Repo created

This commit is contained in:
Fr4nz D13trich 2025-11-22 14:04:28 +01:00
parent 81b91f4139
commit f8c34fa5ee
22732 changed files with 4815320 additions and 2 deletions

View file

@ -0,0 +1 @@
file://base/SECURITY_OWNERS

View file

@ -0,0 +1,42 @@
// Copyright 2019 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#ifndef BASE_JSON_JSON_COMMON_H_
#define BASE_JSON_JSON_COMMON_H_
#include <stddef.h>
#include "base/logging.h"
#include "base/macros.h"
namespace base {
namespace internal {
// Chosen to support 99.9% of documents found in the wild late 2016.
// http://crbug.com/673263
const size_t kAbsoluteMaxDepth = 200;
// Simple class that checks for maximum recursion/stack overflow.
class StackMarker {
public:
StackMarker(size_t max_depth, size_t* depth)
: max_depth_(max_depth), depth_(depth) {
++(*depth_);
DCHECK_LE(*depth_, max_depth_);
}
~StackMarker() { --(*depth_); }
bool IsTooDeep() const { return *depth_ >= max_depth_; }
private:
const size_t max_depth_;
size_t* const depth_;
DISALLOW_COPY_AND_ASSIGN(StackMarker);
};
} // namespace internal
} // namespace base
#endif // BASE_JSON_JSON_COMMON_H_

View file

@ -0,0 +1,58 @@
// Copyright 2016 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
// A fuzzer that checks correctness of json parser/writer.
// The fuzzer input is passed through parsing twice,
// so that presumably valid json is parsed/written again.
#include <stddef.h>
#include <stdint.h>
#include <string>
#include "base/json/json_reader.h"
#include "base/json/json_writer.h"
#include "base/json/string_escape.h"
#include "base/logging.h"
#include "base/values.h"
// Entry point for libFuzzer.
// We will use the last byte of data as parsing options.
// The rest will be used as text input to the parser.
extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
if (size < 2)
return 0;
// Create a copy of input buffer, as otherwise we don't catch
// overflow that touches the last byte (which is used in options).
std::unique_ptr<char[]> input(new char[size - 1]);
memcpy(input.get(), data, size - 1);
base::StringPiece input_string(input.get(), size - 1);
const int options = data[size - 1];
base::JSONReader::ValueWithError result =
base::JSONReader::ReadAndReturnValueWithError(input_string, options);
if (!result.value)
return 0;
std::string parsed_output;
bool b = base::JSONWriter::Write(*result.value, &parsed_output);
LOG_ASSERT(b);
base::JSONReader::ValueWithError double_result =
base::JSONReader::ReadAndReturnValueWithError(parsed_output, options);
LOG_ASSERT(double_result.value);
std::string double_parsed_output;
bool b2 =
base::JSONWriter::Write(*double_result.value, &double_parsed_output);
LOG_ASSERT(b2);
LOG_ASSERT(parsed_output == double_parsed_output)
<< "Parser/Writer mismatch."
<< "\nInput=" << base::GetQuotedJSONString(parsed_output)
<< "\nOutput=" << base::GetQuotedJSONString(double_parsed_output);
return 0;
}

View file

@ -0,0 +1,113 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "base/json/json_file_value_serializer.h"
#include "base/files/file_util.h"
#include "base/json/json_string_value_serializer.h"
#include "base/logging.h"
#include "build/build_config.h"
using base::FilePath;
const char JSONFileValueDeserializer::kAccessDenied[] = "Access denied.";
const char JSONFileValueDeserializer::kCannotReadFile[] = "Can't read file.";
const char JSONFileValueDeserializer::kFileLocked[] = "File locked.";
const char JSONFileValueDeserializer::kNoSuchFile[] = "File doesn't exist.";
JSONFileValueSerializer::JSONFileValueSerializer(
const base::FilePath& json_file_path)
: json_file_path_(json_file_path) {
}
JSONFileValueSerializer::~JSONFileValueSerializer() = default;
bool JSONFileValueSerializer::Serialize(const base::Value& root) {
return SerializeInternal(root, false);
}
bool JSONFileValueSerializer::SerializeAndOmitBinaryValues(
const base::Value& root) {
return SerializeInternal(root, true);
}
bool JSONFileValueSerializer::SerializeInternal(const base::Value& root,
bool omit_binary_values) {
std::string json_string;
JSONStringValueSerializer serializer(&json_string);
serializer.set_pretty_print(true);
bool result = omit_binary_values ?
serializer.SerializeAndOmitBinaryValues(root) :
serializer.Serialize(root);
if (!result)
return false;
int data_size = static_cast<int>(json_string.size());
if (base::WriteFile(json_file_path_, json_string.data(), data_size) !=
data_size)
return false;
return true;
}
JSONFileValueDeserializer::JSONFileValueDeserializer(
const base::FilePath& json_file_path,
int options)
: json_file_path_(json_file_path), options_(options), last_read_size_(0U) {}
JSONFileValueDeserializer::~JSONFileValueDeserializer() = default;
int JSONFileValueDeserializer::ReadFileToString(std::string* json_string) {
DCHECK(json_string);
if (!base::ReadFileToString(json_file_path_, json_string)) {
#if defined(OS_WIN)
int error = ::GetLastError();
if (error == ERROR_SHARING_VIOLATION || error == ERROR_LOCK_VIOLATION) {
return JSON_FILE_LOCKED;
} else if (error == ERROR_ACCESS_DENIED) {
return JSON_ACCESS_DENIED;
}
#endif
return base::PathExists(json_file_path_) ? JSON_CANNOT_READ_FILE
: JSON_NO_SUCH_FILE;
}
last_read_size_ = json_string->size();
return JSON_NO_ERROR;
}
const char* JSONFileValueDeserializer::GetErrorMessageForCode(int error_code) {
switch (error_code) {
case JSON_NO_ERROR:
return "";
case JSON_ACCESS_DENIED:
return kAccessDenied;
case JSON_CANNOT_READ_FILE:
return kCannotReadFile;
case JSON_FILE_LOCKED:
return kFileLocked;
case JSON_NO_SUCH_FILE:
return kNoSuchFile;
default:
NOTREACHED();
return "";
}
}
std::unique_ptr<base::Value> JSONFileValueDeserializer::Deserialize(
int* error_code,
std::string* error_str) {
std::string json_string;
int error = ReadFileToString(&json_string);
if (error != JSON_NO_ERROR) {
if (error_code)
*error_code = error;
if (error_str)
*error_str = GetErrorMessageForCode(error);
return nullptr;
}
JSONStringValueDeserializer deserializer(json_string, options_);
return deserializer.Deserialize(error_code, error_str);
}

View file

@ -0,0 +1,103 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#ifndef BASE_JSON_JSON_FILE_VALUE_SERIALIZER_H_
#define BASE_JSON_JSON_FILE_VALUE_SERIALIZER_H_
#include <stddef.h>
#include <string>
#include "base/base_export.h"
#include "base/files/file_path.h"
#include "base/macros.h"
#include "base/values.h"
class BASE_EXPORT JSONFileValueSerializer : public base::ValueSerializer {
public:
// |json_file_path_| is the path of a file that will be destination of the
// serialization. The serializer will attempt to create the file at the
// specified location.
explicit JSONFileValueSerializer(const base::FilePath& json_file_path);
~JSONFileValueSerializer() override;
// DO NOT USE except in unit tests to verify the file was written properly.
// We should never serialize directly to a file since this will block the
// thread. Instead, serialize to a string and write to the file you want on
// the file thread.
//
// Attempt to serialize the data structure represented by Value into
// JSON. If the return value is true, the result will have been written
// into the file whose name was passed into the constructor.
bool Serialize(const base::Value& root) override;
// Equivalent to Serialize(root) except binary values are omitted from the
// output.
bool SerializeAndOmitBinaryValues(const base::Value& root);
private:
bool SerializeInternal(const base::Value& root, bool omit_binary_values);
const base::FilePath json_file_path_;
DISALLOW_IMPLICIT_CONSTRUCTORS(JSONFileValueSerializer);
};
class BASE_EXPORT JSONFileValueDeserializer : public base::ValueDeserializer {
public:
// |json_file_path_| is the path of a file that will be source of the
// deserialization. |options| is a bitmask of JSONParserOptions.
explicit JSONFileValueDeserializer(const base::FilePath& json_file_path,
int options = 0);
~JSONFileValueDeserializer() override;
// Attempt to deserialize the data structure encoded in the file passed
// in to the constructor into a structure of Value objects. If the return
// value is NULL, and if |error_code| is non-null, |error_code| will
// contain an integer error code (either JsonFileError or JsonParseError).
// If |error_message| is non-null, it will be filled in with a formatted
// error message including the location of the error if appropriate.
// The caller takes ownership of the returned value.
std::unique_ptr<base::Value> Deserialize(int* error_code,
std::string* error_message) override;
// This enum is designed to safely overlap with JSONReader::JsonParseError.
enum JsonFileError {
JSON_NO_ERROR = 0,
JSON_ACCESS_DENIED = 1000,
JSON_CANNOT_READ_FILE,
JSON_FILE_LOCKED,
JSON_NO_SUCH_FILE
};
// File-specific error messages that can be returned.
static const char kAccessDenied[];
static const char kCannotReadFile[];
static const char kFileLocked[];
static const char kNoSuchFile[];
// Convert an error code into an error message. |error_code| is assumed to
// be a JsonFileError.
static const char* GetErrorMessageForCode(int error_code);
// Returns the size (in bytes) of JSON string read from disk in the last
// successful |Deserialize()| call.
size_t get_last_read_size() const { return last_read_size_; }
private:
// A wrapper for ReadFileToString which returns a non-zero JsonFileError if
// there were file errors.
int ReadFileToString(std::string* json_string);
const base::FilePath json_file_path_;
const int options_;
size_t last_read_size_;
DISALLOW_IMPLICIT_CONSTRUCTORS(JSONFileValueDeserializer);
};
#endif // BASE_JSON_JSON_FILE_VALUE_SERIALIZER_H_

