JSON_Parser_rvalue_stack_type = {
"JSON::Ext::Parser/rvalue_stack",
{
.dmark = rvalue_stack_mark,
.dfree = rvalue_stack_free,
.dsize = rvalue_stack_memsize,
},
0, 0,
RUBY_TYPED_FREE_IMMEDIATELY,
};
static rvalue_stack *rvalue_stack_spill(rvalue_stack *old_stack, VALUE *handle, rvalue_stack **stack_ref)
{
rvalue_stack *stack;
*handle = TypedData_Make_Struct(0, rvalue_stack, &JSON_Parser_rvalue_stack_type, stack);
*stack_ref = stack;
MEMCPY(stack, old_stack, rvalue_stack, 1);
stack->capa = old_stack->capa << 1;
stack->ptr = ALLOC_N(VALUE, stack->capa);
stack->type = RVALUE_STACK_HEAP_ALLOCATED;
MEMCPY(stack->ptr, old_stack->ptr, VALUE, old_stack->head);
return stack;
}
static void rvalue_stack_eagerly_release(VALUE handle)
{
rvalue_stack *stack;
TypedData_Get_Struct(handle, rvalue_stack, &JSON_Parser_rvalue_stack_type, stack);
RTYPEDDATA_DATA(handle) = NULL;
rvalue_stack_free(stack);
}
/* unicode */
static const signed 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 uint32_t unescape_unicode(const unsigned char *p)
{
const uint32_t replacement_char = 0xFFFD;
signed char b;
uint32_t result = 0;
b = digit_values[p[0]];
if (b < 0) return replacement_char;
result = (result << 4) | (unsigned char)b;
b = digit_values[p[1]];
if (b < 0) return replacement_char;
result = (result << 4) | (unsigned char)b;
b = digit_values[p[2]];
if (b < 0) return replacement_char;
result = (result << 4) | (unsigned char)b;
b = digit_values[p[3]];
if (b < 0) return replacement_char;
result = (result << 4) | (unsigned char)b;
return result;
}
static int convert_UTF32_to_UTF8(char *buf, uint32_t 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;
}
typedef struct JSON_ParserStruct {
VALUE Vsource;
char *source;
long len;
char *memo;
VALUE create_id;
VALUE object_class;
VALUE array_class;
VALUE decimal_class;
VALUE match_string;
FBuffer fbuffer;
int in_array;
int max_nesting;
bool allow_nan;
bool allow_trailing_comma;
bool parsing_name;
bool symbolize_names;
bool freeze;
bool create_additions;
bool deprecated_create_additions;
rvalue_cache name_cache;
rvalue_stack *stack;
VALUE stack_handle;
} JSON_Parser;
#define GET_PARSER \
GET_PARSER_INIT; \
if (!json->Vsource) rb_raise(rb_eTypeError, "uninitialized instance")
#define GET_PARSER_INIT \
JSON_Parser *json; \
TypedData_Get_Struct(self, JSON_Parser, &JSON_Parser_type, json)
#define MinusInfinity "-Infinity"
#define EVIL 0x666
static const rb_data_type_t JSON_Parser_type;
static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result);
static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result, int current_nesting);
static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result, int current_nesting);
static char *JSON_parse_number(JSON_Parser *json, char *p, char *pe, VALUE *result);
static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result, int current_nesting);
#define PARSE_ERROR_FRAGMENT_LEN 32
#ifdef RBIMPL_ATTR_NORETURN
RBIMPL_ATTR_NORETURN()
#endif
static void raise_parse_error(const char *format, const char *start)
{
char buffer[PARSE_ERROR_FRAGMENT_LEN + 1];
size_t len = strnlen(start, PARSE_ERROR_FRAGMENT_LEN);
const char *ptr = start;
if (len == PARSE_ERROR_FRAGMENT_LEN) {
MEMCPY(buffer, start, char, PARSE_ERROR_FRAGMENT_LEN);
buffer[PARSE_ERROR_FRAGMENT_LEN] = '\0';
ptr = buffer;
}
rb_enc_raise(enc_utf8, rb_path2class("JSON::ParserError"), format, ptr);
}
#line 475 "parser.rl"
#line 457 "parser.c"
enum {JSON_object_start = 1};
enum {JSON_object_first_final = 32};
enum {JSON_object_error = 0};
enum {JSON_object_en_main = 1};
#line 515 "parser.rl"
#define PUSH(result) rvalue_stack_push(json->stack, result, &json->stack_handle, &json->stack)
static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result, int current_nesting)
{
int cs = EVIL;
if (json->max_nesting && current_nesting > json->max_nesting) {
rb_raise(eNestingError, "nesting of %d is too deep", current_nesting);
}
long stack_head = json->stack->head;
#line 481 "parser.c"
{
cs = JSON_object_start;
}
#line 530 "parser.rl"
#line 488 "parser.c"
{
short _widec;
if ( p == pe )
goto _test_eof;
switch ( cs )
{
case 1:
if ( (*p) == 123 )
goto st2;
goto st0;
st0:
cs = 0;
goto _out;
st2:
if ( ++p == pe )
goto _test_eof2;
case 2:
switch( (*p) ) {
case 13: goto st2;
case 32: goto st2;
case 34: goto tr2;
case 47: goto st28;
case 125: goto tr4;
}
if ( 9 <= (*p) && (*p) <= 10 )
goto st2;
goto st0;
tr2:
#line 494 "parser.rl"
{
char *np;
json->parsing_name = true;
np = JSON_parse_string(json, p, pe, result);
json->parsing_name = false;
if (np == NULL) { p--; {p++; cs = 3; goto _out;} } else {
PUSH(*result);
{p = (( np))-1;}
}
}
goto st3;
st3:
if ( ++p == pe )
goto _test_eof3;
case 3:
#line 533 "parser.c"
switch( (*p) ) {
case 13: goto st3;
case 32: goto st3;
case 47: goto st4;
case 58: goto st8;
}
if ( 9 <= (*p) && (*p) <= 10 )
goto st3;
goto st0;
st4:
if ( ++p == pe )
goto _test_eof4;
case 4:
switch( (*p) ) {
case 42: goto st5;
case 47: goto st7;
}
goto st0;
st5:
if ( ++p == pe )
goto _test_eof5;
case 5:
if ( (*p) == 42 )
goto st6;
goto st5;
st6:
if ( ++p == pe )
goto _test_eof6;
case 6:
switch( (*p) ) {
case 42: goto st6;
case 47: goto st3;
}
goto st5;
st7:
if ( ++p == pe )
goto _test_eof7;
case 7:
if ( (*p) == 10 )
goto st3;
goto st7;
st8:
if ( ++p == pe )
goto _test_eof8;
case 8:
switch( (*p) ) {
case 13: goto st8;
case 32: goto st8;
case 34: goto tr11;
case 45: goto tr11;
case 47: goto st24;
case 73: goto tr11;
case 78: goto tr11;
case 91: goto tr11;
case 102: goto tr11;
case 110: goto tr11;
case 116: goto tr11;
case 123: goto tr11;
}
if ( (*p) > 10 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto tr11;
} else if ( (*p) >= 9 )
goto st8;
goto st0;
tr11:
#line 483 "parser.rl"
{
char *np = JSON_parse_value(json, p, pe, result, current_nesting);
if (np == NULL) {
p--; {p++; cs = 9; goto _out;}
} else {
{p = (( np))-1;}
}
}
goto st9;
st9:
if ( ++p == pe )
goto _test_eof9;
case 9:
#line 614 "parser.c"
_widec = (*p);
if ( (*p) < 13 ) {
if ( (*p) > 9 ) {
if ( 10 <= (*p) && (*p) <= 10 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) >= 9 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) > 13 ) {
if ( (*p) < 44 ) {
if ( 32 <= (*p) && (*p) <= 32 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) > 44 ) {
if ( 47 <= (*p) && (*p) <= 47 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 125: goto tr4;
case 269: goto st10;
case 288: goto st10;
case 300: goto st11;
case 303: goto st16;
case 525: goto st9;
case 544: goto st9;
case 556: goto st2;
case 559: goto st20;
}
if ( _widec > 266 ) {
if ( 521 <= _widec && _widec <= 522 )
goto st9;
} else if ( _widec >= 265 )
goto st10;
goto st0;
tr4:
#line 505 "parser.rl"
{ p--; {p++; cs = 32; goto _out;} }
goto st32;
st32:
if ( ++p == pe )
goto _test_eof32;
case 32:
#line 682 "parser.c"
goto st0;
st10:
if ( ++p == pe )
goto _test_eof10;
case 10:
switch( (*p) ) {
case 13: goto st10;
case 32: goto st10;
case 44: goto st11;
