json/include/nlohmann/detail/input/json_sax.hpp

761 lines
22 KiB
C++

#pragma once
#include <cassert> // assert
#include <cstddef>
#include <string> // string
#include <utility> // move
#include <vector> // vector
#include <nlohmann/detail/exceptions.hpp>
#include <nlohmann/detail/macro_scope.hpp>
namespace nlohmann
{
/*!
@brief SAX interface
This class describes the SAX interface used by @ref nlohmann::json::sax_parse.
Each function is called in different situations while the input is parsed. The
boolean return value informs the parser whether to continue processing the
input.
*/
template<typename BasicJsonType>
struct json_sax
{
/// type for (signed) integers
using number_integer_t = typename BasicJsonType::number_integer_t;
/// type for unsigned integers
using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
/// type for floating-point numbers
using number_float_t = typename BasicJsonType::number_float_t;
/// type for strings
using string_t = typename BasicJsonType::string_t;
using binary_t = typename BasicJsonType::binary_t;
/*!
@brief a null value was read
@return whether parsing should proceed
*/
virtual bool null() = 0;
/*!
@brief a boolean value was read
@param[in] val boolean value
@return whether parsing should proceed
*/
virtual bool boolean(bool val) = 0;
/*!
@brief an integer number was read
@param[in] val integer value
@return whether parsing should proceed
*/
virtual bool number_integer(number_integer_t val) = 0;
/*!
@brief an unsigned integer number was read
@param[in] val unsigned integer value
@return whether parsing should proceed
*/
virtual bool number_unsigned(number_unsigned_t val) = 0;
/*!
@brief an floating-point number was read
@param[in] val floating-point value
@param[in] s raw token value
@return whether parsing should proceed
*/
virtual bool number_float(number_float_t val, const string_t& s) = 0;
/*!
@brief a string was read
@param[in] val string value
@return whether parsing should proceed
@note It is safe to move the passed string.
*/
virtual bool string(string_t& val) = 0;
/*!
@brief a binary string was read
@param[in] val binary value
@return whether parsing should proceed
@note It is safe to move the passed binary.
*/
virtual bool binary(binary_t& val) = 0;
/*!
@brief the beginning of an object was read
@param[in] elements number of object elements or -1 if unknown
@return whether parsing should proceed
@note binary formats may report the number of elements
*/
virtual bool start_object(std::size_t elements) = 0;
/*!
@brief an object key was read
@param[in] val object key
@return whether parsing should proceed
@note It is safe to move the passed string.
*/
virtual bool key(string_t& val) = 0;
/*!
@brief the end of an object was read
@return whether parsing should proceed
*/
virtual bool end_object() = 0;
/*!
@brief the beginning of an array was read
@param[in] elements number of array elements or -1 if unknown
@return whether parsing should proceed
@note binary formats may report the number of elements
*/
virtual bool start_array(std::size_t elements) = 0;
/*!
@brief the end of an array was read
@return whether parsing should proceed
*/
virtual bool end_array() = 0;
/*!
@brief a parse error occurred
@param[in] position the position in the input where the error occurs
@param[in] last_token the last read token
@param[in] ex an exception object describing the error
@return whether parsing should proceed (must return false)
*/
virtual bool parse_error(std::size_t position,
const std::string& last_token,
const detail::exception& ex) = 0;
virtual ~json_sax() = default;
};
namespace detail
{
/*!
@brief SAX implementation to create a JSON value from SAX events
This class implements the @ref json_sax interface and processes the SAX events
to create a JSON value which makes it basically a DOM parser. The structure or
hierarchy of the JSON value is managed by the stack `ref_stack` which contains
a pointer to the respective array or object for each recursion depth.
After successful parsing, the value that is passed by reference to the
constructor contains the parsed value.
@tparam BasicJsonType the JSON type
*/
template<typename BasicJsonType>
class json_sax_dom_parser
{
public:
using number_integer_t = typename BasicJsonType::number_integer_t;
using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
using number_float_t = typename BasicJsonType::number_float_t;
using string_t = typename BasicJsonType::string_t;
using binary_t = typename BasicJsonType::binary_t;
/*!