View file

@ -0,0 +1,746 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "base/json/json_parser.h"
#include <cmath>
#include <utility>
#include <vector>
#include "base/logging.h"
#include "base/macros.h"
#include "base/numerics/safe_conversions.h"
#include "base/strings/string_number_conversions.h"
#include "base/strings/string_piece.h"
#include "base/strings/string_util.h"
#include "base/strings/stringprintf.h"
#include "base/strings/utf_string_conversion_utils.h"
#include "base/strings/utf_string_conversions.h"
#include "base/third_party/icu/icu_utf.h"
#include "base/values.h"
namespace base {
namespace internal {
namespace {
const int32_t kExtendedASCIIStart = 0x80;
constexpr uint32_t kUnicodeReplacementPoint = 0xFFFD;
// UnprefixedHexStringToInt acts like |HexStringToInt|, but enforces that the
// input consists purely of hex digits. I.e. no "0x" nor "OX" prefix is
// permitted.
bool UnprefixedHexStringToInt(StringPiece input, int* output) {
for (size_t i = 0; i < input.size(); i++) {
if (!IsHexDigit(input[i])) {
return false;
}
}
return HexStringToInt(input, output);
}
} // namespace
// This is U+FFFD.
const char kUnicodeReplacementString[] = "\xEF\xBF\xBD";
JSONParser::JSONParser(int options, size_t max_depth)
: options_(options),
max_depth_(max_depth),
index_(0),
stack_depth_(0),
line_number_(0),
index_last_line_(0),
error_code_(JSONReader::JSON_NO_ERROR),
error_line_(0),
error_column_(0) {
CHECK_LE(max_depth, kAbsoluteMaxDepth);
}
JSONParser::~JSONParser() = default;
Optional<Value> JSONParser::Parse(StringPiece input) {
input_ = input;
index_ = 0;
line_number_ = 1;
index_last_line_ = 0;
error_code_ = JSONReader::JSON_NO_ERROR;
error_line_ = 0;
error_column_ = 0;
// ICU and ReadUnicodeCharacter() use int32_t for lengths, so ensure
// that the index_ will not overflow when parsing.
if (!base::IsValueInRangeForNumericType<int32_t>(input.length())) {
ReportError(JSONReader::JSON_TOO_LARGE, 0);
return nullopt;
}
// When the input JSON string starts with a UTF-8 Byte-Order-Mark,
// advance the start position to avoid the ParseNextToken function mis-
// treating a Unicode BOM as an invalid character and returning NULL.
ConsumeIfMatch("\xEF\xBB\xBF");
// Parse the first and any nested tokens.
Optional<Value> root(ParseNextToken());
if (!root)
return nullopt;
// Make sure the input stream is at an end.
if (GetNextToken() != T_END_OF_INPUT) {
ReportError(JSONReader::JSON_UNEXPECTED_DATA_AFTER_ROOT, 1);
return nullopt;
}
return root;
}
JSONReader::JsonParseError JSONParser::error_code() const {
return error_code_;
}
std::string JSONParser::GetErrorMessage() const {
return FormatErrorMessage(error_line_, error_column_,
JSONReader::ErrorCodeToString(error_code_));
}
int JSONParser::error_line() const {
return error_line_;
}
int JSONParser::error_column() const {
return error_column_;
}
// StringBuilder ///////////////////////////////////////////////////////////////
JSONParser::StringBuilder::StringBuilder() : StringBuilder(nullptr) {}
JSONParser::StringBuilder::StringBuilder(const char* pos)
: pos_(pos), length_(0) {}
JSONParser::StringBuilder::~StringBuilder() = default;
JSONParser::StringBuilder& JSONParser::StringBuilder::operator=(
StringBuilder&& other) = default;
void JSONParser::StringBuilder::Append(uint32_t point) {
DCHECK(IsValidCodepoint(point));
if (point < kExtendedASCIIStart && !string_) {
DCHECK_EQ(static_cast<char>(point), pos_[length_]);
++length_;
} else {
Convert();
if (UNLIKELY(point == kUnicodeReplacementPoint)) {
string_->append(kUnicodeReplacementString);
} else {
WriteUnicodeCharacter(point, &*string_);
}
}
}
void JSONParser::StringBuilder::Convert() {
if (string_)
return;
string_.emplace(pos_, length_);
}
std::string JSONParser::StringBuilder::DestructiveAsString() {
if (string_)
return std::move(*string_);
return std::string(pos_, length_);
}
// JSONParser private //////////////////////////////////////////////////////////
Optional<StringPiece> JSONParser::PeekChars(size_t count) {
if (index_ + count > input_.length())
return nullopt;
// Using StringPiece::substr() is significantly slower (according to
// base_perftests) than constructing a substring manually.
return StringPiece(input_.data() + index_, count);
}
Optional<char> JSONParser::PeekChar() {
Optional<StringPiece> chars = PeekChars(1);
if (chars)
return (*chars)[0];
return nullopt;
}
Optional<StringPiece> JSONParser::ConsumeChars(size_t count) {
Optional<StringPiece> chars = PeekChars(count);
if (chars)
index_ += count;
return chars;
}
Optional<char> JSONParser::ConsumeChar() {
Optional<StringPiece> chars = ConsumeChars(1);
if (chars)
return (*chars)[0];
return nullopt;
}
const char* JSONParser::pos() {
CHECK_LE(static_cast<size_t>(index_), input_.length());
return input_.data() + index_;
}
JSONParser::Token JSONParser::GetNextToken() {
EatWhitespaceAndComments();
Optional<char> c = PeekChar();
if (!c)
return T_END_OF_INPUT;
switch (*c) {
case '{':
return T_OBJECT_BEGIN;
case '}':
return T_OBJECT_END;
case '[':
return T_ARRAY_BEGIN;
case ']':
return T_ARRAY_END;
case '"':
return T_STRING;
case '0':
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
case '8':
case '9':
case '-':
return T_NUMBER;
case 't':
return T_BOOL_TRUE;
case 'f':
return T_BOOL_FALSE;
case 'n':
return T_NULL;
case ',':
return T_LIST_SEPARATOR;
case ':':
return T_OBJECT_PAIR_SEPARATOR;
default:
return T_INVALID_TOKEN;
}
}
void JSONParser::EatWhitespaceAndComments() {
while (Optional<char> c = PeekChar()) {
switch (*c) {
case '\r':
case '\n':
index_last_line_ = index_;
// Don't increment line_number_ twice for "\r\n".
if (!(c == '\n' && index_ > 0 && input_[index_ - 1] == '\r')) {
++line_number_;
}
FALLTHROUGH;
case ' ':
case '\t':
ConsumeChar();
break;
case '/':
if (!EatComment())
return;
break;
default:
return;
}
}
}
bool JSONParser::EatComment() {
Optional<StringPiece> comment_start = PeekChars(2);
if (!comment_start)
return false;
if (comment_start == "//") {
ConsumeChars(2);
// Single line comment, read to newline.
while (Optional<char> c = PeekChar()) {
if (c == '\n' || c == '\r')
return true;
ConsumeChar();
}
} else if (comment_start == "/*") {
ConsumeChars(2);
char previous_char = '\0';
// Block comment, read until end marker.
while (Optional<char> c = PeekChar()) {
if (previous_char == '*' && c == '/') {
// EatWhitespaceAndComments will inspect pos(), which will still be on
// the last / of the comment, so advance once more (which may also be
// end of input).
ConsumeChar();
return true;
}
previous_char = *ConsumeChar();
}
// If the comment is unterminated, GetNextToken will report T_END_OF_INPUT.
}
return false;
}
Optional<Value> JSONParser::ParseNextToken() {
return ParseToken(GetNextToken());
}
Optional<Value> JSONParser::ParseToken(Token token) {
switch (token) {
case T_OBJECT_BEGIN:
return ConsumeDictionary();
case T_ARRAY_BEGIN:
return ConsumeList();
case T_STRING:
return ConsumeString();
case T_NUMBER:
return ConsumeNumber();
case T_BOOL_TRUE:
case T_BOOL_FALSE:
case T_NULL:
return ConsumeLiteral();
default:
ReportError(JSONReader::JSON_UNEXPECTED_TOKEN, 1);
return nullopt;
}
}
Optional<Value> JSONParser::ConsumeDictionary() {
if (ConsumeChar() != '{') {
ReportError(JSONReader::JSON_UNEXPECTED_TOKEN, 1);
return nullopt;
}
StackMarker depth_check(max_depth_, &stack_depth_);
if (depth_check.IsTooDeep()) {
ReportError(JSONReader::JSON_TOO_MUCH_NESTING, 0);
return nullopt;
}
std::vector<Value::DictStorage::value_type> dict_storage;
Token token = GetNextToken();
while (token != T_OBJECT_END) {
if (token != T_STRING) {
ReportError(JSONReader::JSON_UNQUOTED_DICTIONARY_KEY, 1);
return nullopt;
}
// First consume the key.
StringBuilder key;
if (!ConsumeStringRaw(&key)) {
return nullopt;
}
// Read the separator.
token = GetNextToken();
if (token != T_OBJECT_PAIR_SEPARATOR) {
ReportError(JSONReader::JSON_SYNTAX_ERROR, 1);
return nullopt;
}
// The next token is the value. Ownership transfers to |dict|.
ConsumeChar();
Optional<Value> value = ParseNextToken();
if (!value) {
// ReportError from deeper level.
return nullopt;
}
dict_storage.emplace_back(key.DestructiveAsString(),
std::make_unique<Value>(std::move(*value)));
token = GetNextToken();
if (token == T_LIST_SEPARATOR) {
ConsumeChar();
token = GetNextToken();
if (token == T_OBJECT_END && !(options_ & JSON_ALLOW_TRAILING_COMMAS)) {
ReportError(JSONReader::JSON_TRAILING_COMMA, 1);
return nullopt;
}
} else if (token != T_OBJECT_END) {