case 47: goto st16;
case 125: goto tr4;
}
if ( 9 <= (*p) && (*p) <= 10 )
goto st10;
goto st0;
st11:
if ( ++p == pe )
goto _test_eof11;
case 11:
switch( (*p) ) {
case 13: goto st11;
case 32: goto st11;
case 34: goto tr2;
case 47: goto st12;
}
if ( 9 <= (*p) && (*p) <= 10 )
goto st11;
goto st0;
st12:
if ( ++p == pe )
goto _test_eof12;
case 12:
switch( (*p) ) {
case 42: goto st13;
case 47: goto st15;
}
goto st0;
st13:
if ( ++p == pe )
goto _test_eof13;
case 13:
if ( (*p) == 42 )
goto st14;
goto st13;
st14:
if ( ++p == pe )
goto _test_eof14;
case 14:
switch( (*p) ) {
case 42: goto st14;
case 47: goto st11;
}
goto st13;
st15:
if ( ++p == pe )
goto _test_eof15;
case 15:
if ( (*p) == 10 )
goto st11;
goto st15;
st16:
if ( ++p == pe )
goto _test_eof16;
case 16:
switch( (*p) ) {
case 42: goto st17;
case 47: goto st19;
}
goto st0;
st17:
if ( ++p == pe )
goto _test_eof17;
case 17:
if ( (*p) == 42 )
goto st18;
goto st17;
st18:
if ( ++p == pe )
goto _test_eof18;
case 18:
switch( (*p) ) {
case 42: goto st18;
case 47: goto st10;
}
goto st17;
st19:
if ( ++p == pe )
goto _test_eof19;
case 19:
if ( (*p) == 10 )
goto st10;
goto st19;
st20:
if ( ++p == pe )
goto _test_eof20;
case 20:
_widec = (*p);
if ( (*p) > 42 ) {
if ( 47 <= (*p) && (*p) <= 47 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) >= 42 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 298: goto st17;
case 303: goto st19;
case 554: goto st21;
case 559: goto st23;
}
goto st0;
st21:
if ( ++p == pe )
goto _test_eof21;
case 21:
_widec = (*p);
if ( (*p) < 42 ) {
if ( (*p) <= 41 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) > 42 ) {
if ( 43 <= (*p) )
{ _widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 298: goto st18;
case 554: goto st22;
}
if ( _widec > 383 ) {
if ( 384 <= _widec && _widec <= 639 )
goto st21;
} else if ( _widec >= 128 )
goto st17;
goto st0;
st22:
if ( ++p == pe )
goto _test_eof22;
case 22:
_widec = (*p);
if ( (*p) < 43 ) {
if ( (*p) > 41 ) {
if ( 42 <= (*p) && (*p) <= 42 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) > 46 ) {
if ( (*p) > 47 ) {
if ( 48 <= (*p) )
{ _widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) >= 47 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 298: goto st18;
case 303: goto st10;
case 554: goto st22;
case 559: goto st9;
}
if ( _widec > 383 ) {
if ( 384 <= _widec && _widec <= 639 )
goto st21;
} else if ( _widec >= 128 )
goto st17;
goto st0;
st23:
if ( ++p == pe )
goto _test_eof23;
case 23:
_widec = (*p);
if ( (*p) < 10 ) {
if ( (*p) <= 9 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) > 10 ) {
if ( 11 <= (*p) )
{ _widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 492 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 266: goto st10;
case 522: goto st9;
}
if ( _widec > 383 ) {
if ( 384 <= _widec && _widec <= 639 )
goto st23;
} else if ( _widec >= 128 )
goto st19;
goto st0;
st24:
if ( ++p == pe )
goto _test_eof24;
case 24:
switch( (*p) ) {
case 42: goto st25;
case 47: goto st27;
}
goto st0;
st25:
if ( ++p == pe )
goto _test_eof25;
case 25:
if ( (*p) == 42 )
goto st26;
goto st25;
st26:
if ( ++p == pe )
goto _test_eof26;
case 26:
switch( (*p) ) {
case 42: goto st26;
case 47: goto st8;
}
goto st25;
st27:
if ( ++p == pe )
goto _test_eof27;
case 27:
if ( (*p) == 10 )
goto st8;
goto st27;
st28:
if ( ++p == pe )
goto _test_eof28;
case 28:
switch( (*p) ) {
case 42: goto st29;
case 47: goto st31;
}
goto st0;
st29:
if ( ++p == pe )
goto _test_eof29;
case 29:
if ( (*p) == 42 )
goto st30;
goto st29;
st30:
if ( ++p == pe )
goto _test_eof30;
case 30:
switch( (*p) ) {
case 42: goto st30;
case 47: goto st2;
}
goto st29;
st31:
if ( ++p == pe )
goto _test_eof31;
case 31:
if ( (*p) == 10 )
goto st2;
goto st31;
}
_test_eof2: cs = 2; goto _test_eof;
_test_eof3: cs = 3; goto _test_eof;
_test_eof4: cs = 4; goto _test_eof;
_test_eof5: cs = 5; goto _test_eof;
_test_eof6: cs = 6; goto _test_eof;
_test_eof7: cs = 7; goto _test_eof;
_test_eof8: cs = 8; goto _test_eof;
_test_eof9: cs = 9; goto _test_eof;
_test_eof32: cs = 32; goto _test_eof;
_test_eof10: cs = 10; goto _test_eof;
_test_eof11: cs = 11; goto _test_eof;
_test_eof12: cs = 12; goto _test_eof;
_test_eof13: cs = 13; goto _test_eof;
_test_eof14: cs = 14; goto _test_eof;
_test_eof15: cs = 15; goto _test_eof;
_test_eof16: cs = 16; goto _test_eof;
_test_eof17: cs = 17; goto _test_eof;
_test_eof18: cs = 18; goto _test_eof;
_test_eof19: cs = 19; goto _test_eof;
_test_eof20: cs = 20; goto _test_eof;
_test_eof21: cs = 21; goto _test_eof;
_test_eof22: cs = 22; goto _test_eof;
_test_eof23: cs = 23; goto _test_eof;
_test_eof24: cs = 24; goto _test_eof;
_test_eof25: cs = 25; goto _test_eof;
_test_eof26: cs = 26; goto _test_eof;
_test_eof27: cs = 27; goto _test_eof;
_test_eof28: cs = 28; goto _test_eof;
_test_eof29: cs = 29; goto _test_eof;
_test_eof30: cs = 30; goto _test_eof;
_test_eof31: cs = 31; goto _test_eof;
_test_eof: {}
_out: {}
}
#line 531 "parser.rl"
if (cs >= JSON_object_first_final) {
long count = json->stack->head - stack_head;
if (RB_UNLIKELY(json->object_class)) {
VALUE object = rb_class_new_instance(0, 0, json->object_class);
long index = 0;
VALUE *items = rvalue_stack_peek(json->stack, count);
while (index < count) {
VALUE name = items[index++];
VALUE value = items[index++];
rb_funcall(object, i_aset, 2, name, value);
}
*result = object;
} else {
VALUE hash;
#ifdef HAVE_RB_HASH_NEW_CAPA
hash = rb_hash_new_capa(count >> 1);
#else
hash = rb_hash_new();
#endif
rb_hash_bulk_insert(count, rvalue_stack_peek(json->stack, count), hash);
*result = hash;
}
rvalue_stack_pop(json->stack, count);
if (RB_UNLIKELY(json->create_additions)) {
VALUE klassname;
if (json->object_class) {
klassname = rb_funcall(*result, i_aref, 1, json->create_id);
} else {
klassname = rb_hash_aref(*result, json->create_id);
}
if (!NIL_P(klassname)) {
VALUE klass = rb_funcall(mJSON, i_deep_const_get, 1, klassname);
if (RTEST(rb_funcall(klass, i_json_creatable_p, 0))) {
if (json->deprecated_create_additions) {
json_deprecated(deprecated_create_additions_warning);
}
*result = rb_funcall(klass, i_json_create, 1, *result);
}
}
}
return p + 1;
} else {
return NULL;
}
}
#line 1073 "parser.c"
enum {JSON_value_start = 1};
enum {JSON_value_first_final = 29};
enum {JSON_value_error = 0};
enum {JSON_value_en_main = 1};
#line 664 "parser.rl"
static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result, int current_nesting)
{
int cs = EVIL;
#line 1089 "parser.c"
{
cs = JSON_value_start;
}
#line 671 "parser.rl"
#line 1096 "parser.c"
{
if ( p == pe )
goto _test_eof;
switch ( cs )
{
st1:
if ( ++p == pe )
goto _test_eof1;
case 1:
switch( (*p) ) {
case 13: goto st1;
case 32: goto st1;
case 34: goto tr2;
case 45: goto tr3;
case 47: goto st6;
case 73: goto st10;
case 78: goto st17;
case 91: goto tr7;
case 102: goto st19;
case 110: goto st23;
case 116: goto st26;
case 123: goto tr11;
}
if ( (*p) > 10 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto tr3;