@param[in, out] r reference to a JSON value that is manipulated while
parsing
@param[in] allow_exceptions_ whether parse errors yield exceptions
*/
explicit json_sax_dom_parser(BasicJsonType& r, const bool allow_exceptions_ = true)
: root(r), allow_exceptions(allow_exceptions_)
{}
// make class move-only
json_sax_dom_parser(const json_sax_dom_parser&) = delete;
json_sax_dom_parser(json_sax_dom_parser&&) = default;
json_sax_dom_parser& operator=(const json_sax_dom_parser&) = delete;
json_sax_dom_parser& operator=(json_sax_dom_parser&&) = default;
~json_sax_dom_parser() = default;
bool null()
{
handle_value(nullptr);
return true;
}
bool boolean(bool val)
{
handle_value(val);
return true;
}
bool number_integer(number_integer_t val)
{
handle_value(val);
return true;
}
bool number_unsigned(number_unsigned_t val)
{
handle_value(val);
return true;
}
bool number_float(number_float_t val, const string_t& /*unused*/)
{
handle_value(val);
return true;
}
bool string(string_t& val)
{
handle_value(val);
return true;
}
bool binary(binary_t& val)
{
handle_binary(val);
return true;
}
bool start_object(std::size_t len)
{
ref_stack.push_back(handle_value(BasicJsonType::value_t::object));
if (JSON_HEDLEY_UNLIKELY(len != std::size_t(-1) and len > ref_stack.back()->max_size()))
{
JSON_THROW(out_of_range::create(408,
"excessive object size: " + std::to_string(len)));
}
return true;
}
bool key(string_t& val)
{
// add null at given key and store the reference for later
object_element = &(ref_stack.back()->m_value.object->operator[](val));
return true;
}
bool end_object()
{
ref_stack.pop_back();
return true;
}
bool start_array(std::size_t len)
{
ref_stack.push_back(handle_value(BasicJsonType::value_t::array));
if (JSON_HEDLEY_UNLIKELY(len != std::size_t(-1) and len > ref_stack.back()->max_size()))
{
JSON_THROW(out_of_range::create(408,
"excessive array size: " + std::to_string(len)));
}
return true;
}
bool end_array()
{
ref_stack.pop_back();
return true;
}
bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/,
const detail::exception& ex)
{
errored = true;
if (allow_exceptions)
{
// determine the proper exception type from the id
switch ((ex.id / 100) % 100)
{
case 1:
JSON_THROW(*static_cast<const detail::parse_error*>(&ex));
case 4:
JSON_THROW(*static_cast<const detail::out_of_range*>(&ex));
// LCOV_EXCL_START
case 2:
JSON_THROW(*static_cast<const detail::invalid_iterator*>(&ex));
case 3:
JSON_THROW(*static_cast<const detail::type_error*>(&ex));
case 5:
JSON_THROW(*static_cast<const detail::other_error*>(&ex));
default:
assert(false);
// LCOV_EXCL_STOP
}
}
return false;
}
constexpr bool is_errored() const
{
return errored;
}
private:
/*!
@invariant If the ref stack is empty, then the passed value will be the new
root.
@invariant If the ref stack contains a value, then it is an array or an
object to which we can add elements
*/
template<typename Value>
JSON_HEDLEY_RETURNS_NON_NULL
BasicJsonType* handle_value(Value&& v)
{
if (ref_stack.empty())
{
root = BasicJsonType(std::forward<Value>(v));
return &root;
}
assert(ref_stack.back()->is_array() or ref_stack.back()->is_object());
if (ref_stack.back()->is_array())
{
ref_stack.back()->m_value.array->emplace_back(std::forward<Value>(v));
return &(ref_stack.back()->m_value.array->back());
}
assert(ref_stack.back()->is_object());
assert(object_element);
*object_element = BasicJsonType(std::forward<Value>(v));
return object_element;
}
/*!
@invariant If the ref stack is empty, then the passed value will be the new
root.