ReportError(JSONReader::JSON_SYNTAX_ERROR, 0);
return nullopt;
}
}
ConsumeChar(); // Closing '}'.
// Reverse |dict_storage| to keep the last of elements with the same key in
// the input.
std::reverse(dict_storage.begin(), dict_storage.end());
return Value(Value::DictStorage(std::move(dict_storage)));
}
Optional<Value> JSONParser::ConsumeList() {
if (ConsumeChar() != '[') {
ReportError(JSONReader::JSON_UNEXPECTED_TOKEN, 1);
return nullopt;
}
StackMarker depth_check(max_depth_, &stack_depth_);
if (depth_check.IsTooDeep()) {
ReportError(JSONReader::JSON_TOO_MUCH_NESTING, 0);
return nullopt;
}
Value::ListStorage list_storage;
Token token = GetNextToken();
while (token != T_ARRAY_END) {
Optional<Value> item = ParseToken(token);
if (!item) {
// ReportError from deeper level.
return nullopt;
}
list_storage.push_back(std::move(*item));
token = GetNextToken();
if (token == T_LIST_SEPARATOR) {
ConsumeChar();
token = GetNextToken();
if (token == T_ARRAY_END && !(options_ & JSON_ALLOW_TRAILING_COMMAS)) {
ReportError(JSONReader::JSON_TRAILING_COMMA, 1);
return nullopt;
}
} else if (token != T_ARRAY_END) {
ReportError(JSONReader::JSON_SYNTAX_ERROR, 1);
return nullopt;
}
}
ConsumeChar(); // Closing ']'.
return Value(std::move(list_storage));
}
Optional<Value> JSONParser::ConsumeString() {
StringBuilder string;
if (!ConsumeStringRaw(&string))
return nullopt;
return Value(string.DestructiveAsString());
}
bool JSONParser::ConsumeStringRaw(StringBuilder* out) {
if (ConsumeChar() != '"') {
ReportError(JSONReader::JSON_UNEXPECTED_TOKEN, 1);
return false;
}
// StringBuilder will internally build a StringPiece unless a UTF-16
// conversion occurs, at which point it will perform a copy into a
// std::string.
StringBuilder string(pos());
while (PeekChar()) {
uint32_t next_char = 0;
if (!ReadUnicodeCharacter(input_.data(),
static_cast<int32_t>(input_.length()), &index_,
&next_char) ||
!IsValidCodepoint(next_char)) {
if ((options_ & JSON_REPLACE_INVALID_CHARACTERS) == 0) {
ReportError(JSONReader::JSON_UNSUPPORTED_ENCODING, 1);
return false;
}
ConsumeChar();
string.Append(kUnicodeReplacementPoint);
continue;
}
if (next_char == '"') {
ConsumeChar();
*out = std::move(string);
return true;
}
if (next_char != '\\') {
// If this character is not an escape sequence...
ConsumeChar();
string.Append(next_char);
} else {
// And if it is an escape sequence, the input string will be adjusted
// (either by combining the two characters of an encoded escape sequence,
// or with a UTF conversion), so using StringPiece isn't possible -- force
// a conversion.
string.Convert();
// Read past the escape '\' and ensure there's a character following.
Optional<StringPiece> escape_sequence = ConsumeChars(2);
if (!escape_sequence) {
ReportError(JSONReader::JSON_INVALID_ESCAPE, 0);
return false;
}
switch ((*escape_sequence)[1]) {
// Allowed esape sequences:
case 'x': { // UTF-8 sequence.
// UTF-8 \x escape sequences are not allowed in the spec, but they
// are supported here for backwards-compatiblity with the old parser.
escape_sequence = ConsumeChars(2);
if (!escape_sequence) {
ReportError(JSONReader::JSON_INVALID_ESCAPE, -2);
return false;
}
int hex_digit = 0;
if (!UnprefixedHexStringToInt(*escape_sequence, &hex_digit) ||
!IsValidCharacter(hex_digit)) {
ReportError(JSONReader::JSON_INVALID_ESCAPE, -2);
return false;
}
string.Append(hex_digit);
break;
}
case 'u': { // UTF-16 sequence.
// UTF units are of the form \uXXXX.
uint32_t code_point;
if (!DecodeUTF16(&code_point)) {
ReportError(JSONReader::JSON_INVALID_ESCAPE, 0);
return false;
}
string.Append(code_point);
break;
}
case '"':
string.Append('"');
break;
case '\\':
string.Append('\\');
break;
case '/':
string.Append('/');
break;
case 'b':
string.Append('\b');
break;
case 'f':
string.Append('\f');
break;
case 'n':
string.Append('\n');
break;
case 'r':
string.Append('\r');
break;
case 't':
string.Append('\t');
break;
case 'v': // Not listed as valid escape sequence in the RFC.
string.Append('\v');
break;
// All other escape squences are illegal.
default:
ReportError(JSONReader::JSON_INVALID_ESCAPE, 0);
return false;
}
}
}
ReportError(JSONReader::JSON_SYNTAX_ERROR, 0);
return false;
}
// Entry is at the first X in \uXXXX.
bool JSONParser::DecodeUTF16(uint32_t* out_code_point) {
Optional<StringPiece> escape_sequence = ConsumeChars(4);
if (!escape_sequence)
return false;
// Consume the UTF-16 code unit, which may be a high surrogate.
int code_unit16_high = 0;
if (!UnprefixedHexStringToInt(*escape_sequence, &code_unit16_high))
return false;
// If this is a high surrogate, consume the next code unit to get the
// low surrogate.
if (CBU16_IS_SURROGATE(code_unit16_high)) {
// Make sure this is the high surrogate. If not, it's an encoding
// error.
if (!CBU16_IS_SURROGATE_LEAD(code_unit16_high))
return false;
// Make sure that the token has more characters to consume the
// lower surrogate.
if (!ConsumeIfMatch("\\u")) {
if ((options_ & JSON_REPLACE_INVALID_CHARACTERS) == 0)
return false;
*out_code_point = kUnicodeReplacementPoint;
return true;
}
escape_sequence = ConsumeChars(4);
if (!escape_sequence)
return false;
int code_unit16_low = 0;
if (!UnprefixedHexStringToInt(*escape_sequence, &code_unit16_low))
return false;
if (!CBU16_IS_TRAIL(code_unit16_low)) {
if ((options_ & JSON_REPLACE_INVALID_CHARACTERS) == 0)
return false;
*out_code_point = kUnicodeReplacementPoint;
return true;
}
uint32_t code_point =
CBU16_GET_SUPPLEMENTARY(code_unit16_high, code_unit16_low);
*out_code_point = code_point;
} else {
// Not a surrogate.
DCHECK(CBU16_IS_SINGLE(code_unit16_high));
*out_code_point = code_unit16_high;
}
return true;
}
Optional<Value> JSONParser::ConsumeNumber() {
const char* num_start = pos();
const int start_index = index_;
int end_index = start_index;
if (PeekChar() == '-')
ConsumeChar();
if (!ReadInt(false)) {
ReportError(JSONReader::JSON_SYNTAX_ERROR, 1);
return nullopt;
}
end_index = index_;
// The optional fraction part.
if (PeekChar() == '.') {
ConsumeChar();
if (!ReadInt(true)) {
ReportError(JSONReader::JSON_SYNTAX_ERROR, 1);
return nullopt;
}
end_index = index_;
}
// Optional exponent part.
Optional<char> c = PeekChar();
if (c == 'e' || c == 'E') {
ConsumeChar();
if (PeekChar() == '-' || PeekChar() == '+') {
ConsumeChar();
}
if (!ReadInt(true)) {
ReportError(JSONReader::JSON_SYNTAX_ERROR, 1);
return nullopt;
}
end_index = index_;
}
// ReadInt is greedy because numbers have no easily detectable sentinel,
// so save off where the parser should be on exit (see Consume invariant at
// the top of the header), then make sure the next token is one which is
// valid.
int exit_index = index_;
switch (GetNextToken()) {
case T_OBJECT_END:
case T_ARRAY_END:
case T_LIST_SEPARATOR:
case T_END_OF_INPUT:
break;
default:
ReportError(JSONReader::JSON_SYNTAX_ERROR, 1);
return nullopt;
}
index_ = exit_index;
StringPiece num_string(num_start, end_index - start_index);
int num_int;
if (StringToInt(num_string, &num_int))
return Value(num_int);
double num_double;
if (StringToDouble(num_string.as_string(), &num_double) &&
std::isfinite(num_double)) {
return Value(num_double);
}
ReportError(JSONReader::JSON_UNREPRESENTABLE_NUMBER, 1);
return nullopt;
}
bool JSONParser::ReadInt(bool allow_leading_zeros) {
size_t len = 0;
char first = 0;
while (Optional<char> c = PeekChar()) {
if (!IsAsciiDigit(c))
break;
if (len == 0)
first = *c;
++len;
ConsumeChar();
}
if (len == 0)
return false;
if (!allow_leading_zeros && len > 1 && first == '0')
return false;
return true;
}
Optional<Value> JSONParser::ConsumeLiteral() {
if (ConsumeIfMatch("true"))
return Value(true);
if (ConsumeIfMatch("false"))
return Value(false);
if (ConsumeIfMatch("null"))
return Value(Value::Type::NONE);
ReportError(JSONReader::JSON_SYNTAX_ERROR, 1);
return nullopt;
}
bool JSONParser::ConsumeIfMatch(StringPiece match) {
if (match == PeekChars(match.size())) {
ConsumeChars(match.size());
return true;
}
return false;
}
void JSONParser::ReportError(JSONReader::JsonParseError code,
int column_adjust) {
error_code_ = code;
error_line_ = line_number_;
error_column_ = index_ - index_last_line_ + column_adjust;
}
// static
std::string JSONParser::FormatErrorMessage(int line, int column,
const std::string& description) {
if (line || column) {
return StringPrintf("Line: %i, column: %i, %s",
line, column, description.c_str());
}
return description;
}
} // namespace internal
} // namespace base