} else if ( (*p) >= 9 )
goto st1;
goto st0;
st0:
cs = 0;
goto _out;
tr2:
#line 609 "parser.rl"
{
char *np = JSON_parse_string(json, p, pe, result);
if (np == NULL) {
p--;
{p++; cs = 29; goto _out;}
} else {
{p = (( np))-1;}
}
}
goto st29;
tr3:
#line 619 "parser.rl"
{
char *np;
if(pe > p + 8 && !strncmp(MinusInfinity, p, 9)) {
if (json->allow_nan) {
*result = CMinusInfinity;
{p = (( p + 10))-1;}
p--; {p++; cs = 29; goto _out;}
} else {
raise_parse_error("unexpected token at '%s'", p);
}
}
np = JSON_parse_number(json, p, pe, result);
if (np != NULL) {
{p = (( np))-1;}
}
p--; {p++; cs = 29; goto _out;}
}
goto st29;
tr7:
#line 637 "parser.rl"
{
char *np;
json->in_array++;
np = JSON_parse_array(json, p, pe, result, current_nesting + 1);
json->in_array--;
if (np == NULL) { p--; {p++; cs = 29; goto _out;} } else {p = (( np))-1;}
}
goto st29;
tr11:
#line 645 "parser.rl"
{
char *np;
np = JSON_parse_object(json, p, pe, result, current_nesting + 1);
if (np == NULL) { p--; {p++; cs = 29; goto _out;} } else {p = (( np))-1;}
}
goto st29;
tr25:
#line 602 "parser.rl"
{
if (json->allow_nan) {
*result = CInfinity;
} else {
raise_parse_error("unexpected token at '%s'", p - 7);
}
}
goto st29;
tr27:
#line 595 "parser.rl"
{
if (json->allow_nan) {
*result = CNaN;
} else {
raise_parse_error("unexpected token at '%s'", p - 2);
}
}
goto st29;
tr31:
#line 589 "parser.rl"
{
*result = Qfalse;
}
goto st29;
tr34:
#line 586 "parser.rl"
{
*result = Qnil;
}
goto st29;
tr37:
#line 592 "parser.rl"
{
*result = Qtrue;
}
goto st29;
st29:
if ( ++p == pe )
goto _test_eof29;
case 29:
#line 651 "parser.rl"
{ p--; {p++; cs = 29; goto _out;} }
#line 1223 "parser.c"
switch( (*p) ) {
case 13: goto st29;
case 32: goto st29;
case 47: goto st2;
}
if ( 9 <= (*p) && (*p) <= 10 )
goto st29;
goto st0;
st2:
if ( ++p == pe )
goto _test_eof2;
case 2:
switch( (*p) ) {
case 42: goto st3;
case 47: goto st5;
}
goto st0;
st3:
if ( ++p == pe )
goto _test_eof3;
case 3:
if ( (*p) == 42 )
goto st4;
goto st3;
st4:
if ( ++p == pe )
goto _test_eof4;
case 4:
switch( (*p) ) {
case 42: goto st4;
case 47: goto st29;
}
goto st3;
st5:
if ( ++p == pe )
goto _test_eof5;
case 5:
if ( (*p) == 10 )
goto st29;
goto st5;
st6:
if ( ++p == pe )
goto _test_eof6;
case 6:
switch( (*p) ) {
case 42: goto st7;
case 47: goto st9;
}
goto st0;
st7:
if ( ++p == pe )
goto _test_eof7;
case 7:
if ( (*p) == 42 )
goto st8;
goto st7;
st8:
if ( ++p == pe )
goto _test_eof8;
case 8:
switch( (*p) ) {
case 42: goto st8;
case 47: goto st1;
}
goto st7;
st9:
if ( ++p == pe )
goto _test_eof9;
case 9:
if ( (*p) == 10 )
goto st1;
goto st9;
st10:
if ( ++p == pe )
goto _test_eof10;
case 10:
if ( (*p) == 110 )
goto st11;
goto st0;
st11:
if ( ++p == pe )
goto _test_eof11;
case 11:
if ( (*p) == 102 )
goto st12;
goto st0;
st12:
if ( ++p == pe )
goto _test_eof12;
case 12:
if ( (*p) == 105 )
goto st13;
goto st0;
st13:
if ( ++p == pe )
goto _test_eof13;
case 13:
if ( (*p) == 110 )
goto st14;
goto st0;
st14:
if ( ++p == pe )
goto _test_eof14;
case 14:
if ( (*p) == 105 )
goto st15;
goto st0;
st15:
if ( ++p == pe )
goto _test_eof15;
case 15:
if ( (*p) == 116 )
goto st16;
goto st0;
st16:
if ( ++p == pe )
goto _test_eof16;
case 16:
if ( (*p) == 121 )
goto tr25;
goto st0;
st17:
if ( ++p == pe )
goto _test_eof17;
case 17:
if ( (*p) == 97 )
goto st18;
goto st0;
st18:
if ( ++p == pe )
goto _test_eof18;
case 18:
if ( (*p) == 78 )
goto tr27;
goto st0;
st19:
if ( ++p == pe )
goto _test_eof19;
case 19:
if ( (*p) == 97 )
goto st20;
goto st0;
st20:
if ( ++p == pe )
goto _test_eof20;
case 20:
if ( (*p) == 108 )
goto st21;
goto st0;
st21:
if ( ++p == pe )
goto _test_eof21;
case 21:
if ( (*p) == 115 )
goto st22;
goto st0;
st22:
if ( ++p == pe )
goto _test_eof22;
case 22:
if ( (*p) == 101 )
goto tr31;
goto st0;
st23:
if ( ++p == pe )
goto _test_eof23;
case 23:
if ( (*p) == 117 )
goto st24;
goto st0;
st24:
if ( ++p == pe )
goto _test_eof24;
case 24:
if ( (*p) == 108 )
goto st25;
goto st0;
st25:
if ( ++p == pe )
goto _test_eof25;
case 25:
if ( (*p) == 108 )
goto tr34;
goto st0;
st26:
if ( ++p == pe )
goto _test_eof26;
case 26:
if ( (*p) == 114 )
goto st27;
goto st0;
st27:
if ( ++p == pe )
goto _test_eof27;
case 27:
if ( (*p) == 117 )
goto st28;
goto st0;
st28:
if ( ++p == pe )
goto _test_eof28;
case 28:
if ( (*p) == 101 )
goto tr37;
goto st0;
}
_test_eof1: cs = 1; goto _test_eof;
_test_eof29: cs = 29; goto _test_eof;
_test_eof2: cs = 2; goto _test_eof;
_test_eof3: cs = 3; goto _test_eof;
_test_eof4: cs = 4; goto _test_eof;
_test_eof5: cs = 5; goto _test_eof;
_test_eof6: cs = 6; goto _test_eof;
_test_eof7: cs = 7; goto _test_eof;
_test_eof8: cs = 8; goto _test_eof;
_test_eof9: cs = 9; goto _test_eof;
_test_eof10: cs = 10; goto _test_eof;
_test_eof11: cs = 11; goto _test_eof;
_test_eof12: cs = 12; goto _test_eof;
_test_eof13: cs = 13; goto _test_eof;
_test_eof14: cs = 14; goto _test_eof;
_test_eof15: cs = 15; goto _test_eof;
_test_eof16: cs = 16; goto _test_eof;
_test_eof17: cs = 17; goto _test_eof;
_test_eof18: cs = 18; goto _test_eof;
_test_eof19: cs = 19; goto _test_eof;
_test_eof20: cs = 20; goto _test_eof;
_test_eof21: cs = 21; goto _test_eof;
_test_eof22: cs = 22; goto _test_eof;
_test_eof23: cs = 23; goto _test_eof;
_test_eof24: cs = 24; goto _test_eof;
_test_eof25: cs = 25; goto _test_eof;
_test_eof26: cs = 26; goto _test_eof;
_test_eof27: cs = 27; goto _test_eof;
_test_eof28: cs = 28; goto _test_eof;
_test_eof: {}
_out: {}
}
#line 672 "parser.rl"
if (json->freeze) {
OBJ_FREEZE(*result);
}
if (cs >= JSON_value_first_final) {
PUSH(*result);
return p;
} else {
return NULL;
}
}
#line 1479 "parser.c"
enum {JSON_integer_start = 1};
enum {JSON_integer_first_final = 3};
enum {JSON_integer_error = 0};
enum {JSON_integer_en_main = 1};
#line 693 "parser.rl"
#define MAX_FAST_INTEGER_SIZE 18
static inline VALUE fast_parse_integer(char *p, char *pe)
{
bool negative = false;
if (*p == '-') {
negative = true;
p++;
}
long long memo = 0;
while (p < pe) {
memo *= 10;
memo += *p - '0';
p++;
}
if (negative) {
memo = -memo;
}
return LL2NUM(memo);
}
static char *JSON_decode_integer(JSON_Parser *json, char *p, VALUE *result)
{
long len = p - json->memo;
if (RB_LIKELY(len < MAX_FAST_INTEGER_SIZE)) {
*result = fast_parse_integer(json->memo, p);
} else {
fbuffer_clear(&json->fbuffer);
fbuffer_append(&json->fbuffer, json->memo, len);
fbuffer_append_char(&json->fbuffer, '\0');
*result = rb_cstr2inum(FBUFFER_PTR(&json->fbuffer), 10);
}
return p + 1;
}
#line 1527 "parser.c"
enum {JSON_float_start = 1};
enum {JSON_float_first_final = 6};
enum {JSON_float_error = 0};
enum {JSON_float_en_main = 1};
#line 745 "parser.rl"
static char *JSON_parse_number(JSON_Parser *json, char *p, char *pe, VALUE *result)
{
int cs = EVIL;
bool is_float = false;
#line 1544 "parser.c"
{
cs = JSON_float_start;
}
#line 753 "parser.rl"
json->memo = p;
#line 1552 "parser.c"
{
if ( p == pe )
goto _test_eof;
switch ( cs )
{
case 1:
switch( (*p) ) {
