mirror of
https://github.com/LadybirdBrowser/ladybird
synced 2026-04-25 17:25:08 +02:00
LibJS: Use simdjson for JSON.parse
Replace the custom AK JSON parser with simdjson for parsing JSON in LibJS. This eliminates the intermediate AK::JsonValue object graph, going directly from JSON text to JS::Value. simdjson's on-demand API parses at ~4GB/s and only materializes values as they are accessed, making this both faster and more memory efficient than the previous approach. The AK JSON parser is still used elsewhere (WebDriver protocol, config files, etc.) but LibJS now uses simdjson exclusively for JSON.parse() and JSON.rawJSON().
This commit is contained in:
Notes:
github-actions[bot]
2026-01-12 18:54:37 +00:00
Author: https://github.com/awesomekling Commit: https://github.com/LadybirdBrowser/ladybird/commit/5e0ee26e8be Pull-request: https://github.com/LadybirdBrowser/ladybird/pull/7436 Reviewed-by: https://github.com/tcl3 ✅ Reviewed-by: https://github.com/trflynn89 ✅
@@ -276,6 +276,10 @@ generate_bytecode_def_derived()
|
||||
set(GENERATED_SOURCES Bytecode/Op.cpp)
|
||||
|
||||
ladybird_lib(LibJS js EXPLICIT_SYMBOL_EXPORT)
|
||||
|
||||
find_package(simdjson CONFIG REQUIRED)
|
||||
target_link_libraries(LibJS PRIVATE simdjson::simdjson)
|
||||
|
||||
target_link_libraries(LibJS PRIVATE LibCore LibCrypto LibFileSystem LibRegex LibSyntax LibGC)
|
||||
|
||||
# Link LibUnicode publicly to ensure ICU data (which is in libicudata.a) is available in any process using LibJS.
|
||||
|
||||
@@ -5,10 +5,9 @@
|
||||
*/
|
||||
|
||||
#include <AK/Function.h>
|
||||
#include <AK/JsonArray.h>
|
||||
#include <AK/JsonObject.h>
|
||||
#include <AK/JsonParser.h>
|
||||
#include <AK/GenericLexer.h>
|
||||
#include <AK/StringBuilder.h>
|
||||
#include <AK/StringConversions.h>
|
||||
#include <AK/TypeCasts.h>
|
||||
#include <AK/Utf16View.h>
|
||||
#include <AK/Utf8View.h>
|
||||
@@ -26,6 +25,8 @@
|
||||
#include <LibJS/Runtime/StringObject.h>
|
||||
#include <LibJS/Runtime/ValueInlines.h>
|
||||
|
||||
#include <simdjson.h>
|
||||
|
||||
namespace JS {
|
||||
|
||||
GC_DEFINE_ALLOCATOR(JSONObject);
|
||||
@@ -447,13 +448,327 @@ JS_DEFINE_NATIVE_FUNCTION(JSONObject::parse)
|
||||
return unfiltered;
|
||||
}
|
||||
|
||||
// Unescape a JSON string, properly handling \uXXXX escape sequences including lone surrogates.
|
||||
// simdjson validates UTF-8 strictly and rejects lone surrogates, but JSON allows them.
|
||||
// Returns {} on malformed escape sequences.