View file

@ -0,0 +1,261 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#ifndef BASE_JSON_JSON_PARSER_H_
#define BASE_JSON_JSON_PARSER_H_
#include <stddef.h>
#include <stdint.h>
#include <memory>
#include <string>
#include "base/base_export.h"
#include "base/compiler_specific.h"
#include "base/gtest_prod_util.h"
#include "base/json/json_common.h"
#include "base/json/json_reader.h"
#include "base/macros.h"
#include "base/optional.h"
#include "base/strings/string_piece.h"
namespace base {
class Value;
namespace internal {
class JSONParserTest;
// The implementation behind the JSONReader interface. This class is not meant
// to be used directly; it encapsulates logic that need not be exposed publicly.
//
// This parser guarantees O(n) time through the input string. Iteration happens
// on the byte level, with the functions ConsumeChars() and ConsumeChar(). The
// conversion from byte to JSON token happens without advancing the parser in
// GetNextToken/ParseToken, that is tokenization operates on the current parser
// position without advancing.
//
// Built on top of these are a family of Consume functions that iterate
// internally. Invariant: on entry of a Consume function, the parser is wound
// to the first byte of a valid JSON token. On exit, it is on the first byte
// after the token that was just consumed, which would likely be the first byte
// of the next token.
class BASE_EXPORT JSONParser {
public:
JSONParser(int options, size_t max_depth = kAbsoluteMaxDepth);
~JSONParser();
// Parses the input string according to the set options and returns the
// result as a Value.
// Wrap this in base::FooValue::From() to check the Value is of type Foo and
// convert to a FooValue at the same time.
Optional<Value> Parse(StringPiece input);
// Returns the error code.
JSONReader::JsonParseError error_code() const;
// Returns the human-friendly error message.
std::string GetErrorMessage() const;
// Returns the error line number if parse error happened. Otherwise always
// returns 0.
int error_line() const;
// Returns the error column number if parse error happened. Otherwise always
// returns 0.
int error_column() const;
private:
enum Token {
T_OBJECT_BEGIN, // {
T_OBJECT_END, // }
T_ARRAY_BEGIN, // [
T_ARRAY_END, // ]
T_STRING,
T_NUMBER,
T_BOOL_TRUE, // true
T_BOOL_FALSE, // false
T_NULL, // null
T_LIST_SEPARATOR, // ,
T_OBJECT_PAIR_SEPARATOR, // :
T_END_OF_INPUT,
T_INVALID_TOKEN,
};
// A helper class used for parsing strings. One optimization performed is to
// create base::Value with a StringPiece to avoid unnecessary std::string
// copies. This is not possible if the input string needs to be decoded from
// UTF-16 to UTF-8, or if an escape sequence causes characters to be skipped.
// This class centralizes that logic.
class StringBuilder {
public:
// Empty constructor. Used for creating a builder with which to assign to.
StringBuilder();
// |pos| is the beginning of an input string, excluding the |"|.
explicit StringBuilder(const char* pos);
~StringBuilder();
StringBuilder& operator=(StringBuilder&& other);
// Appends the Unicode code point |point| to the string, either by
// increasing the |length_| of the string if the string has not been
// converted, or by appending the UTF8 bytes for the code point.
void Append(uint32_t point);
// Converts the builder from its default StringPiece to a full std::string,
// performing a copy. Once a builder is converted, it cannot be made a
// StringPiece again.
void Convert();
// Returns the builder as a string, invalidating all state. This allows
// the internal string buffer representation to be destructively moved
// in cases where the builder will not be needed any more.
std::string DestructiveAsString();
private:
// The beginning of the input string.
const char* pos_;
// Number of bytes in |pos_| that make up the string being built.
size_t length_;
// The copied string representation. Will be unset until Convert() is
// called.
base::Optional<std::string> string_;
};
// Returns the next |count| bytes of the input stream, or nullopt if fewer
// than |count| bytes remain.
Optional<StringPiece> PeekChars(size_t count);
// Calls PeekChars() with a |count| of 1.
Optional<char> PeekChar();
// Returns the next |count| bytes of the input stream, or nullopt if fewer
// than |count| bytes remain, and advances the parser position by |count|.
Optional<StringPiece> ConsumeChars(size_t count);
// Calls ConsumeChars() with a |count| of 1.
Optional<char> ConsumeChar();
// Returns a pointer to the current character position.
const char* pos();
// Skips over whitespace and comments to find the next token in the stream.
// This does not advance the parser for non-whitespace or comment chars.
Token GetNextToken();
// Consumes whitespace characters and comments until the next non-that is
// encountered.
void EatWhitespaceAndComments();
// Helper function that consumes a comment, assuming that the parser is
// currently wound to a '/'.
bool EatComment();
// Calls GetNextToken() and then ParseToken().
Optional<Value> ParseNextToken();
// Takes a token that represents the start of a Value ("a structural token"
// in RFC terms) and consumes it, returning the result as a Value.
Optional<Value> ParseToken(Token token);
// Assuming that the parser is currently wound to '{', this parses a JSON
// object into a Value.
Optional<Value> ConsumeDictionary();
// Assuming that the parser is wound to '[', this parses a JSON list into a
// Value.
Optional<Value> ConsumeList();
// Calls through ConsumeStringRaw and wraps it in a value.
Optional<Value> ConsumeString();
// Assuming that the parser is wound to a double quote, this parses a string,
// decoding any escape sequences and converts UTF-16 to UTF-8. Returns true on
// success and places result into |out|. Returns false on failure with
// error information set.
bool ConsumeStringRaw(StringBuilder* out);
// Helper function for ConsumeStringRaw() that consumes the next four or 10
// bytes (parser is wound to the first character of a HEX sequence, with the
// potential for consuming another \uXXXX for a surrogate). Returns true on
// success and places the code point |out_code_point|, and false on failure.
bool DecodeUTF16(uint32_t* out_code_point);
// Assuming that the parser is wound to the start of a valid JSON number,
// this parses and converts it to either an int or double value.
Optional<Value> ConsumeNumber();
// Helper that reads characters that are ints. Returns true if a number was
// read and false on error.
bool ReadInt(bool allow_leading_zeros);
// Consumes the literal values of |true|, |false|, and |null|, assuming the
// parser is wound to the first character of any of those.
Optional<Value> ConsumeLiteral();
// Helper function that returns true if the byte squence |match| can be
// consumed at the current parser position. Returns false if there are fewer
// than |match|-length bytes or if the sequence does not match, and the
// parser state is unchanged.
bool ConsumeIfMatch(StringPiece match);
// Sets the error information to |code| at the current column, based on
// |index_| and |index_last_line_|, with an optional positive/negative
// adjustment by |column_adjust|.
void ReportError(JSONReader::JsonParseError code, int column_adjust);
// Given the line and column number of an error, formats one of the error
// message contants from json_reader.h for human display.
static std::string FormatErrorMessage(int line, int column,
const std::string& description);
// base::JSONParserOptions that control parsing.
const int options_;
// Maximum depth to parse.
const size_t max_depth_;
// The input stream being parsed. Note: Not guaranteed to NUL-terminated.
StringPiece input_;
// The index in the input stream to which the parser is wound.
int index_;
// The number of times the parser has recursed (current stack depth).
size_t stack_depth_;
// The line number that the parser is at currently.
int line_number_;
// The last value of |index_| on the previous line.
int index_last_line_;
// Error information.
JSONReader::JsonParseError error_code_;
int error_line_;
int error_column_;
friend class JSONParserTest;
FRIEND_TEST_ALL_PREFIXES(JSONParserTest, NextChar);
FRIEND_TEST_ALL_PREFIXES(JSONParserTest, ConsumeDictionary);
FRIEND_TEST_ALL_PREFIXES(JSONParserTest, ConsumeList);
FRIEND_TEST_ALL_PREFIXES(JSONParserTest, ConsumeString);
FRIEND_TEST_ALL_PREFIXES(JSONParserTest, ConsumeLiterals);
FRIEND_TEST_ALL_PREFIXES(JSONParserTest, ConsumeNumbers);
FRIEND_TEST_ALL_PREFIXES(JSONParserTest, ErrorMessages);
FRIEND_TEST_ALL_PREFIXES(JSONParserTest, ReplaceInvalidCharacters);
FRIEND_TEST_ALL_PREFIXES(JSONParserTest, ReplaceInvalidUTF16EscapeSequence);
DISALLOW_COPY_AND_ASSIGN(JSONParser);
};
// Used when decoding and an invalid utf-8 sequence is encountered.
BASE_EXPORT extern const char kUnicodeReplacementString[];
} // namespace internal
} // namespace base
#endif // BASE_JSON_JSON_PARSER_H_