case 45: goto st2;
case 48: goto st6;
}
if ( 49 <= (*p) && (*p) <= 57 )
goto st10;
goto st0;
st0:
cs = 0;
goto _out;
st2:
if ( ++p == pe )
goto _test_eof2;
case 2:
if ( (*p) == 48 )
goto st6;
if ( 49 <= (*p) && (*p) <= 57 )
goto st10;
goto st0;
st6:
if ( ++p == pe )
goto _test_eof6;
case 6:
switch( (*p) ) {
case 45: goto st0;
case 46: goto tr8;
case 69: goto tr9;
case 101: goto tr9;
}
if ( 48 <= (*p) && (*p) <= 57 )
goto st0;
goto tr7;
tr7:
#line 737 "parser.rl"
{ p--; {p++; cs = 7; goto _out;} }
goto st7;
st7:
if ( ++p == pe )
goto _test_eof7;
case 7:
#line 1599 "parser.c"
goto st0;
tr8:
#line 738 "parser.rl"
{ is_float = true; }
goto st3;
st3:
if ( ++p == pe )
goto _test_eof3;
case 3:
#line 1609 "parser.c"
if ( 48 <= (*p) && (*p) <= 57 )
goto st8;
goto st0;
st8:
if ( ++p == pe )
goto _test_eof8;
case 8:
switch( (*p) ) {
case 69: goto st4;
case 101: goto st4;
}
if ( (*p) > 46 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto st8;
} else if ( (*p) >= 45 )
goto st0;
goto tr7;
tr9:
#line 738 "parser.rl"
{ is_float = true; }
goto st4;
st4:
if ( ++p == pe )
goto _test_eof4;
case 4:
#line 1635 "parser.c"
switch( (*p) ) {
case 43: goto st5;
case 45: goto st5;
}
if ( 48 <= (*p) && (*p) <= 57 )
goto st9;
goto st0;
st5:
if ( ++p == pe )
goto _test_eof5;
case 5:
if ( 48 <= (*p) && (*p) <= 57 )
goto st9;
goto st0;
st9:
if ( ++p == pe )
goto _test_eof9;
case 9:
switch( (*p) ) {
case 69: goto st0;
case 101: goto st0;
}
if ( (*p) > 46 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto st9;
} else if ( (*p) >= 45 )
goto st0;
goto tr7;
st10:
if ( ++p == pe )
goto _test_eof10;
case 10:
switch( (*p) ) {
case 45: goto st0;
case 46: goto tr8;
case 69: goto tr9;
case 101: goto tr9;
}
if ( 48 <= (*p) && (*p) <= 57 )
goto st10;
goto tr7;
}
_test_eof2: cs = 2; goto _test_eof;
_test_eof6: cs = 6; goto _test_eof;
_test_eof7: cs = 7; goto _test_eof;
_test_eof3: cs = 3; goto _test_eof;
_test_eof8: cs = 8; goto _test_eof;
_test_eof4: cs = 4; goto _test_eof;
_test_eof5: cs = 5; goto _test_eof;
_test_eof9: cs = 9; goto _test_eof;
_test_eof10: cs = 10; goto _test_eof;
_test_eof: {}
_out: {}
}
#line 755 "parser.rl"
if (cs >= JSON_float_first_final) {
if (!is_float) {
return JSON_decode_integer(json, p, result);
}
VALUE mod = Qnil;
ID method_id = 0;
if (json->decimal_class) {
if (rb_respond_to(json->decimal_class, i_try_convert)) {
mod = json->decimal_class;
method_id = i_try_convert;
} else if (rb_respond_to(json->decimal_class, i_new)) {
mod = json->decimal_class;
method_id = i_new;
} else if (RB_TYPE_P(json->decimal_class, T_CLASS)) {
VALUE name = rb_class_name(json->decimal_class);
const char *name_cstr = RSTRING_PTR(name);
const char *last_colon = strrchr(name_cstr, ':');
if (last_colon) {
const char *mod_path_end = last_colon - 1;
VALUE mod_path = rb_str_substr(name, 0, mod_path_end - name_cstr);
mod = rb_path_to_class(mod_path);
const char *method_name_beg = last_colon + 1;
long before_len = method_name_beg - name_cstr;
long len = RSTRING_LEN(name) - before_len;
VALUE method_name = rb_str_substr(name, before_len, len);
method_id = SYM2ID(rb_str_intern(method_name));
} else {
mod = rb_mKernel;
method_id = SYM2ID(rb_str_intern(name));
}
}
}
long len = p - json->memo;
fbuffer_clear(&json->fbuffer);
fbuffer_append(&json->fbuffer, json->memo, len);
fbuffer_append_char(&json->fbuffer, '\0');
if (method_id) {
VALUE text = rb_str_new2(FBUFFER_PTR(&json->fbuffer));
*result = rb_funcallv(mod, method_id, 1, &text);
} else {
*result = DBL2NUM(rb_cstr_to_dbl(FBUFFER_PTR(&json->fbuffer), 1));
}
return p + 1;
} else {
return NULL;
}
}
#line 1748 "parser.c"
enum {JSON_array_start = 1};
enum {JSON_array_first_final = 22};
enum {JSON_array_error = 0};
enum {JSON_array_en_main = 1};
#line 835 "parser.rl"
static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result, int current_nesting)
{
int cs = EVIL;
if (json->max_nesting && current_nesting > json->max_nesting) {
rb_raise(eNestingError, "nesting of %d is too deep", current_nesting);
}
long stack_head = json->stack->head;
#line 1769 "parser.c"
{
cs = JSON_array_start;
}
#line 847 "parser.rl"
#line 1776 "parser.c"
{
short _widec;
if ( p == pe )
goto _test_eof;
switch ( cs )
{
case 1:
if ( (*p) == 91 )
goto st2;
goto st0;
st0:
cs = 0;
goto _out;
st2:
if ( ++p == pe )
goto _test_eof2;
case 2:
switch( (*p) ) {
case 13: goto st2;
case 32: goto st2;
case 34: goto tr2;
case 45: goto tr2;
case 47: goto st18;
case 73: goto tr2;
case 78: goto tr2;
case 91: goto tr2;
case 93: goto tr4;
case 102: goto tr2;
case 110: goto tr2;
case 116: goto tr2;
case 123: goto tr2;
}
if ( (*p) > 10 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto tr2;
} else if ( (*p) >= 9 )
goto st2;
goto st0;
tr2:
#line 815 "parser.rl"
{
VALUE v = Qnil;
char *np = JSON_parse_value(json, p, pe, &v, current_nesting);
if (np == NULL) {
p--; {p++; cs = 3; goto _out;}
} else {
{p = (( np))-1;}
}
}
goto st3;
st3:
if ( ++p == pe )
goto _test_eof3;
case 3:
#line 1831 "parser.c"
_widec = (*p);
if ( 44 <= (*p) && (*p) <= 44 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 13: goto st3;
case 32: goto st3;
case 47: goto st4;
case 93: goto tr4;
case 300: goto st8;
case 556: goto st13;
}
if ( 9 <= _widec && _widec <= 10 )
goto st3;
goto st0;
st4:
if ( ++p == pe )
goto _test_eof4;
case 4:
switch( (*p) ) {
case 42: goto st5;
case 47: goto st7;
}
goto st0;
st5:
if ( ++p == pe )
goto _test_eof5;
case 5:
if ( (*p) == 42 )
goto st6;
goto st5;
st6:
if ( ++p == pe )
goto _test_eof6;
case 6:
switch( (*p) ) {
case 42: goto st6;
case 47: goto st3;
}
goto st5;
st7:
if ( ++p == pe )
goto _test_eof7;
case 7:
if ( (*p) == 10 )
goto st3;
goto st7;
tr4:
#line 827 "parser.rl"
{ p--; {p++; cs = 22; goto _out;} }
goto st22;
st22:
if ( ++p == pe )
goto _test_eof22;
case 22:
#line 1890 "parser.c"
goto st0;
st8:
if ( ++p == pe )
goto _test_eof8;
case 8:
switch( (*p) ) {
case 13: goto st8;
case 32: goto st8;
case 34: goto tr2;
case 45: goto tr2;
case 47: goto st9;
case 73: goto tr2;
case 78: goto tr2;
case 91: goto tr2;
case 102: goto tr2;
case 110: goto tr2;
case 116: goto tr2;
case 123: goto tr2;
}
if ( (*p) > 10 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto tr2;
} else if ( (*p) >= 9 )
goto st8;
goto st0;
st9:
if ( ++p == pe )
goto _test_eof9;
case 9:
switch( (*p) ) {
case 42: goto st10;
case 47: goto st12;
}
goto st0;
st10:
if ( ++p == pe )
goto _test_eof10;
case 10:
if ( (*p) == 42 )
goto st11;
goto st10;
st11:
if ( ++p == pe )
goto _test_eof11;
case 11:
switch( (*p) ) {
case 42: goto st11;
case 47: goto st8;
}
goto st10;
st12:
if ( ++p == pe )
goto _test_eof12;
case 12:
if ( (*p) == 10 )
goto st8;
goto st12;
st13:
if ( ++p == pe )
goto _test_eof13;
case 13:
_widec = (*p);
if ( (*p) < 13 ) {
if ( (*p) > 9 ) {
if ( 10 <= (*p) && (*p) <= 10 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) >= 9 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) > 13 ) {
if ( (*p) > 32 ) {