|
||||
static Optional<Utf16String> unescape_json_string(StringView raw)
|
||||
{
|
||||
StringBuilder builder(StringBuilder::Mode::UTF16, raw.length());
|
||||
|
||||
GenericLexer lexer { raw };
|
||||
|
||||
auto consume_hex4 = [&]() -> Optional<u16> {
|
||||
if (lexer.tell_remaining() < 4)
|
||||
return {};
|
||||
u16 value = 0;
|
||||
for (int i = 0; i < 4; ++i) {
|
||||
auto ch = lexer.consume();
|
||||
value <<= 4;
|
||||
if (ch >= '0' && ch <= '9')
|
||||
value |= ch - '0';
|
||||
else if (ch >= 'a' && ch <= 'f')
|
||||
value |= ch - 'a' + 10;
|
||||
else if (ch >= 'A' && ch <= 'F')
|
||||
value |= ch - 'A' + 10;
|
||||
else
|
||||
return {};
|
||||
}
|
||||
return value;
|
||||
};
|
||||
|
||||
while (!lexer.is_eof()) {
|
||||
if (lexer.consume_specific('\\')) {
|
||||
if (lexer.is_eof())
|
||||
return {};
|
||||
auto escaped = lexer.consume();
|
||||
switch (escaped) {
|
||||
case '"':
|
||||
builder.append_code_unit('"');
|
||||
break;
|
||||
case '\\':
|
||||
builder.append_code_unit('\\');
|
||||
break;
|
||||
case '/':
|
||||
builder.append_code_unit('/');
|
||||
break;
|
||||
case 'b':
|
||||
builder.append_code_unit('\b');
|
||||
break;
|
||||
case 'f':
|
||||
builder.append_code_unit('\f');
|
||||
break;
|
||||
case 'n':
|
||||
builder.append_code_unit('\n');
|
||||
break;
|
||||
case 'r':
|
||||
builder.append_code_unit('\r');
|
||||
break;
|
||||
case 't':
|
||||
builder.append_code_unit('\t');
|
||||
break;
|
||||
case 'u': {
|
||||
auto code_unit = consume_hex4();
|
||||
if (!code_unit.has_value())
|
||||
return {};
|
||||
builder.append_code_unit(*code_unit);
|
||||
break;
|
||||
}
|
||||
default:
|
||||
return {};
|
||||
}
|
||||
} else {
|
||||
// Non-escaped character - copy UTF-8 code point to UTF-16
|
||||
auto ch = lexer.consume();
|
||||
if ((ch & 0x80) == 0) {
|
||||
// ASCII
|
||||
builder.append_code_unit(ch);
|
||||
} else if ((ch & 0xE0) == 0xC0) {
|
||||
// 2-byte UTF-8
|
||||
if (lexer.is_eof())
|
||||
return {};
|
||||
auto ch2 = lexer.consume();
|
||||
u32 code_point = ((ch & 0x1F) << 6) | (ch2 & 0x3F);
|
||||
builder.append_code_unit(code_point);
|
||||
} else if ((ch & 0xF0) == 0xE0) {
|
||||
// 3-byte UTF-8
|
||||
if (lexer.tell_remaining() < 2)
|
||||
return {};
|
||||
auto ch2 = lexer.consume();
|
||||
auto ch3 = lexer.consume();
|
||||
u32 code_point = ((ch & 0x0F) << 12) | ((ch2 & 0x3F) << 6) | (ch3 & 0x3F);
|
||||
builder.append_code_unit(code_point);
|
||||
} else if ((ch & 0xF8) == 0xF0) {
|
||||
// 4-byte UTF-8 (needs surrogate pair)
|
||||
if (lexer.tell_remaining() < 3)
|
||||
return {};
|
||||
auto ch2 = lexer.consume();
|
||||
auto ch3 = lexer.consume();
|
||||
auto ch4 = lexer.consume();
|
||||
u32 code_point = ((ch & 0x07) << 18) | ((ch2 & 0x3F) << 12) | ((ch3 & 0x3F) << 6) | (ch4 & 0x3F);
|
||||
builder.append_code_point(code_point);
|
||||
} else {
|
||||
return {};
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return builder.to_utf16_string();
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
static ALWAYS_INLINE ThrowCompletionOr<void> ensure_simdjson_fully_parsed(VM& vm, T& value)
|
||||
{
|
||||
if constexpr (IsSame<T, simdjson::ondemand::document>) {
|
||||
if (!value.at_end())
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
}
|
||||
return {};
|
||||
}
|
||||
|
||||
static ThrowCompletionOr<Value> parse_simdjson_value(VM&, simdjson::ondemand::value);
|
||||
|
||||
template<typename T>
|
||||
static ThrowCompletionOr<Value> parse_simdjson_number(VM& vm, T& value, StringView raw_sv)
|
||||
{
|
||||
// Validate JSON number format (simdjson is more lenient than spec)
|
||||
// - No leading zeros (except "0" or "0.xxx")
|
||||
// - No trailing decimal point (e.g., "1." is invalid)
|
||||
size_t i = 0;
|
||||
if (i < raw_sv.length() && raw_sv[i] == '-')
|
||||
++i;
|
||||
if (i < raw_sv.length() && raw_sv[i] == '0' && i + 1 < raw_sv.length() && is_ascii_digit(raw_sv[i + 1]))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed); // Leading zero
|
||||
while (i < raw_sv.length() && is_ascii_digit(raw_sv[i]))