View file

@ -0,0 +1,158 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "base/json/json_reader.h"
#include <utility>
#include <vector>
#include "base/json/json_parser.h"
#include "base/logging.h"
#include "base/optional.h"
namespace base {
// Values 1000 and above are used by JSONFileValueSerializer::JsonFileError.
static_assert(JSONReader::JSON_PARSE_ERROR_COUNT < 1000,
"JSONReader error out of bounds");
const char JSONReader::kInvalidEscape[] =
"Invalid escape sequence.";
const char JSONReader::kSyntaxError[] =
"Syntax error.";
const char JSONReader::kUnexpectedToken[] =
"Unexpected token.";
const char JSONReader::kTrailingComma[] =
"Trailing comma not allowed.";
const char JSONReader::kTooMuchNesting[] =
"Too much nesting.";
const char JSONReader::kUnexpectedDataAfterRoot[] =
"Unexpected data after root element.";
const char JSONReader::kUnsupportedEncoding[] =
"Unsupported encoding. JSON must be UTF-8.";
const char JSONReader::kUnquotedDictionaryKey[] =
"Dictionary keys must be quoted.";
const char JSONReader::kInputTooLarge[] =
"Input string is too large (>2GB).";
const char JSONReader::kUnrepresentableNumber[] =
"Number cannot be represented.";
JSONReader::ValueWithError::ValueWithError() = default;
JSONReader::ValueWithError::ValueWithError(ValueWithError&& other) = default;
JSONReader::ValueWithError::~ValueWithError() = default;
JSONReader::ValueWithError& JSONReader::ValueWithError::operator=(
ValueWithError&& other) = default;
JSONReader::JSONReader(int options, size_t max_depth)
: parser_(new internal::JSONParser(options, max_depth)) {}
JSONReader::~JSONReader() = default;
// static
Optional<Value> JSONReader::Read(StringPiece json,
int options,
size_t max_depth) {
internal::JSONParser parser(options, max_depth);
return parser.Parse(json);
}
std::unique_ptr<Value> JSONReader::ReadDeprecated(StringPiece json,
int options,
size_t max_depth) {
Optional<Value> value = Read(json, options, max_depth);
return value ? Value::ToUniquePtrValue(std::move(*value)) : nullptr;
}
// static
JSONReader::ValueWithError JSONReader::ReadAndReturnValueWithError(
StringPiece json,
int options) {
ValueWithError ret;
internal::JSONParser parser(options);
ret.value = parser.Parse(json);
if (!ret.value) {
ret.error_message = parser.GetErrorMessage();
ret.error_code = parser.error_code();
ret.error_line = parser.error_line();
ret.error_column = parser.error_column();
}
return ret;
}
// static
std::unique_ptr<Value> JSONReader::ReadAndReturnErrorDeprecated(
StringPiece json,
int options,
int* error_code_out,
std::string* error_msg_out,
int* error_line_out,
int* error_column_out) {
ValueWithError ret = ReadAndReturnValueWithError(json, options);
if (ret.value)
return Value::ToUniquePtrValue(std::move(*ret.value));
if (error_code_out)
*error_code_out = ret.error_code;
if (error_msg_out)
*error_msg_out = ret.error_message;
if (error_line_out)
*error_line_out = ret.error_line;
if (error_column_out)
*error_column_out = ret.error_column;
return nullptr;
}
// static
std::string JSONReader::ErrorCodeToString(JsonParseError error_code) {
switch (error_code) {
case JSON_NO_ERROR:
return std::string();
case JSON_INVALID_ESCAPE:
return kInvalidEscape;
case JSON_SYNTAX_ERROR:
return kSyntaxError;
case JSON_UNEXPECTED_TOKEN:
return kUnexpectedToken;
case JSON_TRAILING_COMMA:
return kTrailingComma;
case JSON_TOO_MUCH_NESTING:
return kTooMuchNesting;
case JSON_UNEXPECTED_DATA_AFTER_ROOT:
return kUnexpectedDataAfterRoot;
case JSON_UNSUPPORTED_ENCODING:
return kUnsupportedEncoding;
case JSON_UNQUOTED_DICTIONARY_KEY:
return kUnquotedDictionaryKey;
case JSON_TOO_LARGE:
return kInputTooLarge;
case JSON_UNREPRESENTABLE_NUMBER:
return kUnrepresentableNumber;
case JSON_PARSE_ERROR_COUNT:
break;
}
NOTREACHED();
return std::string();
}
Optional<Value> JSONReader::ReadToValue(StringPiece json) {
return parser_->Parse(json);
}
std::unique_ptr<Value> JSONReader::ReadToValueDeprecated(StringPiece json) {
Optional<Value> value = parser_->Parse(json);
return value ? std::make_unique<Value>(std::move(*value)) : nullptr;
}
JSONReader::JsonParseError JSONReader::error_code() const {
return parser_->error_code();
}
std::string JSONReader::GetErrorMessage() const {
return parser_->GetErrorMessage();
}
} // namespace base

