1
0
mirror of https://github.com/nlohmann/json synced 2024-12-11 12:10:08 +00:00
nlohmannjson/develop/detail/serializer.hpp

608 lines
22 KiB
C++

#pragma once
#include <algorithm> // reverse, remove, fill, find, none_of
#include <array> // array
#include <cassert> // assert
#include <ciso646> // and, or
#include <clocale> // localeconv, lconv
#include <cmath> // labs, isfinite, isnan, signbit
#include <cstddef> // size_t, ptrdiff_t
#include <cstdint> // uint8_t
#include <cstdio> // snprintf
#include <iterator> // next
#include <limits> // numeric_limits
#include <string> // string
#include <type_traits> // is_same
#include "detail/macro_scope.hpp"
#include "detail/meta.hpp"
#include "detail/parsing/output_adapters.hpp"
#include "detail/value_t.hpp"
namespace nlohmann
{
namespace detail
{
///////////////////
// serialization //
///////////////////
template<typename BasicJsonType>
class serializer
{
using string_t = typename BasicJsonType::string_t;
using number_float_t = typename BasicJsonType::number_float_t;
using number_integer_t = typename BasicJsonType::number_integer_t;
using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
static constexpr uint8_t UTF8_ACCEPT = 0;
static constexpr uint8_t UTF8_REJECT = 1;
public:
/*!
@param[in] s output stream to serialize to
@param[in] ichar indentation character to use
*/
serializer(output_adapter_t<char> s, const char ichar)
: o(std::move(s)), loc(std::localeconv()),
thousands_sep(loc->thousands_sep == nullptr ? '\0' : * (loc->thousands_sep)),
decimal_point(loc->decimal_point == nullptr ? '\0' : * (loc->decimal_point)),
indent_char(ichar), indent_string(512, indent_char)
{}
// delete because of pointer members
serializer(const serializer&) = delete;
serializer& operator=(const serializer&) = delete;
/*!
@brief internal implementation of the serialization function
This function is called by the public member function dump and organizes
the serialization internally. The indentation level is propagated as
additional parameter. In case of arrays and objects, the function is
called recursively.
- strings and object keys are escaped using `escape_string()`
- integer numbers are converted implicitly via `operator<<`
- floating-point numbers are converted to a string using `"%g"` format
@param[in] val value to serialize
@param[in] pretty_print whether the output shall be pretty-printed
@param[in] indent_step the indent level
@param[in] current_indent the current indent level (only used internally)
*/
void dump(const BasicJsonType& val, const bool pretty_print,
const bool ensure_ascii,
const unsigned int indent_step,
const unsigned int current_indent = 0)
{
switch (val.m_type)
{
case value_t::object:
{
if (val.m_value.object->empty())
{
o->write_characters("{}", 2);
return;
}
if (pretty_print)
{
o->write_characters("{\n", 2);
// variable to hold indentation for recursive calls
const auto new_indent = current_indent + indent_step;
if (JSON_UNLIKELY(indent_string.size() < new_indent))
{
indent_string.resize(indent_string.size() * 2, ' ');
}
// first n-1 elements
auto i = val.m_value.object->cbegin();
for (std::size_t cnt = 0; cnt < val.m_value.object->size() - 1; ++cnt, ++i)
{
o->write_characters(indent_string.c_str(), new_indent);
o->write_character('\"');
dump_escaped(i->first, ensure_ascii);
o->write_characters("\": ", 3);
dump(i->second, true, ensure_ascii, indent_step, new_indent);
o->write_characters(",\n", 2);
}
// last element
assert(i != val.m_value.object->cend());