|
||||
++i;
|
||||
if (i < raw_sv.length() && raw_sv[i] == '.') {
|
||||
++i;
|
||||
if (i >= raw_sv.length() || !is_ascii_digit(raw_sv[i]))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed); // Trailing decimal
|
||||
}
|
||||
|
||||
double double_value;
|
||||
auto error = value.get_double().get(double_value);
|
||||
if (!error) {
|
||||
TRY(ensure_simdjson_fully_parsed(vm, value));
|
||||
return Value(double_value);
|
||||
}
|
||||
|
||||
// Handle overflow to infinity (e.g., 1e309)
|
||||
// simdjson returns NUMBER_ERROR for numbers that overflow double
|
||||
// Use parse_first_number as fallback - it handles overflow correctly
|
||||
if (error == simdjson::NUMBER_ERROR) {
|
||||
auto result = parse_first_number<double>(raw_sv, TrimWhitespace::No);
|
||||
if (result.has_value() && result->characters_parsed == raw_sv.length())
|
||||
return Value(result->value);
|
||||
}
|
||||
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
static ThrowCompletionOr<Value> parse_simdjson_string(VM& vm, T& value)
|
||||
{
|
||||
// Use get_raw_json_string() to get the raw JSON string content (without quotes, with escapes),
|
||||
// then unescape ourselves to properly handle lone surrogates like \uD800 which simdjson rejects.
|
||||
simdjson::ondemand::raw_json_string raw_string;
|
||||
if (value.get_raw_json_string().get(raw_string))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
char const* raw = raw_string.raw();
|
||||
// Find the length by looking for the closing quote (simdjson validated the structure)
|
||||
size_t length = 0;
|
||||
while (raw[length] != '"') {
|
||||
if (raw[length] == '\\')
|
||||
++length; // Skip escaped character
|
||||
++length;
|
||||
}
|
||||
auto unescaped = unescape_json_string({ raw, length });
|
||||
if (!unescaped.has_value())
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
return PrimitiveString::create(vm, unescaped.release_value());
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
static ThrowCompletionOr<Value> parse_simdjson_array(VM& vm, T& value)
|
||||
{
|
||||
auto& realm = *vm.current_realm();
|
||||
|
||||
simdjson::ondemand::array simdjson_array;
|
||||
if (value.get_array().get(simdjson_array))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
|
||||
auto array = MUST(Array::create(realm, 0));
|
||||
size_t index = 0;
|
||||
|
||||
for (auto element : simdjson_array) {
|
||||
simdjson::ondemand::value element_value;
|
||||
if (element.get(element_value))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
auto parsed = TRY(parse_simdjson_value(vm, element_value));
|
||||
array->define_direct_property(index++, parsed, default_attributes);
|
||||
}
|
||||
|
||||
TRY(ensure_simdjson_fully_parsed(vm, value));
|
||||
return array;
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
static ThrowCompletionOr<Value> parse_simdjson_object(VM& vm, T& value)
|
||||
{
|
||||
auto& realm = *vm.current_realm();
|
||||
|
||||
simdjson::ondemand::object simdjson_object;
|
||||
if (value.get_object().get(simdjson_object))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
|
||||
auto object = Object::create(realm, realm.intrinsics().object_prototype());
|
||||
|
||||
for (auto field : simdjson_object) {
|
||||
// Use escaped_key() to get the raw JSON key (with escapes), then unescape ourselves
|
||||
std::string_view raw_key;
|
||||
if (field.escaped_key().get(raw_key))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
auto unescaped_key = unescape_json_string({ raw_key.data(), raw_key.size() });
|
||||
if (!unescaped_key.has_value())
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
simdjson::ondemand::value field_value;
|
||||
if (field.value().get(field_value))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
auto parsed = TRY(parse_simdjson_value(vm, field_value));
|
||||
object->define_direct_property(unescaped_key.release_value(), parsed, default_attributes);
|
||||
}
|
||||
|
||||
TRY(ensure_simdjson_fully_parsed(vm, value));
|
||||
return object;
|
||||
}
|
||||
|
||||