View file

@ -0,0 +1,176 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
//
// A JSON parser. Converts strings of JSON into a Value object (see
// base/values.h).
// http://www.ietf.org/rfc/rfc4627.txt?number=4627
//
// Known limitations/deviations from the RFC:
// - Only knows how to parse ints within the range of a signed 32 bit int and
// decimal numbers within a double.
// - Assumes input is encoded as UTF8. The spec says we should allow UTF-16
// (BE or LE) and UTF-32 (BE or LE) as well.
// - We limit nesting to 100 levels to prevent stack overflow (this is allowed
// by the RFC).
// - A Unicode FAQ ("http://unicode.org/faq/utf_bom.html") writes a data
// stream may start with a Unicode Byte-Order-Mark (U+FEFF), i.e. the input
// UTF-8 string for the JSONReader::JsonToValue() function may start with a
// UTF-8 BOM (0xEF, 0xBB, 0xBF).
// To avoid the function from mis-treating a UTF-8 BOM as an invalid
// character, the function skips a Unicode BOM at the beginning of the
// Unicode string (converted from the input UTF-8 string) before parsing it.
//
// TODO(tc): Add a parsing option to to relax object keys being wrapped in
// double quotes
// TODO(tc): Add an option to disable comment stripping
#ifndef BASE_JSON_JSON_READER_H_
#define BASE_JSON_JSON_READER_H_
#include <memory>
#include <string>
#include "base/base_export.h"
#include "base/json/json_common.h"
#include "base/optional.h"
#include "base/strings/string_piece.h"
#include "base/values.h"
namespace base {
namespace internal {
class JSONParser;
}
enum JSONParserOptions {
// Parses the input strictly according to RFC 4627, except for where noted
// above.
JSON_PARSE_RFC = 0,
// Allows commas to exist after the last element in structures.
JSON_ALLOW_TRAILING_COMMAS = 1 << 0,
// If set the parser replaces invalid code points (i.e. lone
// surrogates) with the Unicode replacement character (U+FFFD). If
// not set, invalid code points trigger a hard error and parsing
// fails.
JSON_REPLACE_INVALID_CHARACTERS = 1 << 1,
};
class BASE_EXPORT JSONReader {
public:
// Error codes during parsing.
enum JsonParseError {
JSON_NO_ERROR = 0,
JSON_INVALID_ESCAPE,
JSON_SYNTAX_ERROR,
JSON_UNEXPECTED_TOKEN,
JSON_TRAILING_COMMA,
JSON_TOO_MUCH_NESTING,
JSON_UNEXPECTED_DATA_AFTER_ROOT,
JSON_UNSUPPORTED_ENCODING,
JSON_UNQUOTED_DICTIONARY_KEY,
JSON_TOO_LARGE,
JSON_UNREPRESENTABLE_NUMBER,
JSON_PARSE_ERROR_COUNT
};
struct BASE_EXPORT ValueWithError {
ValueWithError();
ValueWithError(ValueWithError&& other);
ValueWithError& operator=(ValueWithError&& other);
~ValueWithError();
Optional<Value> value;
// Contains default values if |value| exists, or the error status if |value|
// is base::nullopt.
JsonParseError error_code = JSON_NO_ERROR;
std::string error_message;
int error_line = 0;
int error_column = 0;
DISALLOW_COPY_AND_ASSIGN(ValueWithError);
};
// String versions of parse error codes.
static const char kInvalidEscape[];
static const char kSyntaxError[];
static const char kUnexpectedToken[];
static const char kTrailingComma[];
static const char kTooMuchNesting[];
static const char kUnexpectedDataAfterRoot[];
static const char kUnsupportedEncoding[];
static const char kUnquotedDictionaryKey[];
static const char kInputTooLarge[];
static const char kUnrepresentableNumber[];
// Constructs a reader.
JSONReader(int options = JSON_PARSE_RFC,
size_t max_depth = internal::kAbsoluteMaxDepth);
~JSONReader();
// Reads and parses |json|, returning a Value.
// If |json| is not a properly formed JSON string, returns base::nullopt.
static Optional<Value> Read(StringPiece json,
int options = JSON_PARSE_RFC,
size_t max_depth = internal::kAbsoluteMaxDepth);
// Deprecated. Use the Read() method above.
// Reads and parses |json|, returning a Value.
// If |json| is not a properly formed JSON string, returns nullptr.
// Wrap this in base::FooValue::From() to check the Value is of type Foo and
// convert to a FooValue at the same time.
static std::unique_ptr<Value> ReadDeprecated(
StringPiece json,
int options = JSON_PARSE_RFC,
size_t max_depth = internal::kAbsoluteMaxDepth);
// Reads and parses |json| like Read(). Returns a ValueWithError, which on
// error, will be populated with a formatted error message, an error code, and
// the error location if appropriate.
static ValueWithError ReadAndReturnValueWithError(
StringPiece json,
int options = JSON_PARSE_RFC);
// Deprecated. Use the ReadAndReturnValueWithError() method above.
// Reads and parses |json| like Read(). |error_code_out| and |error_msg_out|
// are optional. If specified and nullptr is returned, they will be populated
// an error code and a formatted error message (including error location if
// appropriate). Otherwise, they will be unmodified.
static std::unique_ptr<Value> ReadAndReturnErrorDeprecated(
StringPiece json,
int options, // JSONParserOptions
int* error_code_out,
std::string* error_msg_out,
int* error_line_out = nullptr,
int* error_column_out = nullptr);
// Converts a JSON parse error code into a human readable message.
// Returns an empty string if error_code is JSON_NO_ERROR.
static std::string ErrorCodeToString(JsonParseError error_code);
// Non-static version of Read() above.
Optional<Value> ReadToValue(StringPiece json);
// Deprecated. Use the ReadToValue() method above.
// Non-static version of Read() above.
std::unique_ptr<Value> ReadToValueDeprecated(StringPiece json);
// Returns the error code if the last call to ReadToValue() failed.
// Returns JSON_NO_ERROR otherwise.
JsonParseError error_code() const;
// Converts error_code_ to a human-readable string, including line and column
// numbers if appropriate.
std::string GetErrorMessage() const;
private:
std::unique_ptr<internal::JSONParser> parser_;
};
} // namespace base
#endif // BASE_JSON_JSON_READER_H_

View file

@ -0,0 +1,45 @@
// Copyright 2015 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "base/json/json_reader.h"
#include "base/json/json_writer.h"
#include "base/values.h"
namespace base {
// Entry point for LibFuzzer.
extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
if (size < 2)
return 0;
// Create a copy of input buffer, as otherwise we don't catch
// overflow that touches the last byte (which is used in options).
std::unique_ptr<char[]> input(new char[size - 1]);
memcpy(input.get(), data, size - 1);
StringPiece input_string(input.get(), size - 1);
const int options = data[size - 1];
JSONReader::ValueWithError json_val =
JSONReader::ReadAndReturnValueWithError(input_string, options);
CHECK((json_val.error_code == JSONReader::JSON_NO_ERROR) ==
json_val.value.has_value());
if (json_val.value) {
// Check that the value can be serialized and deserialized back to an
// equivalent |Value|.
const Value& value = json_val.value.value();
std::string serialized;
CHECK(JSONWriter::Write(value, &serialized));
Optional<Value> deserialized = JSONReader::Read(StringPiece(serialized));
CHECK(deserialized);
CHECK(value.Equals(&deserialized.value()));
}
return 0;
}
} // namespace base

View file

@ -0,0 +1,55 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "base/json/json_string_value_serializer.h"
#include "base/json/json_reader.h"
#include "base/json/json_writer.h"
#include "base/logging.h"
using base::Value;
JSONStringValueSerializer::JSONStringValueSerializer(std::string* json_string)
: json_string_(json_string),
pretty_print_(false) {
}
JSONStringValueSerializer::~JSONStringValueSerializer() = default;
bool JSONStringValueSerializer::Serialize(const Value& root) {
return SerializeInternal(root, false);
}
bool JSONStringValueSerializer::SerializeAndOmitBinaryValues(
const Value& root) {
return SerializeInternal(root, true);
}
bool JSONStringValueSerializer::SerializeInternal(const Value& root,
bool omit_binary_values) {
if (!json_string_)
return false;
int options = 0;
if (omit_binary_values)
options |= base::JSONWriter::OPTIONS_OMIT_BINARY_VALUES;
if (pretty_print_)
options |= base::JSONWriter::OPTIONS_PRETTY_PRINT;
return base::JSONWriter::WriteWithOptions(root, options, json_string_);
}
JSONStringValueDeserializer::JSONStringValueDeserializer(
const base::StringPiece& json_string,
int options)
: json_string_(json_string), options_(options) {}
JSONStringValueDeserializer::~JSONStringValueDeserializer() = default;
std::unique_ptr<Value> JSONStringValueDeserializer::Deserialize(
int* error_code,
std::string* error_str) {
return base::JSONReader::ReadAndReturnErrorDeprecated(json_string_, options_,
error_code, error_str);
}

View file

@ -0,0 +1,75 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#ifndef BASE_JSON_JSON_STRING_VALUE_SERIALIZER_H_
#define BASE_JSON_JSON_STRING_VALUE_SERIALIZER_H_
#include <string>
#include "base/base_export.h"
#include "base/files/file_path.h"
#include "base/macros.h"
#include "base/strings/string_piece.h"
#include "base/values.h"
class BASE_EXPORT JSONStringValueSerializer : public base::ValueSerializer {
public:
// |json_string| is the string that will be the destination of the
// serialization. The caller of the constructor retains ownership of the
// string. |json_string| must not be null.
explicit JSONStringValueSerializer(std::string* json_string);
~JSONStringValueSerializer() override;
// Attempt to serialize the data structure represented by Value into
// JSON. If the return value is true, the result will have been written
// into the string passed into the constructor.
bool Serialize(const base::Value& root) override;
// Equivalent to Serialize(root) except binary values are omitted from the
// output.
bool SerializeAndOmitBinaryValues(const base::Value& root);
void set_pretty_print(bool new_value) { pretty_print_ = new_value; }
bool pretty_print() { return pretty_print_; }
private:
bool SerializeInternal(const base::Value& root, bool omit_binary_values);
// Owned by the caller of the constructor.
std::string* json_string_;
bool pretty_print_; // If true, serialization will span multiple lines.
DISALLOW_COPY_AND_ASSIGN(JSONStringValueSerializer);
};
class BASE_EXPORT JSONStringValueDeserializer : public base::ValueDeserializer {
public:
// This retains a reference to the contents of |json_string|, so the data
// must outlive the JSONStringValueDeserializer. |options| is a bitmask of
// JSONParserOptions.
explicit JSONStringValueDeserializer(const base::StringPiece& json_string,
int options = 0);
~JSONStringValueDeserializer() override;
// Attempt to deserialize the data structure encoded in the string passed
// in to the constructor into a structure of Value objects. If the return
// value is null, and if |error_code| is non-null, |error_code| will
// contain an integer error code (a JsonParseError in this case).
// If |error_message| is non-null, it will be filled in with a formatted
// error message including the location of the error if appropriate.
// The caller takes ownership of the returned value.
std::unique_ptr<base::Value> Deserialize(int* error_code,
std::string* error_message) override;
private:
// Data is owned by the caller of the constructor.
base::StringPiece json_string_;
const int options_;
DISALLOW_COPY_AND_ASSIGN(JSONStringValueDeserializer);
};
#endif // BASE_JSON_JSON_STRING_VALUE_SERIALIZER_H_