assert(std::next(i) == val.m_value.object->cend());
o->write_characters(indent_string.c_str(), new_indent);
o->write_character('\"');
dump_escaped(i->first, ensure_ascii);
o->write_characters("\": ", 3);
dump(i->second, true, ensure_ascii, indent_step, new_indent);
o->write_character('\n');
o->write_characters(indent_string.c_str(), current_indent);
o->write_character('}');
}
else
{
o->write_character('{');
// first n-1 elements
auto i = val.m_value.object->cbegin();
for (std::size_t cnt = 0; cnt < val.m_value.object->size() - 1; ++cnt, ++i)
{
o->write_character('\"');
dump_escaped(i->first, ensure_ascii);
o->write_characters("\":", 2);
dump(i->second, false, ensure_ascii, indent_step, current_indent);
o->write_character(',');
}
// last element
assert(i != val.m_value.object->cend());
assert(std::next(i) == val.m_value.object->cend());
o->write_character('\"');
dump_escaped(i->first, ensure_ascii);
o->write_characters("\":", 2);
dump(i->second, false, ensure_ascii, indent_step, current_indent);
o->write_character('}');
}
return;
}
case value_t::array:
{
if (val.m_value.array->empty())
{
o->write_characters("[]", 2);
return;
}
if (pretty_print)
{
o->write_characters("[\n", 2);
// variable to hold indentation for recursive calls
const auto new_indent = current_indent + indent_step;
if (JSON_UNLIKELY(indent_string.size() < new_indent))
{
indent_string.resize(indent_string.size() * 2, ' ');
}
// first n-1 elements
for (auto i = val.m_value.array->cbegin();
i != val.m_value.array->cend() - 1; ++i)
{
o->write_characters(indent_string.c_str(), new_indent);
dump(*i, true, ensure_ascii, indent_step, new_indent);
o->write_characters(",\n", 2);
}
// last element
assert(not val.m_value.array->empty());
o->write_characters(indent_string.c_str(), new_indent);
dump(val.m_value.array->back(), true, ensure_ascii, indent_step, new_indent);
o->write_character('\n');
o->write_characters(indent_string.c_str(), current_indent);
o->write_character(']');
}
else
{
o->write_character('[');
// first n-1 elements
for (auto i = val.m_value.array->cbegin();
i != val.m_value.array->cend() - 1; ++i)
{
dump(*i, false, ensure_ascii, indent_step, current_indent);
o->write_character(',');
}
// last element
assert(not val.m_value.array->empty());
dump(val.m_value.array->back(), false, ensure_ascii, indent_step, current_indent);
o->write_character(']');
}
return;
}
case value_t::string:
{
o->write_character('\"');
dump_escaped(*val.m_value.string, ensure_ascii);
o->write_character('\"');
return;
}
case value_t::boolean:
{
if (val.m_value.boolean)
{
o->write_characters("true", 4);
}
else
{
o->write_characters("false", 5);
}
return;
}
case value_t::number_integer:
{
dump_integer(val.m_value.number_integer);
return;
}
case value_t::number_unsigned:
{
dump_integer(val.m_value.number_unsigned);
return;
}
case value_t::number_float:
{
dump_float(val.m_value.number_float);
return;
}
case value_t::discarded:
{
o->write_characters("<discarded>", 11);
return;
}
case value_t::null:
{
o->write_characters("null", 4);
return;
}
}
}
private:
/*!
@brief dump escaped string
Escape a string by replacing certain special characters by a sequence of an
escape character (backslash) and another character and other control
characters by a sequence of "\u" followed by a four-digit hex
representation. The escaped string is written to output stream @a o.
@param[in] s the string to escape
@param[in] ensure_ascii whether to escape non-ASCII characters with
\uXXXX sequences
@complexity Linear in the length of string @a s.