static ThrowCompletionOr<Value> parse_simdjson_value(VM& vm, simdjson::ondemand::value value)
|
||||
{
|
||||
simdjson::ondemand::json_type type;
|
||||
if (value.type().get(type))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
|
||||
switch (type) {
|
||||
case simdjson::ondemand::json_type::null:
|
||||
return js_null();
|
||||
case simdjson::ondemand::json_type::boolean: {
|
||||
bool boolean_value;
|
||||
if (value.get_bool().get(boolean_value))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
return Value(boolean_value);
|
||||
}
|
||||
case simdjson::ondemand::json_type::number: {
|
||||
auto raw = value.raw_json_token();
|
||||
StringView raw_sv { raw.data(), raw.size() };
|
||||
return parse_simdjson_number(vm, value, raw_sv);
|
||||
}
|
||||
case simdjson::ondemand::json_type::string:
|
||||
return parse_simdjson_string(vm, value);
|
||||
case simdjson::ondemand::json_type::array:
|
||||
return parse_simdjson_array(vm, value);
|
||||
case simdjson::ondemand::json_type::object:
|
||||
return parse_simdjson_object(vm, value);
|
||||
}
|
||||
VERIFY_NOT_REACHED();
|
||||
}
|
||||
|
||||
static ThrowCompletionOr<Value> parse_simdjson_document(VM& vm, simdjson::ondemand::document& document)
|
||||
{
|
||||
simdjson::ondemand::json_type type;
|
||||
if (document.type().get(type))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
|
||||
switch (type) {
|
||||
case simdjson::ondemand::json_type::null: {
|
||||
if (document.is_null().error())
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
if (!document.at_end())
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
return js_null();
|
||||
}
|
||||
case simdjson::ondemand::json_type::boolean: {
|
||||
bool boolean_value;
|
||||
if (document.get_bool().get(boolean_value))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
if (!document.at_end())
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
return Value(boolean_value);
|
||||
}
|
||||
case simdjson::ondemand::json_type::number: {
|
||||
// Get raw token first in case get_double fails (e.g., overflow)
|
||||
std::string_view raw;
|
||||
if (document.raw_json_token().get(raw))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
StringView raw_sv { raw.data(), raw.size() };
|
||||
auto trimmed = raw_sv.trim_whitespace();
|
||||
return parse_simdjson_number(vm, document, trimmed);
|
||||
}
|
||||
case simdjson::ondemand::json_type::string:
|
||||
return parse_simdjson_string(vm, document);
|
||||
case simdjson::ondemand::json_type::array:
|
||||
return parse_simdjson_array(vm, document);
|
||||
case simdjson::ondemand::json_type::object:
|
||||
return parse_simdjson_object(vm, document);
|
||||
}
|
||||
VERIFY_NOT_REACHED();
|
||||
}
|
||||
|
||||
// 25.5.1.1 ParseJSON ( text ), https://tc39.es/ecma262/#sec-ParseJSON
|
||||
ThrowCompletionOr<Value> JSONObject::parse_json(VM& vm, StringView text)
|
||||
{
|
||||
auto json = JsonValue::from_string(text);
|
||||
|
||||
// 1. If StringToCodePoints(text) is not a valid JSON text as specified in ECMA-404, throw a SyntaxError exception.
|
||||
if (json.is_error())
|
||||
// NB: Per ECMA-404, the BOM is not valid JSON whitespace. simdjson silently skips it, so we must reject it explicitly.
|
||||
if (text.length() >= 3
|
||||
&& static_cast<u8>(text[0]) == 0xEF
|
||||
&& static_cast<u8>(text[1]) == 0xBB
|
||||
&& static_cast<u8>(text[2]) == 0xBF) {
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
}
|
||||
|
||||
simdjson::ondemand::parser parser;
|
||||
simdjson::padded_string padded(text.characters_without_null_termination(), text.length());
|
||||
|
||||
simdjson::ondemand::document document;
|
||||
if (parser.iterate(padded).get(document))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
|
||||
// 2. Let scriptString be the string-concatenation of "(", text, and ");".
|
||||
@@ -461,7 +776,7 @@ ThrowCompletionOr<Value> JSONObject::parse_json(VM& vm, StringView text)
|
||||
// 4. NOTE: The early error rules defined in 13.2.5.1 have special handling for the above invocation of ParseText.