if ( 47 <= (*p) && (*p) <= 47 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) >= 32 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 34: goto tr2;
case 45: goto tr2;
case 73: goto tr2;
case 78: goto tr2;
case 91: goto tr2;
case 93: goto tr4;
case 102: goto tr2;
case 110: goto tr2;
case 116: goto tr2;
case 123: goto tr2;
case 269: goto st8;
case 288: goto st8;
case 303: goto st9;
case 525: goto st13;
case 544: goto st13;
case 559: goto st14;
}
if ( _widec < 265 ) {
if ( 48 <= _widec && _widec <= 57 )
goto tr2;
} else if ( _widec > 266 ) {
if ( 521 <= _widec && _widec <= 522 )
goto st13;
} else
goto st8;
goto st0;
st14:
if ( ++p == pe )
goto _test_eof14;
case 14:
_widec = (*p);
if ( (*p) > 42 ) {
if ( 47 <= (*p) && (*p) <= 47 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) >= 42 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 298: goto st10;
case 303: goto st12;
case 554: goto st15;
case 559: goto st17;
}
goto st0;
st15:
if ( ++p == pe )
goto _test_eof15;
case 15:
_widec = (*p);
if ( (*p) < 42 ) {
if ( (*p) <= 41 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) > 42 ) {
if ( 43 <= (*p) )
{ _widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 298: goto st11;
case 554: goto st16;
}
if ( _widec > 383 ) {
if ( 384 <= _widec && _widec <= 639 )
goto st15;
} else if ( _widec >= 128 )
goto st10;
goto st0;
st16:
if ( ++p == pe )
goto _test_eof16;
case 16:
_widec = (*p);
if ( (*p) < 43 ) {
if ( (*p) > 41 ) {
if ( 42 <= (*p) && (*p) <= 42 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) > 46 ) {
if ( (*p) > 47 ) {
if ( 48 <= (*p) )
{ _widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) >= 47 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 298: goto st11;
case 303: goto st8;
case 554: goto st16;
case 559: goto st13;
}
if ( _widec > 383 ) {
if ( 384 <= _widec && _widec <= 639 )
goto st15;
} else if ( _widec >= 128 )
goto st10;
goto st0;
st17:
if ( ++p == pe )
goto _test_eof17;
case 17:
_widec = (*p);
if ( (*p) < 10 ) {
if ( (*p) <= 9 ) {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else if ( (*p) > 10 ) {
if ( 11 <= (*p) )
{ _widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
} else {
_widec = (short)(128 + ((*p) - -128));
if (
#line 825 "parser.rl"
json->allow_trailing_comma ) _widec += 256;
}
switch( _widec ) {
case 266: goto st8;
case 522: goto st13;
}
if ( _widec > 383 ) {
if ( 384 <= _widec && _widec <= 639 )
goto st17;
} else if ( _widec >= 128 )
goto st12;
goto st0;
st18:
if ( ++p == pe )
goto _test_eof18;
case 18:
switch( (*p) ) {
case 42: goto st19;
case 47: goto st21;
}
goto st0;
st19:
if ( ++p == pe )
goto _test_eof19;
case 19:
if ( (*p) == 42 )
goto st20;
goto st19;
st20:
if ( ++p == pe )
goto _test_eof20;
case 20:
switch( (*p) ) {
case 42: goto st20;
case 47: goto st2;
}
goto st19;
st21:
if ( ++p == pe )
goto _test_eof21;
case 21:
if ( (*p) == 10 )
goto st2;
goto st21;
}
_test_eof2: cs = 2; goto _test_eof;
_test_eof3: cs = 3; goto _test_eof;
_test_eof4: cs = 4; goto _test_eof;
_test_eof5: cs = 5; goto _test_eof;
_test_eof6: cs = 6; goto _test_eof;
_test_eof7: cs = 7; goto _test_eof;
_test_eof22: cs = 22; goto _test_eof;
_test_eof8: cs = 8; goto _test_eof;
_test_eof9: cs = 9; goto _test_eof;
_test_eof10: cs = 10; goto _test_eof;
_test_eof11: cs = 11; goto _test_eof;
_test_eof12: cs = 12; goto _test_eof;
_test_eof13: cs = 13; goto _test_eof;
_test_eof14: cs = 14; goto _test_eof;
_test_eof15: cs = 15; goto _test_eof;
_test_eof16: cs = 16; goto _test_eof;
_test_eof17: cs = 17; goto _test_eof;
_test_eof18: cs = 18; goto _test_eof;
_test_eof19: cs = 19; goto _test_eof;
_test_eof20: cs = 20; goto _test_eof;
_test_eof21: cs = 21; goto _test_eof;
_test_eof: {}
_out: {}
}
#line 848 "parser.rl"
if(cs >= JSON_array_first_final) {
long count = json->stack->head - stack_head;
if (RB_UNLIKELY(json->array_class)) {
VALUE array = rb_class_new_instance(0, 0, json->array_class);
VALUE *items = rvalue_stack_peek(json->stack, count);
long index;
for (index = 0; index < count; index++) {
rb_funcall(array, i_leftshift, 1, items[index]);
}
*result = array;
} else {
VALUE array = rb_ary_new_from_values(count, rvalue_stack_peek(json->stack, count));
*result = array;
}
rvalue_stack_pop(json->stack, count);
return p + 1;
} else {
raise_parse_error("unexpected token at '%s'", p);
return NULL;
}
}
static inline VALUE build_string(const char *start, const char *end, bool intern, bool symbolize)
{
if (symbolize) {
intern = true;
}
VALUE result;
# ifdef HAVE_RB_ENC_INTERNED_STR
if (intern) {
result = rb_enc_interned_str(start, (long)(end - start), enc_utf8);
} else {
result = rb_utf8_str_new(start, (long)(end - start));
}
# else
result = rb_utf8_str_new(start, (long)(end - start));
if (intern) {
result = rb_funcall(rb_str_freeze(result), i_uminus, 0);
}
# endif
if (symbolize) {
result = rb_str_intern(result);
}
return result;
}
static VALUE json_string_fastpath(JSON_Parser *json, char *string, char *stringEnd, bool is_name, bool intern, bool symbolize)
{
size_t bufferSize = stringEnd - string;
if (is_name && json->in_array) {
VALUE cached_key;
if (RB_UNLIKELY(symbolize)) {
cached_key = rsymbol_cache_fetch(&json->name_cache, string, bufferSize);
} else {
cached_key = rstring_cache_fetch(&json->name_cache, string, bufferSize);
}
if (RB_LIKELY(cached_key)) {
return cached_key;
}
}
return build_string(string, stringEnd, intern, symbolize);
}
static VALUE json_string_unescape(JSON_Parser *json, char *string, char *stringEnd, bool is_name, bool intern, bool symbolize)
{
size_t bufferSize = stringEnd - string;
char *p = string, *pe = string, *unescape, *bufferStart, *buffer;
int unescape_len;
char buf[4];
if (is_name && json->in_array) {
VALUE cached_key;
if (RB_UNLIKELY(symbolize)) {
cached_key = rsymbol_cache_fetch(&json->name_cache, string, bufferSize);
} else {
cached_key = rstring_cache_fetch(&json->name_cache, string, bufferSize);
}
if (RB_LIKELY(cached_key)) {
return cached_key;
}
}
pe = memchr(p, '\\', bufferSize);
if (RB_UNLIKELY(pe == NULL)) {
return build_string(string, stringEnd, intern, symbolize);
}
VALUE result = rb_str_buf_new(bufferSize);
rb_enc_associate_index(result, utf8_encindex);
buffer = bufferStart = RSTRING_PTR(result);
while (pe < stringEnd) {
if (*pe == '\\') {
unescape = (char *) "?";
unescape_len = 1;
if (pe > p) {
MEMCPY(buffer, p, char, pe - p);
buffer += 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 '\\':
unescape = (char *) "\\";
break;
case 'b':
unescape = (char *) "\b";
break;
case 'f':
unescape = (char *) "\f";
break;
case 'u':
if (pe > stringEnd - 4) {
raise_parse_error("incomplete unicode character escape sequence at '%s'", p);
} else {
uint32_t ch = unescape_unicode((unsigned char *) ++pe);
pe += 3;
/* To handle values above U+FFFF, we take a sequence of
* \uXXXX escapes in the U+D800..U+DBFF then
* U+DC00..U+DFFF ranges, take the low 10 bits from each
* to make a 20-bit number, then add 0x10000 to get the
* final codepoint.