*/
void dump_escaped(const string_t& s, const bool ensure_ascii)
{
uint32_t codepoint;
uint8_t state = UTF8_ACCEPT;
std::size_t bytes = 0; // number of bytes written to string_buffer
for (std::size_t i = 0; i < s.size(); ++i)
{
const auto byte = static_cast<uint8_t>(s[i]);
switch (decode(state, codepoint, byte))
{
case UTF8_ACCEPT: // decode found a new code point
{
switch (codepoint)
{
case 0x08: // backspace
{
string_buffer[bytes++] = '\\';
string_buffer[bytes++] = 'b';
break;
}
case 0x09: // horizontal tab
{
string_buffer[bytes++] = '\\';
string_buffer[bytes++] = 't';
break;
}
case 0x0A: // newline
{
string_buffer[bytes++] = '\\';
string_buffer[bytes++] = 'n';
break;
}
case 0x0C: // formfeed
{
string_buffer[bytes++] = '\\';
string_buffer[bytes++] = 'f';
break;
}
case 0x0D: // carriage return
{
string_buffer[bytes++] = '\\';
string_buffer[bytes++] = 'r';
break;
}
case 0x22: // quotation mark
{
string_buffer[bytes++] = '\\';
string_buffer[bytes++] = '\"';
break;
}
case 0x5C: // reverse solidus
{
string_buffer[bytes++] = '\\';
string_buffer[bytes++] = '\\';
break;
}
default:
{
// escape control characters (0x00..0x1F) or, if
// ensure_ascii parameter is used, non-ASCII characters
if ((codepoint <= 0x1F) or (ensure_ascii and (codepoint >= 0x7F)))
{
if (codepoint <= 0xFFFF)
{
std::snprintf(string_buffer.data() + bytes, 7, "\\u%04x", codepoint);
bytes += 6;
}
else
{
std::snprintf(string_buffer.data() + bytes, 13, "\\u%04x\\u%04x",
(0xD7C0 + (codepoint >> 10)),
(0xDC00 + (codepoint & 0x3FF)));
bytes += 12;
}
}
else
{
// copy byte to buffer (all previous bytes
// been copied have in default case above)
string_buffer[bytes++] = s[i];
}
break;
}
}
// write buffer and reset index; there must be 13 bytes
// left, as this is the maximal number of bytes to be
// written ("\uxxxx\uxxxx\0") for one code point
if (string_buffer.size() - bytes < 13)
{
o->write_characters(string_buffer.data(), bytes);
bytes = 0;
}
break;
}
case UTF8_REJECT: // decode found invalid UTF-8 byte
{
std::stringstream ss;
ss << std::setw(2) << std::uppercase << std::setfill('0') << std::hex << static_cast<int>(byte);
JSON_THROW(type_error::create(316, "invalid UTF-8 byte at index " + std::to_string(i) + ": 0x" + ss.str()));
}
default: // decode found yet incomplete multi-byte code point
{
if (not ensure_ascii)
{
// code point will not be escaped - copy byte to buffer
string_buffer[bytes++] = s[i];
}
break;
}
}
}
if (JSON_LIKELY(state == UTF8_ACCEPT))
{
// write buffer
if (bytes > 0)
{
o->write_characters(string_buffer.data(), bytes);
}
}
else
{
// we finish reading, but do not accept: string was incomplete
std::stringstream ss;
ss << std::setw(2) << std::uppercase << std::setfill('0') << std::hex << static_cast<int>(static_cast<uint8_t>(s.back()));
JSON_THROW(type_error::create(316, "incomplete UTF-8 string; last byte: 0x" + ss.str()));
}
}
/*!
@brief dump an integer
Dump a given integer to output stream @a o. Works internally with
@a number_buffer.
@param[in] x integer number (signed or unsigned) to dump
@tparam NumberType either @a number_integer_t or @a number_unsigned_t
*/
template<typename NumberType, detail::enable_if_t<
std::is_same<NumberType, number_unsigned_t>::value or
std::is_same<NumberType, number_integer_t>::value,
int> = 0>
void dump_integer(NumberType x)
{
// special case for "0"
if (x == 0)
{
o->write_character('0');
return;
}
const bool is_negative = (x <= 0) and (x != 0); // see issue #755
std::size_t i = 0;
while (x != 0)
{
// spare 1 byte for '\0'
assert(i < number_buffer.size() - 1);
const auto digit = std::labs(static_cast<long>(x % 10));
number_buffer[i++] = static_cast<char>('0' + digit);
x /= 10;
}
if (is_negative)
{
// make sure there is capacity for the '-'
assert(i < number_buffer.size() - 2);
number_buffer[i++] = '-';
}
std::reverse(number_buffer.begin(), number_buffer.begin() + i);
o->write_characters(number_buffer.data(), i);
}
/*!
@brief dump a floating-point number
Dump a given floating-point number to output stream @a o. Works internally
with @a number_buffer.