@invariant If the ref stack contains a value, then it is an array or an
object to which we can add elements
*/
template<typename BinaryValue>
JSON_HEDLEY_RETURNS_NON_NULL
BasicJsonType* handle_binary(BinaryValue&& v)
{
if (ref_stack.empty())
{
root = BasicJsonType::binary_array(std::forward<BinaryValue>(v));
return &root;
}
assert(ref_stack.back()->is_array() or ref_stack.back()->is_object());
if (ref_stack.back()->is_array())
{
ref_stack.back()->m_value.array->emplace_back(BasicJsonType::binary_array(std::forward<BinaryValue>(v)));
return &(ref_stack.back()->m_value.array->back());
}
assert(ref_stack.back()->is_object());
assert(object_element);
*object_element = BasicJsonType::binary_array(std::forward<BinaryValue>(v));
return object_element;
}
/// the parsed JSON value
BasicJsonType& root;
/// stack to model hierarchy of values
std::vector<BasicJsonType*> ref_stack {};
/// helper to hold the reference for the next object element
BasicJsonType* object_element = nullptr;
/// whether a syntax error occurred
bool errored = false;
/// whether to throw exceptions in case of errors
const bool allow_exceptions = true;
};
template<typename BasicJsonType>
class json_sax_dom_callback_parser
{
public:
using number_integer_t = typename BasicJsonType::number_integer_t;
using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
using number_float_t = typename BasicJsonType::number_float_t;
using string_t = typename BasicJsonType::string_t;
using binary_t = typename BasicJsonType::binary_t;
using parser_callback_t = typename BasicJsonType::parser_callback_t;
using parse_event_t = typename BasicJsonType::parse_event_t;
json_sax_dom_callback_parser(BasicJsonType& r,
const parser_callback_t cb,
const bool allow_exceptions_ = true)
: root(r), callback(cb), allow_exceptions(allow_exceptions_)
{
keep_stack.push_back(true);
}
// make class move-only
json_sax_dom_callback_parser(const json_sax_dom_callback_parser&) = delete;
json_sax_dom_callback_parser(json_sax_dom_callback_parser&&) = default;
json_sax_dom_callback_parser& operator=(const json_sax_dom_callback_parser&) = delete;
json_sax_dom_callback_parser& operator=(json_sax_dom_callback_parser&&) = default;
~json_sax_dom_callback_parser() = default;
bool null()
{
handle_value(nullptr);
return true;
}
bool boolean(bool val)
{
handle_value(val);
return true;
}
bool number_integer(number_integer_t val)
{
handle_value(val);
return true;
}
bool number_unsigned(number_unsigned_t val)
{
handle_value(val);
return true;
}
bool number_float(number_float_t val, const string_t& /*unused*/)
{
handle_value(val);
return true;
}
bool string(string_t& val)
{
handle_value(val);
return true;
}
bool binary(binary_t& val)
{
handle_value(val);
return true;
}
bool start_object(std::size_t len)
{
// check callback for object start
const bool keep = callback(static_cast<int>(ref_stack.size()), parse_event_t::object_start, discarded);
keep_stack.push_back(keep);
auto val = handle_value(BasicJsonType::value_t::object, true);
ref_stack.push_back(val.second);
// check object limit
if (ref_stack.back() and JSON_HEDLEY_UNLIKELY(len != std::size_t(-1) and len > ref_stack.back()->max_size()))
{
JSON_THROW(out_of_range::create(408, "excessive object size: " + std::to_string(len)));
}
return true;
}
bool key(string_t& val)
{
BasicJsonType k = BasicJsonType(val);
// check callback for key
const bool keep = callback(static_cast<int>(ref_stack.size()), parse_event_t::key, k);
key_keep_stack.push_back(keep);
// add discarded value at given key and store the reference for later
if (keep and ref_stack.back())
{
object_element = &(ref_stack.back()->m_value.object->operator[](val) = discarded);
}
return true;
}
bool end_object()
{
if (ref_stack.back() and not callback(static_cast<int>(ref_stack.size()) - 1, parse_event_t::object_end, *ref_stack.back()))
{
// discard object
*ref_stack.back() = discarded;
}
assert(not ref_stack.empty());
assert(not keep_stack.empty());
ref_stack.pop_back();
keep_stack.pop_back();
if (not ref_stack.empty() and ref_stack.back() and ref_stack.back()->is_object())
{
// remove discarded value
for (auto it = ref_stack.back()->begin(); it != ref_stack.back()->end(); ++it)
{
if (it->is_discarded())
{
ref_stack.back()->erase(it);
break;
}
}
}
return true;
}
bool start_array(std::size_t len)
{
const bool keep = callback(static_cast<int>(ref_stack.size()), parse_event_t::array_start, discarded);
keep_stack.push_back(keep);
auto val = handle_value(BasicJsonType::value_t::array, true);
ref_stack.push_back(val.second);
// check array limit
if (ref_stack.back() and JSON_HEDLEY_UNLIKELY(len != std::size_t(-1) and len > ref_stack.back()->max_size()))
{
JSON_THROW(out_of_range::create(408, "excessive array size: " + std::to_string(len)));
}
return true;
}
bool end_array()
{
bool keep = true;
if (ref_stack.back())
{
keep = callback(static_cast<int>(ref_stack.size()) - 1, parse_event_t::array_end, *ref_stack.back());
if (not keep)
{
// discard array
*ref_stack.back() = discarded;
}
}
assert(not ref_stack.empty());
assert(not keep_stack.empty());
ref_stack.pop_back();
keep_stack.pop_back();
// remove discarded value
if (not keep and not ref_stack.empty() and ref_stack.back()->is_array())
{
ref_stack.back()->m_value.array->pop_back();
}
return true;
}
bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/,
const detail::exception& ex)
{
errored = true;
if (allow_exceptions)
{
// determine the proper exception type from the id
switch ((ex.id / 100) % 100)
{
case 1:
JSON_THROW(*static_cast<const detail::parse_error*>(&ex));
case 4:
JSON_THROW(*static_cast<const detail::out_of_range*>(&ex));
// LCOV_EXCL_START
case 2:
JSON_THROW(*static_cast<const detail::invalid_iterator*>(&ex));
case 3:
JSON_THROW(*static_cast<const detail::type_error*>(&ex));
case 5:
JSON_THROW(*static_cast<const detail::other_error*>(&ex));
default:
assert(false);
// LCOV_EXCL_STOP
}
}
return false;
}
constexpr bool is_errored() const
{
return errored;
}
private:
/*!