*
* See Unicode 15: 3.8 "Surrogates", 5.3 "Handling
* Surrogate Pairs in UTF-16", and 23.6 "Surrogates
* Area".
*/
if ((ch & 0xFC00) == 0xD800) {
pe++;
if (pe > stringEnd - 6) {
raise_parse_error("incomplete surrogate pair at '%s'", p);
}
if (pe[0] == '\\' && pe[1] == 'u') {
uint32_t 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:
p = pe;
continue;
}
MEMCPY(buffer, unescape, char, unescape_len);
buffer += unescape_len;
p = ++pe;
} else {
pe++;
}
}
if (pe > p) {
MEMCPY(buffer, p, char, pe - p);
buffer += pe - p;
}
rb_str_set_len(result, buffer - bufferStart);
if (symbolize) {
result = rb_str_intern(result);
} else if (intern) {
result = rb_funcall(rb_str_freeze(result), i_uminus, 0);
}
return result;
}
#line 2413 "parser.c"
enum {JSON_string_start = 1};
enum {JSON_string_first_final = 9};
enum {JSON_string_error = 0};
enum {JSON_string_en_main = 1};
#line 1071 "parser.rl"
static int
match_i(VALUE regexp, VALUE klass, VALUE memo)
{
if (regexp == Qundef) return ST_STOP;
if (RTEST(rb_funcall(klass, i_json_creatable_p, 0)) &&
RTEST(rb_funcall(regexp, i_match, 1, rb_ary_entry(memo, 0)))) {
rb_ary_push(memo, klass);
return ST_STOP;
}
return ST_CONTINUE;
}
static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result)
{
int cs = EVIL;
VALUE match_string;
#line 2442 "parser.c"
{
cs = JSON_string_start;
}
#line 1091 "parser.rl"
json->memo = p;
#line 2450 "parser.c"
{
if ( p == pe )
goto _test_eof;
switch ( cs )
{
case 1:
if ( (*p) == 34 )
goto st2;
goto st0;
st0:
cs = 0;
goto _out;
st2:
if ( ++p == pe )
goto _test_eof2;
case 2:
switch( (*p) ) {
case 34: goto tr2;
case 92: goto st3;
}
if ( 0 <= (signed char)(*(p)) && (*(p)) <= 31 )
goto st0;
goto st2;
tr2:
#line 1053 "parser.rl"
{
*result = json_string_fastpath(json, json->memo + 1, p, json->parsing_name, json->parsing_name || json-> freeze, json->parsing_name && json->symbolize_names);
{p = (( p + 1))-1;}
p--;
{p++; cs = 9; goto _out;}
}
#line 1046 "parser.rl"
{
*result = json_string_unescape(json, json->memo + 1, p, json->parsing_name, json->parsing_name || json-> freeze, json->parsing_name && json->symbolize_names);
{p = (( p + 1))-1;}
p--;
{p++; cs = 9; goto _out;}
}
goto st9;
tr6:
#line 1046 "parser.rl"
{
*result = json_string_unescape(json, json->memo + 1, p, json->parsing_name, json->parsing_name || json-> freeze, json->parsing_name && json->symbolize_names);
{p = (( p + 1))-1;}
p--;
{p++; cs = 9; goto _out;}
}
goto st9;
st9:
if ( ++p == pe )
goto _test_eof9;
case 9:
#line 2503 "parser.c"
goto st0;
st3:
if ( ++p == pe )
goto _test_eof3;
case 3:
if ( (*p) == 117 )
goto st5;
if ( 0 <= (signed char)(*(p)) && (*(p)) <= 31 )
goto st0;
goto st4;
st4:
if ( ++p == pe )
goto _test_eof4;
case 4:
switch( (*p) ) {
case 34: goto tr6;
case 92: goto st3;
}
if ( 0 <= (signed char)(*(p)) && (*(p)) <= 31 )
goto st0;
goto st4;
st5:
if ( ++p == pe )
goto _test_eof5;
case 5:
if ( (*p) < 65 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto st6;
} else if ( (*p) > 70 ) {
if ( 97 <= (*p) && (*p) <= 102 )
goto st6;
} else
goto st6;
goto st0;
st6:
if ( ++p == pe )
goto _test_eof6;
case 6:
if ( (*p) < 65 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto st7;
} else if ( (*p) > 70 ) {
if ( 97 <= (*p) && (*p) <= 102 )
goto st7;
} else
goto st7;
goto st0;
st7:
if ( ++p == pe )
goto _test_eof7;
case 7:
if ( (*p) < 65 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto st8;
} else if ( (*p) > 70 ) {
if ( 97 <= (*p) && (*p) <= 102 )
goto st8;
} else
goto st8;
goto st0;
st8:
if ( ++p == pe )
goto _test_eof8;
case 8:
if ( (*p) < 65 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto st4;
} else if ( (*p) > 70 ) {
if ( 97 <= (*p) && (*p) <= 102 )
goto st4;
} else
goto st4;
goto st0;
}
_test_eof2: cs = 2; goto _test_eof;
_test_eof9: cs = 9; goto _test_eof;
_test_eof3: cs = 3; goto _test_eof;
_test_eof4: cs = 4; goto _test_eof;
_test_eof5: cs = 5; goto _test_eof;
_test_eof6: cs = 6; goto _test_eof;
_test_eof7: cs = 7; goto _test_eof;
_test_eof8: cs = 8; goto _test_eof;
_test_eof: {}
_out: {}
}
#line 1093 "parser.rl"
if (json->create_additions && RTEST(match_string = json->match_string)) {
VALUE klass;
VALUE memo = rb_ary_new2(2);
rb_ary_push(memo, *result);
rb_hash_foreach(match_string, match_i, memo);
klass = rb_ary_entry(memo, 1);
if (RTEST(klass)) {
*result = rb_funcall(klass, i_json_create, 1, *result);
}
}
if (cs >= JSON_string_first_final) {
return p + 1;
} else {
return NULL;
}
}
/*
* Document-class: JSON::Ext::Parser
*
* This is the JSON parser implemented as a C extension. It can be configured
* to be used by setting
*
* JSON.parser = JSON::Ext::Parser
*
* with the method parser= in JSON.