View file

@ -0,0 +1,59 @@
// Copyright 2014 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "base/json/json_value_converter.h"
#include "base/strings/utf_string_conversions.h"
namespace base {
namespace internal {
bool BasicValueConverter<int>::Convert(
const base::Value& value, int* field) const {
if (!value.is_int())
return false;
if (field)
*field = value.GetInt();
return true;
}
bool BasicValueConverter<std::string>::Convert(
const base::Value& value, std::string* field) const {
if (!value.is_string())
return false;
if (field)
*field = value.GetString();
return true;
}
bool BasicValueConverter<string16>::Convert(
const base::Value& value, string16* field) const {
if (!value.is_string())
return false;
if (field)
*field = base::UTF8ToUTF16(value.GetString());
return true;
}
bool BasicValueConverter<double>::Convert(
const base::Value& value, double* field) const {
if (!value.is_double() && !value.is_int())
return false;
if (field)
*field = value.GetDouble();
return true;
}
bool BasicValueConverter<bool>::Convert(
const base::Value& value, bool* field) const {
if (!value.is_bool())
return false;
if (field)
*field = value.GetBool();
return true;
}
} // namespace internal
} // namespace base

View file

@ -0,0 +1,206 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "base/json/json_writer.h"
#include <stdint.h>
#include <cmath>
#include <limits>
#include "base/json/string_escape.h"
#include "base/logging.h"
#include "base/strings/string_number_conversions.h"
#include "base/strings/utf_string_conversions.h"
#include "base/values.h"
#include "build/build_config.h"
namespace base {
#if defined(OS_WIN)
const char kPrettyPrintLineEnding[] = "\r\n";
#else
const char kPrettyPrintLineEnding[] = "\n";
#endif
// static
bool JSONWriter::Write(const Value& node, std::string* json, size_t max_depth) {
return WriteWithOptions(node, 0, json, max_depth);
}
// static
bool JSONWriter::WriteWithOptions(const Value& node,
int options,
std::string* json,
size_t max_depth) {
json->clear();
// Is there a better way to estimate the size of the output?
json->reserve(1024);
JSONWriter writer(options, json, max_depth);
bool result = writer.BuildJSONString(node, 0U);
if (options & OPTIONS_PRETTY_PRINT)
json->append(kPrettyPrintLineEnding);
return result;
}
JSONWriter::JSONWriter(int options, std::string* json, size_t max_depth)
: omit_binary_values_((options & OPTIONS_OMIT_BINARY_VALUES) != 0),
omit_double_type_preservation_(
(options & OPTIONS_OMIT_DOUBLE_TYPE_PRESERVATION) != 0),
pretty_print_((options & OPTIONS_PRETTY_PRINT) != 0),
json_string_(json),
max_depth_(max_depth),
stack_depth_(0) {
DCHECK(json);
CHECK_LE(max_depth, internal::kAbsoluteMaxDepth);
}
bool JSONWriter::BuildJSONString(const Value& node, size_t depth) {
internal::StackMarker depth_check(max_depth_, &stack_depth_);
if (depth_check.IsTooDeep())
return false;
switch (node.type()) {
case Value::Type::NONE:
json_string_->append("null");
return true;
case Value::Type::BOOLEAN:
// Note: We are explicitly invoking the std::string constructor in order
// to avoid ASAN errors on Windows: https://crbug.com/900041
json_string_->append(node.GetBool() ? std::string("true")
: std::string("false"));
return true;
case Value::Type::INTEGER:
json_string_->append(NumberToString(node.GetInt()));
return true;
case Value::Type::DOUBLE: {
double value = node.GetDouble();
if (omit_double_type_preservation_ &&
value <= std::numeric_limits<int64_t>::max() &&
value >= std::numeric_limits<int64_t>::min() &&
std::floor(value) == value) {
json_string_->append(NumberToString(static_cast<int64_t>(value)));
return true;
}
std::string real = NumberToString(value);
// Ensure that the number has a .0 if there's no decimal or 'e'. This
// makes sure that when we read the JSON back, it's interpreted as a
// real rather than an int.
if (real.find_first_of(".eE") == std::string::npos)
real.append(".0");
// The JSON spec requires that non-integer values in the range (-1,1)
// have a zero before the decimal point - ".52" is not valid, "0.52" is.
if (real[0] == '.') {
real.insert(static_cast<size_t>(0), static_cast<size_t>(1), '0');
} else if (real.length() > 1 && real[0] == '-' && real[1] == '.') {
// "-.1" bad "-0.1" good
real.insert(static_cast<size_t>(1), static_cast<size_t>(1), '0');
}
json_string_->append(real);
return true;
}
case Value::Type::STRING:
EscapeJSONString(node.GetString(), true, json_string_);
return true;
case Value::Type::LIST: {
json_string_->push_back('[');
if (pretty_print_)
json_string_->push_back(' ');
bool first_value_has_been_output = false;
bool result = true;
for (const auto& value : node.GetList()) {
if (omit_binary_values_ && value.type() == Value::Type::BINARY)
continue;
if (first_value_has_been_output) {
json_string_->push_back(',');
if (pretty_print_)
json_string_->push_back(' ');
}
if (!BuildJSONString(value, depth))
result = false;
first_value_has_been_output = true;
}
if (pretty_print_)
json_string_->push_back(' ');
json_string_->push_back(']');
return result;
}
case Value::Type::DICTIONARY: {
json_string_->push_back('{');
if (pretty_print_)
json_string_->append(kPrettyPrintLineEnding);
bool first_value_has_been_output = false;
bool result = true;
for (const auto& pair : node.DictItems()) {
const auto& key = pair.first;
const auto& value = pair.second;
if (omit_binary_values_ && value.type() == Value::Type::BINARY)
continue;
if (first_value_has_been_output) {
json_string_->push_back(',');
if (pretty_print_)
json_string_->append(kPrettyPrintLineEnding);
}
if (pretty_print_)
IndentLine(depth + 1U);
EscapeJSONString(key, true, json_string_);
json_string_->push_back(':');
if (pretty_print_)
json_string_->push_back(' ');
if (!BuildJSONString(value, depth + 1U))
result = false;
first_value_has_been_output = true;
}
if (pretty_print_) {
json_string_->append(kPrettyPrintLineEnding);
IndentLine(depth);
}
json_string_->push_back('}');
return result;
}
case Value::Type::BINARY:
// Successful only if we're allowed to omit it.
DLOG_IF(ERROR, !omit_binary_values_) << "Cannot serialize binary value.";
return omit_binary_values_;
// TODO(crbug.com/859477): Remove after root cause is found.
case Value::Type::DEAD:
CHECK(false);
return false;
}
// TODO(crbug.com/859477): Revert to NOTREACHED() after root cause is found.
CHECK(false);
return false;
}
void JSONWriter::IndentLine(size_t depth) {
json_string_->append(depth * 3U, ' ');
}
} // namespace base