|
||||
// 5. Assert: script is a Parse Node.
|
||||
// 6. Let result be ! Evaluation of script.
|
||||
auto result = JSONObject::parse_json_value(vm, json.value());
|
||||
auto result = TRY(parse_simdjson_document(vm, document));
|
||||
|
||||
// 7. NOTE: The PropertyDefinitionEvaluation semantics defined in 13.2.5.5 have special handling for the above evaluation.
|
||||
// 8. Assert: result is either a String, a Number, a Boolean, an Object that is defined by either an ArrayLiteral or an ObjectLiteral, or null.
|
||||
@@ -470,44 +785,6 @@ ThrowCompletionOr<Value> JSONObject::parse_json(VM& vm, StringView text)
|
||||
return result;
|
||||
}
|
||||
|
||||
Value JSONObject::parse_json_value(VM& vm, JsonValue const& value)
|
||||
{
|
||||
if (value.is_object())
|
||||
return Value(parse_json_object(vm, value.as_object()));
|
||||
if (value.is_array())
|
||||
return Value(parse_json_array(vm, value.as_array()));
|
||||
if (value.is_null())
|
||||
return js_null();
|
||||
if (auto double_value = value.get_double_with_precision_loss(); double_value.has_value())
|
||||
return Value(double_value.value());
|
||||
if (value.is_string())
|
||||
return PrimitiveString::create(vm, value.as_string());
|
||||
if (value.is_bool())
|
||||
return Value(value.as_bool());
|
||||
VERIFY_NOT_REACHED();
|
||||
}
|
||||
|
||||
Object* JSONObject::parse_json_object(VM& vm, JsonObject const& json_object)
|
||||
{
|
||||
auto& realm = *vm.current_realm();
|
||||
auto object = Object::create(realm, realm.intrinsics().object_prototype());
|
||||
json_object.for_each_member([&](auto& key, auto& value) {
|
||||
object->define_direct_property(Utf16String::from_utf8(key), parse_json_value(vm, value), default_attributes);
|
||||
});
|
||||
return object;
|
||||
}
|
||||
|
||||
Array* JSONObject::parse_json_array(VM& vm, JsonArray const& json_array)
|
||||
{
|
||||
auto& realm = *vm.current_realm();
|
||||
auto array = MUST(Array::create(realm, 0));
|
||||
size_t index = 0;
|
||||
json_array.for_each([&](auto& value) {
|
||||
array->define_direct_property(index++, parse_json_value(vm, value), default_attributes);
|
||||
});
|
||||
return array;
|
||||
}
|
||||
|
||||
// 25.5.1.1 InternalizeJSONProperty ( holder, name, reviver ), https://tc39.es/ecma262/#sec-internalizejsonproperty
|
||||
ThrowCompletionOr<Value> JSONObject::internalize_json_property(VM& vm, Object* holder, PropertyKey const& name, FunctionObject& reviver)
|
||||
{
|
||||
@@ -564,13 +841,41 @@ JS_DEFINE_NATIVE_FUNCTION(JSONObject::raw_json)
|
||||
// 3. Parse StringToCodePoints(jsonString) as a JSON text as specified in ECMA-404. Throw a SyntaxError exception
|
||||
// if it is not a valid JSON text as defined in that specification, or if its outermost value is an object or
|
||||
// array as defined in that specification.
|
||||
auto json = JsonValue::from_string(json_string);
|
||||
if (json.is_error())
|
||||
simdjson::ondemand::parser parser;
|
||||
simdjson::padded_string padded(json_string.bytes_as_string_view().characters_without_null_termination(), json_string.bytes_as_string_view().length());
|
||||
|
||||
simdjson::ondemand::document doc;
|
||||
if (parser.iterate(padded).get(doc))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
|
||||
if (json.value().is_object() || json.value().is_array())
|
||||
simdjson::ondemand::json_type type;
|
||||
if (doc.type().get(type))
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
|
||||
if (type == simdjson::ondemand::json_type::object || type == simdjson::ondemand::json_type::array)
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonRawJSONNonPrimitive);
|
||||
|
||||
// Consume the value to advance past it, then check for trailing content
|
||||
switch (type) {
|
||||
case simdjson::ondemand::json_type::null:
|
||||
(void)doc.is_null();
|
||||
break;
|
||||
case simdjson::ondemand::json_type::boolean:
|
||||
(void)doc.get_bool();
|
||||
break;
|
||||
case simdjson::ondemand::json_type::number:
|
||||
(void)doc.get_double();
|
||||
break;
|
||||
case simdjson::ondemand::json_type::string:
|
||||
(void)doc.get_string();
|
||||
break;
|
||||
default:
|
||||
VERIFY_NOT_REACHED();
|
||||
}
|
||||
|
||||
if (!doc.at_end())
|
||||
return vm.throw_completion<SyntaxError>(ErrorType::JsonMalformed);
|
||||
|
||||
// 4. Let internalSlotsList be « [[IsRawJSON]] ».