*
*/
static VALUE convert_encoding(VALUE source)
{
int encindex = RB_ENCODING_GET(source);
if (RB_LIKELY(encindex == utf8_encindex)) {
return source;
}
if (encindex == binary_encindex) {
// For historical reason, we silently reinterpret binary strings as UTF-8
return rb_enc_associate_index(rb_str_dup(source), utf8_encindex);
}
return rb_funcall(source, i_encode, 1, Encoding_UTF_8);
}
static int configure_parser_i(VALUE key, VALUE val, VALUE data)
{
JSON_Parser *json = (JSON_Parser *)data;
if (key == sym_max_nesting) { json->max_nesting = RTEST(val) ? FIX2INT(val) : 0; }
else if (key == sym_allow_nan) { json->allow_nan = RTEST(val); }
else if (key == sym_allow_trailing_comma) { json->allow_trailing_comma = RTEST(val); }
else if (key == sym_symbolize_names) { json->symbolize_names = RTEST(val); }
else if (key == sym_freeze) { json->freeze = RTEST(val); }
else if (key == sym_create_id) { json->create_id = RTEST(val) ? val : Qfalse; }
else if (key == sym_object_class) { json->object_class = RTEST(val) ? val : Qfalse; }
else if (key == sym_array_class) { json->array_class = RTEST(val) ? val : Qfalse; }
else if (key == sym_decimal_class) { json->decimal_class = RTEST(val) ? val : Qfalse; }
else if (key == sym_match_string) { json->match_string = RTEST(val) ? val : Qfalse; }
else if (key == sym_create_additions) {
if (NIL_P(val)) {
json->create_additions = true;
json->deprecated_create_additions = true;
} else {
json->create_additions = RTEST(val);
json->deprecated_create_additions = false;
}
}
return ST_CONTINUE;
}
static void parser_init(JSON_Parser *json, VALUE source, VALUE opts)
{
if (json->Vsource) {
rb_raise(rb_eTypeError, "already initialized instance");
}
json->fbuffer.initial_length = FBUFFER_INITIAL_LENGTH_DEFAULT;
json->max_nesting = 100;
if (!NIL_P(opts)) {
Check_Type(opts, T_HASH);
if (RHASH_SIZE(opts) > 0) {
// We assume in most cases few keys are set so it's faster to go over
// the provided keys than to check all possible keys.
rb_hash_foreach(opts, configure_parser_i, (VALUE)json);
if (json->symbolize_names && json->create_additions) {
rb_raise(rb_eArgError,
"options :symbolize_names and :create_additions cannot be "
" used in conjunction");
}
if (json->create_additions && !json->create_id) {
json->create_id = rb_funcall(mJSON, i_create_id, 0);
}
}
}
source = convert_encoding(StringValue(source));
StringValue(source);
json->len = RSTRING_LEN(source);
json->source = RSTRING_PTR(source);
json->Vsource = source;
}
/*
* call-seq: new(source, opts => {})
*
* Creates a new JSON::Ext::Parser instance for the string _source_.
*
* It will be configured by the _opts_ hash. _opts_ can have the following
* keys:
*
* _opts_ can have the following keys:
* * *max_nesting*: The maximum depth of nesting allowed in the parsed data
* structures. Disable depth checking with :max_nesting => false|nil|0, it
* defaults to 100.
* * *allow_nan*: If set to true, allow NaN, Infinity and -Infinity in
* defiance of RFC 4627 to be parsed by the Parser. This option defaults to
* false.
* * *symbolize_names*: If set to true, returns symbols for the names
* (keys) in a JSON object. Otherwise strings are returned, which is
* also the default. It's not possible to use this option in
* conjunction with the *create_additions* option.
* * *create_additions*: If set to false, the Parser doesn't create
* additions even if a matching class and create_id was found. This option
* defaults to false.
* * *object_class*: Defaults to Hash. If another type is provided, it will be used
* instead of Hash to represent JSON objects. The type must respond to
* +new+ without arguments, and return an object that respond to +[]=+.
* * *array_class*: Defaults to Array If another type is provided, it will be used
* instead of Hash to represent JSON arrays. The type must respond to
* +new+ without arguments, and return an object that respond to +<<+.
* * *decimal_class*: Specifies which class to use instead of the default
* (Float) when parsing decimal numbers. This class must accept a single
* string argument in its constructor.
*/
static VALUE cParser_initialize(int argc, VALUE *argv, VALUE self)
{
GET_PARSER_INIT;
rb_check_arity(argc, 1, 2);
parser_init(json, argv[0], argc == 2 ? argv[1] : Qnil);
return self;
}
#line 2744 "parser.c"
enum {JSON_start = 1};
enum {JSON_first_final = 10};
enum {JSON_error = 0};
enum {JSON_en_main = 1};
#line 1259 "parser.rl"
/*
* call-seq: parse()
*
* Parses the current JSON text _source_ and returns the complete data
* structure as a result.
* It raises JSON::ParserError if fail to parse.
*/
static VALUE cParser_parse(VALUE self)
{
char *p, *pe;
int cs = EVIL;
VALUE result = Qnil;
GET_PARSER;
char stack_buffer[FBUFFER_STACK_SIZE];
fbuffer_stack_init(&json->fbuffer, FBUFFER_INITIAL_LENGTH_DEFAULT, stack_buffer, FBUFFER_STACK_SIZE);
VALUE rvalue_stack_buffer[RVALUE_STACK_INITIAL_CAPA];
rvalue_stack stack = {
.type = RVALUE_STACK_STACK_ALLOCATED,
.ptr = rvalue_stack_buffer,
.capa = RVALUE_STACK_INITIAL_CAPA,
};
json->stack = &stack;
#line 2781 "parser.c"
{
cs = JSON_start;
}
#line 1287 "parser.rl"
p = json->source;
pe = p + json->len;
#line 2790 "parser.c"
{
if ( p == pe )
goto _test_eof;
switch ( cs )
{
st1:
if ( ++p == pe )
goto _test_eof1;
case 1:
switch( (*p) ) {
case 13: goto st1;
case 32: goto st1;
case 34: goto tr2;
case 45: goto tr2;
case 47: goto st6;
case 73: goto tr2;
case 78: goto tr2;
case 91: goto tr2;
case 102: goto tr2;
case 110: goto tr2;
case 116: goto tr2;
case 123: goto tr2;
}
if ( (*p) > 10 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto tr2;
} else if ( (*p) >= 9 )
goto st1;
goto st0;
st0:
cs = 0;
goto _out;
tr2:
#line 1251 "parser.rl"
{
char *np = JSON_parse_value(json, p, pe, &result, 0);
if (np == NULL) { p--; {p++; cs = 10; goto _out;} } else {p = (( np))-1;}
}
goto st10;
st10:
if ( ++p == pe )
goto _test_eof10;
case 10:
#line 2834 "parser.c"
switch( (*p) ) {
case 13: goto st10;
case 32: goto st10;
case 47: goto st2;
}
if ( 9 <= (*p) && (*p) <= 10 )
goto st10;
goto st0;
st2:
if ( ++p == pe )
goto _test_eof2;
case 2:
switch( (*p) ) {
case 42: goto st3;
case 47: goto st5;
}
goto st0;
st3:
if ( ++p == pe )
goto _test_eof3;
case 3:
if ( (*p) == 42 )
goto st4;
goto st3;
st4:
if ( ++p == pe )
goto _test_eof4;
case 4:
switch( (*p) ) {
case 42: goto st4;
case 47: goto st10;
}
goto st3;
st5:
if ( ++p == pe )
goto _test_eof5;
case 5:
if ( (*p) == 10 )
goto st10;
goto st5;
st6:
if ( ++p == pe )
goto _test_eof6;
case 6:
switch( (*p) ) {
case 42: goto st7;
case 47: goto st9;
}
goto st0;
st7:
if ( ++p == pe )
goto _test_eof7;
case 7:
if ( (*p) == 42 )
goto st8;
goto st7;
st8:
if ( ++p == pe )
goto _test_eof8;
case 8:
switch( (*p) ) {
case 42: goto st8;
case 47: goto st1;
}
goto st7;
st9:
if ( ++p == pe )
goto _test_eof9;
case 9:
if ( (*p) == 10 )
goto st1;
goto st9;
}
_test_eof1: cs = 1; goto _test_eof;
_test_eof10: cs = 10; goto _test_eof;
_test_eof2: cs = 2; goto _test_eof;
_test_eof3: cs = 3; goto _test_eof;