View file

@ -0,0 +1,87 @@
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#ifndef BASE_JSON_JSON_WRITER_H_
#define BASE_JSON_JSON_WRITER_H_
#include <stddef.h>
#include <string>
#include "base/base_export.h"
#include "base/json/json_common.h"
#include "base/macros.h"
namespace base {
class Value;
class BASE_EXPORT JSONWriter {
public:
enum Options {
// This option instructs the writer that if a Binary value is encountered,
// the value (and key if within a dictionary) will be omitted from the
// output, and success will be returned. Otherwise, if a binary value is
// encountered, failure will be returned.
OPTIONS_OMIT_BINARY_VALUES = 1 << 0,
// This option instructs the writer to write doubles that have no fractional
// part as a normal integer (i.e., without using exponential notation
// or appending a '.0') as long as the value is within the range of a
// 64-bit int.
OPTIONS_OMIT_DOUBLE_TYPE_PRESERVATION = 1 << 1,
// Return a slightly nicer formatted json string (pads with whitespace to
// help with readability).
OPTIONS_PRETTY_PRINT = 1 << 2,
};
// Given a root node, generates a JSON string and puts it into |json|.
// The output string is overwritten and not appended.
//
// TODO(tc): Should we generate json if it would be invalid json (e.g.,
// |node| is not a DictionaryValue/ListValue or if there are inf/-inf float
// values)? Return true on success and false on failure.
static bool Write(const Value& node,
std::string* json,
size_t max_depth = internal::kAbsoluteMaxDepth);
// Same as above but with |options| which is a bunch of JSONWriter::Options
// bitwise ORed together. Return true on success and false on failure.
static bool WriteWithOptions(const Value& node,
int options,
std::string* json,
size_t max_depth = internal::kAbsoluteMaxDepth);
private:
JSONWriter(int options,
std::string* json,
size_t max_depth = internal::kAbsoluteMaxDepth);
// Called recursively to build the JSON string. When completed,
// |json_string_| will contain the JSON.
bool BuildJSONString(const Value& node, size_t depth);
// Adds space to json_string_ for the indent level.
void IndentLine(size_t depth);
bool omit_binary_values_;
bool omit_double_type_preservation_;
bool pretty_print_;
// Where we write JSON data as we generate it.
std::string* json_string_;
// Maximum depth to write.
const size_t max_depth_;
// The number of times the writer has recursed (current stack depth).
size_t stack_depth_;
DISALLOW_COPY_AND_ASSIGN(JSONWriter);
};
} // namespace base
#endif // BASE_JSON_JSON_WRITER_H_

View file

@ -0,0 +1,164 @@
// Copyright (c) 2006-2008 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "base/json/string_escape.h"
#include <stddef.h>
#include <stdint.h>
#include <limits>
#include <string>
#include "base/strings/string_util.h"
#include "base/strings/stringprintf.h"
#include "base/strings/utf_string_conversion_utils.h"
#include "base/strings/utf_string_conversions.h"
#include "base/third_party/icu/icu_utf.h"
namespace base {
namespace {
// Format string for printing a \uXXXX escape sequence.
const char kU16EscapeFormat[] = "\\u%04X";
// The code point to output for an invalid input code unit.
const uint32_t kReplacementCodePoint = 0xFFFD;
// Used below in EscapeSpecialCodePoint().
static_assert('<' == 0x3C, "less than sign must be 0x3c");
// Try to escape the |code_point| if it is a known special character. If
// successful, returns true and appends the escape sequence to |dest|. This
// isn't required by the spec, but it's more readable by humans.
bool EscapeSpecialCodePoint(uint32_t code_point, std::string* dest) {
// WARNING: if you add a new case here, you need to update the reader as well.
// Note: \v is in the reader, but not here since the JSON spec doesn't
// allow it.
switch (code_point) {
case '\b':
dest->append("\\b");
break;
case '\f':
dest->append("\\f");
break;
case '\n':
dest->append("\\n");
break;
case '\r':
dest->append("\\r");
break;
case '\t':
dest->append("\\t");
break;
case '\\':
dest->append("\\\\");
break;
case '"':
dest->append("\\\"");
break;
// Escape < to prevent script execution; escaping > is not necessary and
// not doing so save a few bytes.
case '<':
dest->append("\\u003C");
break;
// Escape the "Line Separator" and "Paragraph Separator" characters, since
// they should be treated like a new line \r or \n.
case 0x2028:
dest->append("\\u2028");
break;
case 0x2029:
dest->append("\\u2029");
break;
default:
return false;
}
return true;
}
template <typename S>
bool EscapeJSONStringImpl(const S& str, bool put_in_quotes, std::string* dest) {
bool did_replacement = false;
if (put_in_quotes)
dest->push_back('"');
// Casting is necessary because ICU uses int32_t. Try and do so safely.
CHECK_LE(str.length(),
static_cast<size_t>(std::numeric_limits<int32_t>::max()));
const int32_t length = static_cast<int32_t>(str.length());
for (int32_t i = 0; i < length; ++i) {
uint32_t code_point;
if (!ReadUnicodeCharacter(str.data(), length, &i, &code_point) ||
code_point == static_cast<decltype(code_point)>(CBU_SENTINEL) ||
!IsValidCodepoint(code_point)) {
code_point = kReplacementCodePoint;
did_replacement = true;
}
if (EscapeSpecialCodePoint(code_point, dest))
continue;
// Escape non-printing characters.
if (code_point < 32)
base::StringAppendF(dest, kU16EscapeFormat, code_point);
else
WriteUnicodeCharacter(code_point, dest);
}
if (put_in_quotes)
dest->push_back('"');
return !did_replacement;
}
} // namespace
bool EscapeJSONString(StringPiece str, bool put_in_quotes, std::string* dest) {
return EscapeJSONStringImpl(str, put_in_quotes, dest);
}
bool EscapeJSONString(StringPiece16 str,
bool put_in_quotes,
std::string* dest) {
return EscapeJSONStringImpl(str, put_in_quotes, dest);
}
std::string GetQuotedJSONString(StringPiece str) {
std::string dest;
EscapeJSONStringImpl(str, true, &dest);
return dest;
}
std::string GetQuotedJSONString(StringPiece16 str) {
std::string dest;
EscapeJSONStringImpl(str, true, &dest);
return dest;
}
std::string EscapeBytesAsInvalidJSONString(StringPiece str,
bool put_in_quotes) {
std::string dest;
if (put_in_quotes)
dest.push_back('"');
for (unsigned char c : str) {
if (EscapeSpecialCodePoint(c, &dest))
continue;
if (c < 32 || c > 126)
base::StringAppendF(&dest, kU16EscapeFormat, c);
else
dest.push_back(c);
}
if (put_in_quotes)
dest.push_back('"');
return dest;
}
} // namespace base

View file

@ -0,0 +1,61 @@
// Copyright (c) 2011 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
// This file defines utility functions for escaping strings suitable for JSON.
#ifndef BASE_JSON_STRING_ESCAPE_H_
#define BASE_JSON_STRING_ESCAPE_H_
#include <string>
#include "base/base_export.h"
#include "base/strings/string_piece.h"
namespace base {
// Appends to |dest| an escaped version of |str|. Valid UTF-8 code units and
// characters will pass through from the input to the output. Invalid code
// units and characters will be replaced with the U+FFFD replacement character.
// This function returns true if no replacement was necessary and false if
// there was a lossy replacement. On return, |dest| will contain a valid UTF-8
// JSON string.
//
// Non-printing control characters will be escaped as \uXXXX sequences for
// readability.
//
// If |put_in_quotes| is true, then a leading and trailing double-quote mark
// will be appended to |dest| as well.
BASE_EXPORT bool EscapeJSONString(StringPiece str,
bool put_in_quotes,
std::string* dest);
// Performs a similar function to the UTF-8 StringPiece version above,
// converting UTF-16 code units to UTF-8 code units and escaping non-printing
// control characters. On return, |dest| will contain a valid UTF-8 JSON string.
BASE_EXPORT bool EscapeJSONString(StringPiece16 str,
bool put_in_quotes,
std::string* dest);
// Helper functions that wrap the above two functions but return the value
// instead of appending. |put_in_quotes| is always true.
BASE_EXPORT std::string GetQuotedJSONString(StringPiece str);
BASE_EXPORT std::string GetQuotedJSONString(StringPiece16 str);
// Given an arbitrary byte string |str|, this will escape all non-ASCII bytes
// as \uXXXX escape sequences. This function is *NOT* meant to be used with
// Unicode strings and does not validate |str| as one.
//
// CAVEAT CALLER: The output of this function may not be valid JSON, since
// JSON requires escape sequences to be valid UTF-16 code units. This output
// will be mangled if passed to to the base::JSONReader, since the reader will
// interpret it as UTF-16 and convert it to UTF-8.
//
// The output of this function takes the *appearance* of JSON but is not in
// fact valid according to RFC 4627.
BASE_EXPORT std::string EscapeBytesAsInvalidJSONString(StringPiece str,
bool put_in_quotes);
} // namespace base
#endif // BASE_JSON_STRING_ESCAPE_H_

View file

@ -0,0 +1,37 @@
// Copyright 2018 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "base/json/string_escape.h"
#include <memory>
// Entry point for LibFuzzer.
extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
if (size < 2)
return 0;
const bool put_in_quotes = data[size - 1];
// Create a copy of input buffer, as otherwise we don't catch
// overflow that touches the last byte (which is used in put_in_quotes).
size_t actual_size_char8 = size - 1;
std::unique_ptr<char[]> input(new char[actual_size_char8]);
memcpy(input.get(), data, actual_size_char8);
base::StringPiece input_string(input.get(), actual_size_char8);
std::string escaped_string;
base::EscapeJSONString(input_string, put_in_quotes, &escaped_string);
// Test for wide-strings if available size is even.
if (actual_size_char8 & 1)
return 0;
size_t actual_size_char16 = actual_size_char8 / 2;
base::StringPiece16 input_string16(
reinterpret_cast<base::char16*>(input.get()), actual_size_char16);
escaped_string.clear();
base::EscapeJSONString(input_string16, put_in_quotes, &escaped_string);
return 0;
}