|
||||
// 5. Let obj be OrdinaryObjectCreate(null, internalSlotsList).
|
||||
auto object = RawJSONObject::create(realm, nullptr);
|
||||
|
||||
@@ -24,7 +24,6 @@ public:
|
||||
static ThrowCompletionOr<Optional<String>> stringify_impl(VM&, Value value, Value replacer, Value space);
|
||||
|
||||
static ThrowCompletionOr<Value> parse_json(VM&, StringView text);
|
||||
static Value parse_json_value(VM&, JsonValue const&);
|
||||
|
||||
private:
|
||||
explicit JSONObject(Realm&);
|
||||
@@ -44,8 +43,6 @@ private:
|
||||
static String quote_json_string(Utf16View const&);
|
||||
|
||||
// Parse helpers
|
||||
static Object* parse_json_object(VM&, JsonObject const&);
|
||||
static Array* parse_json_array(VM&, JsonArray const&);
|
||||
static ThrowCompletionOr<Value> internalize_json_property(VM&, Object* holder, PropertyKey const& name, FunctionObject& reviver);
|
||||
|
||||
JS_DECLARE_NATIVE_FUNCTION(stringify);
|
||||
|
||||
@@ -74,3 +74,73 @@ test("does not truncate large integers", () => {
|
||||
expect(JSON.parse("18446744073709551616")).toEqual(18446744073709551616);
|
||||
expect(JSON.parse("18446744073709551617")).toEqual(18446744073709551617);
|
||||
});
|
||||
|
||||
test("number overflow to infinity", () => {
|
||||
expect(JSON.parse("1e309")).toBe(Infinity);
|
||||
expect(JSON.parse("-1e309")).toBe(-Infinity);
|
||||
expect(JSON.parse("1e-400")).toBe(0);
|
||||
});
|
||||
|
||||
test("rejects invalid number formats", () => {
|
||||
// Leading zeros not allowed
|
||||
expect(() => JSON.parse("01")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("-01")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("00")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("007")).toThrow(SyntaxError);
|
||||
|
||||
// Trailing decimal point not allowed
|
||||
expect(() => JSON.parse("1.")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("0.")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("-1.")).toThrow(SyntaxError);
|
||||
|
||||
// Other invalid formats
|
||||
expect(() => JSON.parse("+1")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse(".1")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("1e")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("1e+")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("1e-")).toThrow(SyntaxError);
|
||||
});
|
||||
|
||||
test("rejects trailing content", () => {
|
||||
expect(() => JSON.parse("123 garbage")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("null garbage")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("true garbage")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse('"string" garbage')).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("[] garbage")).toThrow(SyntaxError);
|
||||
expect(() => JSON.parse("{} garbage")).toThrow(SyntaxError);
|
||||
});
|
||||
|
||||
test("string escape sequences", () => {
|
||||
expect(JSON.parse('"\\""')).toBe('"');
|
||||
expect(JSON.parse('"\\\\"')).toBe("\\");
|
||||
expect(JSON.parse('"\\/"')).toBe("/");
|
||||
expect(JSON.parse('"\\b"')).toBe("\b");
|
||||
expect(JSON.parse('"\\f"')).toBe("\f");
|
||||
expect(JSON.parse('"\\n"')).toBe("\n");
|
||||