_test_eof4: cs = 4; goto _test_eof;
_test_eof5: cs = 5; goto _test_eof;
_test_eof6: cs = 6; goto _test_eof;
_test_eof7: cs = 7; goto _test_eof;
_test_eof8: cs = 8; goto _test_eof;
_test_eof9: cs = 9; goto _test_eof;
_test_eof: {}
_out: {}
}
#line 1290 "parser.rl"
if (json->stack_handle) {
rvalue_stack_eagerly_release(json->stack_handle);
}
if (cs >= JSON_first_final && p == pe) {
return result;
} else {
raise_parse_error("unexpected token at '%s'", p);
return Qnil;
}
}
static VALUE cParser_m_parse(VALUE klass, VALUE source, VALUE opts)
{
char *p, *pe;
int cs = EVIL;
VALUE result = Qnil;
JSON_Parser _parser = {0};
JSON_Parser *json = &_parser;
parser_init(json, source, opts);
char stack_buffer[FBUFFER_STACK_SIZE];
fbuffer_stack_init(&json->fbuffer, FBUFFER_INITIAL_LENGTH_DEFAULT, stack_buffer, FBUFFER_STACK_SIZE);
VALUE rvalue_stack_buffer[RVALUE_STACK_INITIAL_CAPA];
rvalue_stack stack = {
.type = RVALUE_STACK_STACK_ALLOCATED,
.ptr = rvalue_stack_buffer,
.capa = RVALUE_STACK_INITIAL_CAPA,
};
json->stack = &stack;
#line 2959 "parser.c"
{
cs = JSON_start;
}
#line 1325 "parser.rl"
p = json->source;
pe = p + json->len;
#line 2968 "parser.c"
{
if ( p == pe )
goto _test_eof;
switch ( cs )
{
st1:
if ( ++p == pe )
goto _test_eof1;
case 1:
switch( (*p) ) {
case 13: goto st1;
case 32: goto st1;
case 34: goto tr2;
case 45: goto tr2;
case 47: goto st6;
case 73: goto tr2;
case 78: goto tr2;
case 91: goto tr2;
case 102: goto tr2;
case 110: goto tr2;
case 116: goto tr2;
case 123: goto tr2;
}
if ( (*p) > 10 ) {
if ( 48 <= (*p) && (*p) <= 57 )
goto tr2;
} else if ( (*p) >= 9 )
goto st1;
goto st0;
st0:
cs = 0;
goto _out;
tr2:
#line 1251 "parser.rl"
{
char *np = JSON_parse_value(json, p, pe, &result, 0);
if (np == NULL) { p--; {p++; cs = 10; goto _out;} } else {p = (( np))-1;}
}
goto st10;
st10:
if ( ++p == pe )
goto _test_eof10;
case 10:
#line 3012 "parser.c"
switch( (*p) ) {
case 13: goto st10;
case 32: goto st10;
case 47: goto st2;
}
if ( 9 <= (*p) && (*p) <= 10 )
goto st10;
goto st0;
st2:
if ( ++p == pe )
goto _test_eof2;
case 2:
switch( (*p) ) {
case 42: goto st3;
case 47: goto st5;
}
goto st0;
st3:
if ( ++p == pe )
goto _test_eof3;
case 3:
if ( (*p) == 42 )
goto st4;
goto st3;
st4:
if ( ++p == pe )
goto _test_eof4;
case 4:
switch( (*p) ) {
case 42: goto st4;
case 47: goto st10;
}
goto st3;
st5:
if ( ++p == pe )
goto _test_eof5;
case 5:
if ( (*p) == 10 )
goto st10;
goto st5;
st6:
if ( ++p == pe )
goto _test_eof6;
case 6:
switch( (*p) ) {
case 42: goto st7;
case 47: goto st9;
}
goto st0;
st7:
if ( ++p == pe )
goto _test_eof7;
case 7:
if ( (*p) == 42 )
goto st8;
goto st7;
st8:
if ( ++p == pe )
goto _test_eof8;
case 8:
switch( (*p) ) {
case 42: goto st8;
case 47: goto st1;
}
goto st7;
st9:
if ( ++p == pe )
goto _test_eof9;
case 9:
if ( (*p) == 10 )
goto st1;
goto st9;
}
_test_eof1: cs = 1; goto _test_eof;
_test_eof10: cs = 10; goto _test_eof;
_test_eof2: cs = 2; goto _test_eof;
_test_eof3: cs = 3; goto _test_eof;
_test_eof4: cs = 4; goto _test_eof;
_test_eof5: cs = 5; goto _test_eof;
_test_eof6: cs = 6; goto _test_eof;
_test_eof7: cs = 7; goto _test_eof;
_test_eof8: cs = 8; goto _test_eof;
_test_eof9: cs = 9; goto _test_eof;
_test_eof: {}
_out: {}
}
#line 1328 "parser.rl"
if (json->stack_handle) {
rvalue_stack_eagerly_release(json->stack_handle);
}
if (cs >= JSON_first_final && p == pe) {
return result;
} else {
raise_parse_error("unexpected token at '%s'", p);
return Qnil;
}
}
static void JSON_mark(void *ptr)
{
JSON_Parser *json = ptr;
rb_gc_mark(json->Vsource);
rb_gc_mark(json->create_id);
rb_gc_mark(json->object_class);
rb_gc_mark(json->array_class);
rb_gc_mark(json->decimal_class);
rb_gc_mark(json->match_string);
rb_gc_mark(json->stack_handle);
const VALUE *name_cache_entries = &json->name_cache.entries[0];
rb_gc_mark_locations(name_cache_entries, name_cache_entries + json->name_cache.length);
}
static void JSON_free(void *ptr)
{
JSON_Parser *json = ptr;
fbuffer_free(&json->fbuffer);
ruby_xfree(json);
}
static size_t JSON_memsize(const void *ptr)
{
const JSON_Parser *json = ptr;
return sizeof(*json) + FBUFFER_CAPA(&json->fbuffer);
}
static const rb_data_type_t JSON_Parser_type = {
"JSON/Parser",
{JSON_mark, JSON_free, JSON_memsize,},
0, 0,
RUBY_TYPED_FREE_IMMEDIATELY,
};
static VALUE cJSON_parser_s_allocate(VALUE klass)
{
JSON_Parser *json;
VALUE obj = TypedData_Make_Struct(klass, JSON_Parser, &JSON_Parser_type, json);
fbuffer_stack_init(&json->fbuffer, 0, NULL, 0);
return obj;
}
/*
* call-seq: source()
*
* Returns a copy of the current _source_ string, that was used to construct
* this Parser.
*/
static VALUE cParser_source(VALUE self)
{
GET_PARSER;
return rb_str_dup(json->Vsource);
}
void Init_parser(void)
{
#ifdef HAVE_RB_EXT_RACTOR_SAFE
rb_ext_ractor_safe(true);
#endif
#undef rb_intern
rb_require("json/common");
mJSON = rb_define_module("JSON");
mExt = rb_define_module_under(mJSON, "Ext");
cParser = rb_define_class_under(mExt, "Parser", rb_cObject);
eNestingError = rb_path2class("JSON::NestingError");
rb_gc_register_mark_object(eNestingError);
rb_define_alloc_func(cParser, cJSON_parser_s_allocate);
rb_define_method(cParser, "initialize", cParser_initialize, -1);
rb_define_method(cParser, "parse", cParser_parse, 0);
rb_define_method(cParser, "source", cParser_source, 0);
rb_define_singleton_method(cParser, "parse", cParser_m_parse, 2);
CNaN = rb_const_get(mJSON, rb_intern("NaN"));
rb_gc_register_mark_object(CNaN);
CInfinity = rb_const_get(mJSON, rb_intern("Infinity"));
rb_gc_register_mark_object(CInfinity);
CMinusInfinity = rb_const_get(mJSON, rb_intern("MinusInfinity"));
rb_gc_register_mark_object(CMinusInfinity);
rb_global_variable(&Encoding_UTF_8);
Encoding_UTF_8 = rb_const_get(rb_path2class("Encoding"), rb_intern("UTF_8"));
sym_max_nesting = ID2SYM(rb_intern("max_nesting"));
sym_allow_nan = ID2SYM(rb_intern("allow_nan"));
sym_allow_trailing_comma = ID2SYM(rb_intern("allow_trailing_comma"));
sym_symbolize_names = ID2SYM(rb_intern("symbolize_names"));
sym_freeze = ID2SYM(rb_intern("freeze"));
sym_create_additions = ID2SYM(rb_intern("create_additions"));
sym_create_id = ID2SYM(rb_intern("create_id"));
sym_object_class = ID2SYM(rb_intern("object_class"));
sym_array_class = ID2SYM(rb_intern("array_class"));
sym_decimal_class = ID2SYM(rb_intern("decimal_class"));
sym_match_string = ID2SYM(rb_intern("match_string"));
i_create_id = rb_intern("create_id");
i_json_creatable_p = rb_intern("json_creatable?");
i_json_create = rb_intern("json_create");
i_chr = rb_intern("chr");
i_match = rb_intern("match");
i_deep_const_get = rb_intern("deep_const_get");
i_aset = rb_intern("[]=");
i_aref = rb_intern("[]");
i_leftshift = rb_intern("<<");
i_new = rb_intern("new");
i_try_convert = rb_intern("try_convert");
i_uminus = rb_intern("-@");
i_encode = rb_intern("encode");
binary_encindex = rb_ascii8bit_encindex();
utf8_encindex = rb_utf8_encindex();
enc_utf8 = rb_utf8_encoding();
}
/*
* Local variables:
* mode: c
* c-file-style: ruby
* indent-tabs-mode: nil
* End:
*/