@param[in] x floating-point number to dump
*/
void dump_float(number_float_t x)
{
// NaN / inf
if (not std::isfinite(x) or std::isnan(x))
{
o->write_characters("null", 4);
return;
}
// get number of digits for a text -> float -> text round-trip
static constexpr auto d = std::numeric_limits<number_float_t>::digits10;
// the actual conversion
std::ptrdiff_t len = snprintf(number_buffer.data(), number_buffer.size(), "%.*g", d, x);
// negative value indicates an error
assert(len > 0);
// check if buffer was large enough
assert(static_cast<std::size_t>(len) < number_buffer.size());
// erase thousands separator
if (thousands_sep != '\0')
{
const auto end = std::remove(number_buffer.begin(),
number_buffer.begin() + len, thousands_sep);
std::fill(end, number_buffer.end(), '\0');
assert((end - number_buffer.begin()) <= len);
len = (end - number_buffer.begin());
}
// convert decimal point to '.'
if (decimal_point != '\0' and decimal_point != '.')
{
const auto dec_pos = std::find(number_buffer.begin(), number_buffer.end(), decimal_point);
if (dec_pos != number_buffer.end())
{
*dec_pos = '.';
}
}
o->write_characters(number_buffer.data(), static_cast<std::size_t>(len));
// determine if need to append ".0"
const bool value_is_int_like =
std::none_of(number_buffer.begin(), number_buffer.begin() + len + 1,
[](char c)
{
return (c == '.' or c == 'e');
});
if (value_is_int_like)
{
o->write_characters(".0", 2);
}
}
/*!
@brief check whether a string is UTF-8 encoded
The function checks each byte of a string whether it is UTF-8 encoded. The
result of the check is stored in the @a state parameter. The function must
be called initially with state 0 (accept). State 1 means the string must
be rejected, because the current byte is not allowed. If the string is
completely processed, but the state is non-zero, the string ended
prematurely; that is, the last byte indicated more bytes should have
followed.
@param[in,out] state the state of the decoding
@param[in,out] codep codepoint (valid only if resulting state is UTF8_ACCEPT)
@param[in] byte next byte to decode
@return new state
@note The function has been edited: a std::array is used.
@copyright Copyright (c) 2008-2009 Bjoern Hoehrmann <bjoern@hoehrmann.de>
@sa http://bjoern.hoehrmann.de/utf-8/decoder/dfa/
*/
static uint8_t decode(uint8_t& state, uint32_t& codep, const uint8_t byte) noexcept
{
static const std::array<uint8_t, 400> utf8d =
{
{
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 00..1F
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 20..3F
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 40..5F
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 60..7F
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, // 80..9F
7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, // A0..BF
8, 8, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, // C0..DF
0xA, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x4, 0x3, 0x3, // E0..EF
0xB, 0x6, 0x6, 0x6, 0x5, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, // F0..FF
0x0, 0x1, 0x2, 0x3, 0x5, 0x8, 0x7, 0x1, 0x1, 0x1, 0x4, 0x6, 0x1, 0x1, 0x1, 0x1, // s0..s0
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, // s1..s2
1, 2, 1, 1, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, // s3..s4
1, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 1, 3, 1, 1, 1, 1, 1, 1, // s5..s6
1, 3, 1, 1, 1, 1, 1, 3, 1, 3, 1, 1, 1, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 // s7..s8
}
};
const uint8_t type = utf8d[byte];
codep = (state != UTF8_ACCEPT)
? (byte & 0x3fu) | (codep << 6)
: (0xff >> type) & (byte);
state = utf8d[256u + state * 16u + type];
return state;
}
private:
/// the output of the serializer
output_adapter_t<char> o = nullptr;
/// a (hopefully) large enough character buffer
std::array<char, 64> number_buffer{{}};
/// the locale
const std::lconv* loc = nullptr;
/// the locale's thousand separator character
const char thousands_sep = '\0';
/// the locale's decimal point character
const char decimal_point = '\0';
/// string buffer
std::array<char, 512> string_buffer{{}};
/// the indentation character
const char indent_char;
/// the indentation string
string_t indent_string;
};
}
}