@param[in] v value to add to the JSON value we build during parsing
@param[in] skip_callback whether we should skip calling the callback
function; this is required after start_array() and
start_object() SAX events, because otherwise we would call the
callback function with an empty array or object, respectively.
@invariant If the ref stack is empty, then the passed value will be the new
root.
@invariant If the ref stack contains a value, then it is an array or an
object to which we can add elements
@return pair of boolean (whether value should be kept) and pointer (to the
passed value in the ref_stack hierarchy; nullptr if not kept)
*/
template<typename Value>
std::pair<bool, BasicJsonType*> handle_value(Value&& v, const bool skip_callback = false)
{
assert(not keep_stack.empty());
// do not handle this value if we know it would be added to a discarded
// container
if (not keep_stack.back())
{
return {false, nullptr};
}
// create value
auto value = BasicJsonType(std::forward<Value>(v));
// check callback
const bool keep = skip_callback or callback(static_cast<int>(ref_stack.size()), parse_event_t::value, value);
// do not handle this value if we just learnt it shall be discarded
if (not keep)
{
return {false, nullptr};
}
if (ref_stack.empty())
{
root = std::move(value);
return {true, &root};
}
// skip this value if we already decided to skip the parent
// (https://github.com/nlohmann/json/issues/971#issuecomment-413678360)
if (not ref_stack.back())
{
return {false, nullptr};
}
// we now only expect arrays and objects
assert(ref_stack.back()->is_array() or ref_stack.back()->is_object());
// array
if (ref_stack.back()->is_array())
{
ref_stack.back()->m_value.array->push_back(std::move(value));
return {true, &(ref_stack.back()->m_value.array->back())};
}
// object
assert(ref_stack.back()->is_object());
// check if we should store an element for the current key
assert(not key_keep_stack.empty());
const bool store_element = key_keep_stack.back();
key_keep_stack.pop_back();
if (not store_element)
{
return {false, nullptr};
}
assert(object_element);
*object_element = std::move(value);
return {true, object_element};
}
/// the parsed JSON value
BasicJsonType& root;
/// stack to model hierarchy of values
std::vector<BasicJsonType*> ref_stack {};
/// stack to manage which values to keep
std::vector<bool> keep_stack {};
/// stack to manage which object keys to keep
std::vector<bool> key_keep_stack {};
/// helper to hold the reference for the next object element
BasicJsonType* object_element = nullptr;
/// whether a syntax error occurred
bool errored = false;
/// callback function
const parser_callback_t callback = nullptr;
/// whether to throw exceptions in case of errors
const bool allow_exceptions = true;
/// a discarded value for the callback
BasicJsonType discarded = BasicJsonType::value_t::discarded;
};
template<typename BasicJsonType>
class json_sax_acceptor
{
public:
using number_integer_t = typename BasicJsonType::number_integer_t;
using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
using number_float_t = typename BasicJsonType::number_float_t;
using string_t = typename BasicJsonType::string_t;
using binary_t = typename BasicJsonType::binary_t;
bool null()
{
return true;
}
bool boolean(bool /*unused*/)
{
return true;
}
bool number_integer(number_integer_t /*unused*/)
{
return true;
}
bool number_unsigned(number_unsigned_t /*unused*/)
{
return true;
}
bool number_float(number_float_t /*unused*/, const string_t& /*unused*/)
{
return true;
}
bool string(string_t& /*unused*/)
{
return true;
}
bool binary(binary_t& /*unused*/)
{
return true;
}
bool start_object(std::size_t /*unused*/ = std::size_t(-1))
{
return true;
}
bool key(string_t& /*unused*/)
{
return true;
}
bool end_object()
{
return true;
}
bool start_array(std::size_t /*unused*/ = std::size_t(-1))
{
return true;
}
bool end_array()
{
return true;
}
bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/, const detail::exception& /*unused*/)
{
return false;
}
};
} // namespace detail
} // namespace nlohmann