expect(JSON.parse('"\\r"')).toBe("\r");
|
||||
expect(JSON.parse('"\\t"')).toBe("\t");
|
||||
expect(JSON.parse('"\\u0041"')).toBe("A");
|
||||
expect(JSON.parse('"\\u0000"')).toBe("\0");
|
||||
});
|
||||
|
||||
test("unicode and surrogate pairs", () => {
|
||||
expect(JSON.parse('"café"')).toBe("café");
|
||||
expect(JSON.parse('"日本語"')).toBe("日本語");
|
||||
expect(JSON.parse('"\\uD83D\\uDE00"')).toBe("😀");
|
||||
expect(JSON.parse('"\\u4e2d\\u6587"')).toBe("中文");
|
||||
|
||||
// Lone surrogates (valid JSON)
|
||||
expect(JSON.parse('"\\uD800"')).toBe("\uD800");
|
||||
expect(JSON.parse('"\\uDFFF"')).toBe("\uDFFF");
|
||||
});
|
||||
|
||||
test("whitespace handling", () => {
|
||||
expect(JSON.parse(" null")).toBe(null);
|
||||
expect(JSON.parse("null ")).toBe(null);
|
||||
expect(JSON.parse(" null ")).toBe(null);
|
||||
expect(JSON.parse("\t123")).toBe(123);
|
||||
expect(JSON.parse("123\n")).toBe(123);
|
||||
expect(JSON.parse("\r\n123\r\n")).toBe(123);
|
||||
expect(JSON.parse(" { } ")).toEqual({});
|
||||
expect(JSON.parse(" [ ] ")).toEqual([]);
|
||||
});
|
||||
|
||||
@@ -10,9 +10,11 @@
|
||||
#include <AK/JsonObject.h>
|
||||
#include <AK/JsonValue.h>
|
||||
#include <AK/NumericLimits.h>
|
||||
#include <AK/Utf16String.h>
|
||||
#include <AK/Variant.h>
|
||||
#include <LibJS/Runtime/Array.h>
|
||||
#include <LibJS/Runtime/JSONObject.h>
|
||||
#include <LibJS/Runtime/Object.h>
|
||||
#include <LibJS/Runtime/PrimitiveString.h>
|
||||
#include <LibWeb/DOM/DOMTokenList.h>
|
||||
#include <LibWeb/DOM/Document.h>
|
||||
#include <LibWeb/DOM/HTMLCollection.h>
|
||||
@@ -67,6 +69,47 @@ static bool is_collection(JS::Object const& value)
|
||||
|| is<DOM::NodeList>(value));
|
||||
}
|
||||
|
||||
// Helper to convert AK::JsonValue to JS::Value (for WebDriver protocol)
|
||||
static JS::Value json_value_to_js_value(JS::VM& vm, JsonValue const& value);
|
||||
|
||||
static JS::Object* json_object_to_js_object(JS::VM& vm, JsonObject const& json_object)
|
||||
{
|
||||
auto& realm = *vm.current_realm();
|
||||
auto object = JS::Object::create(realm, realm.intrinsics().object_prototype());
|
||||
json_object.for_each_member([&](auto& key, auto& value) {
|
||||
object->define_direct_property(Utf16String::from_utf8(key), json_value_to_js_value(vm, value), JS::default_attributes);
|
||||
});
|
||||
return object;
|
||||
}
|
||||
|
||||
static JS::Array* json_array_to_js_array(JS::VM& vm, JsonArray const& json_array)
|
||||
{
|
||||
auto& realm = *vm.current_realm();
|
||||
auto array = MUST(JS::Array::create(realm, 0));
|
||||
size_t index = 0;
|
||||
json_array.for_each([&](auto& value) {
|
||||
array->define_direct_property(index++, json_value_to_js_value(vm, value), JS::default_attributes);
|
||||
});
|
||||
return array;
|
||||
}
|
||||
|
||||
static JS::Value json_value_to_js_value(JS::VM& vm, JsonValue const& value)
|
||||
{
|
||||
if (value.is_object())
|
||||
return JS::Value(json_object_to_js_object(vm, value.as_object()));
|
||||
if (value.is_array())
|
||||
return JS::Value(json_array_to_js_array(vm, value.as_array()));
|
||||
if (value.is_null())
|
||||
return JS::js_null();
|
||||
if (auto double_value = value.get_double_with_precision_loss(); double_value.has_value())
|
||||
return JS::Value(double_value.value());
|
||||
if (value.is_string())
|
||||
return JS::PrimitiveString::create(vm, value.as_string());
|
||||
if (value.is_bool())
|
||||
return JS::Value(value.as_bool());
|
||||
VERIFY_NOT_REACHED();
|
||||
}
|
||||
|
||||
// https://w3c.github.io/webdriver/#dfn-clone-an-object
|
||||
template<typename ResultType, typename CloneAlgorithm>
|
||||
static ErrorOr<ResultType, WebDriver::Error> clone_an_object(HTML::BrowsingContext const& browsing_context, JS::Object const& value, SeenMap& seen, CloneAlgorithm const& clone_algorithm)
|
||||
@@ -334,7 +377,7 @@ ErrorOr<JS::Value, WebDriver::Error> json_deserialize(HTML::BrowsingContext cons
|
||||
auto& vm = browsing_context.vm();
|
||||
|
||||
SeenMap seen;
|
||||
return internal_json_deserialize(browsing_context, JS::JSONObject::parse_json_value(vm, value), seen);
|
||||
return internal_json_deserialize(browsing_context, json_value_to_js_value(vm, value), seen);
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@@ -29,6 +29,25 @@
|
||||
"/share/man"
|
||||
],
|
||||
"modules": [
|
||||
{
|
||||
"name": "simdjson",
|
||||
"buildsystem": "cmake-ninja",
|
||||
"sources": [
|
||||
{
|
||||
"type": "git",
|
||||
"url": "https://github.com/simdjson/simdjson.git",
|
||||
"tag": "v3.12.2"
|
||||
}
|
||||
],
|
||||
"config-opts": [
|
||||
"-DCMAKE_BUILD_TYPE=Release",
|
||||
"-DCMAKE_PREFIX_PATH=/app",
|
||||
"-DCMAKE_INSTALL_LIBDIR=lib",
|
||||
"-DCMAKE_POSITION_INDEPENDENT_CODE=ON",
|
||||
"-DSIMDJSON_BUILD_STATIC_LIB=OFF",
|
||||
"-DSIMDJSON_DEVELOPER_MODE=OFF"
|
||||
]
|
||||
},
|
||||
{
|
||||
"name": "simdutf",
|
||||
"buildsystem": "cmake-ninja",
|
||||
|
||||
@@ -7,7 +7,6 @@
|
||||
* SPDX-License-Identifier: BSD-2-Clause
|
||||
*/
|
||||
|
||||
#include <AK/JsonValue.h>
|
||||
#include <AK/NeverDestroyed.h>
|
||||
#include <AK/Platform.h>
|
||||
#include <AK/StringBuilder.h>
|
||||
@@ -290,11 +289,7 @@ static JS::ThrowCompletionOr<JS::Value> load_json_impl(JS::VM& vm)
|
||||
if (file_contents_or_error.is_error())
|
||||
return vm.throw_completion<JS::Error>(TRY_OR_THROW_OOM(vm, String::formatted("Failed to read '{}': {}", filename, file_contents_or_error.error())));
|
||||
|
||||
auto json = JsonValue::from_string(file_contents_or_error.value());
|
||||
if (json.is_error())
|
||||
return vm.throw_completion<JS::SyntaxError>(JS::ErrorType::JsonMalformed);
|
||||
|
||||
return JS::JSONObject::parse_json_value(vm, json.value());
|
||||
return JS::JSONObject::parse_json(vm, file_contents_or_error.value());
|
||||
}
|
||||
|
||||
void ReplObject::initialize(JS::Realm& realm)
|
||||
|
||||
@@ -163,6 +163,7 @@
|
||||
"name": "sdl3",
|
||||
"default-features": false
|
||||
},
|
||||
"simdjson",
|
||||
"simdutf",
|
||||
{
|
||||
"name": "skia",
|
||||
@@ -307,6 +308,10 @@
|
||||
"name": "sdl3",
|
||||
"version": "3.2.22#0"
|
||||
},
|
||||
{
|
||||
"name": "simdjson",
|
||||
"version": "3.13.0#0"
|
||||
},
|
||||
{
|
||||
"name": "simdutf",
|
||||
"version": "7.4.0#0"
|
||||
|
||||
Reference in New Issue
Block a user