amorfo77/json
1
0 Fork 0
mirror of https://github.com/nlohmann/json.git synced 2026-10-22 22:08:55 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
272d07b48bd7542e615bc734cc89120b573bba52
json/include/nlohmann/detail/view/serializer.hpp
T
Niels Lohmann 272d07b48b Scan the strings of json_view's dump() with ensure_ascii 16 bytes at a time 
With ensure_ascii, a string of the source that has a byte >= 0x7F was written
by a generic path: put() per piece and a validate-and-decode step for every
character beyond ASCII. It now goes through a writer with the cursor in a
local variable. The part of the string that has nothing to escape is found
16 bytes at a time (SSE2 or NEON; the existing 8-byte scan without vector
support) and copied; strings that are ASCII throughout are still decided by
the 8-byte check of dump() with ensure_ascii (a 16-byte scan for them was
2-5 % slower for short strings). The two-byte sequences (lead C2..DF) and
the three-byte sequences whose continuation bytes may have any value (lead
E1..EF except ED) are decoded with a few compares; E0, ED, four-byte
sequences, and damaged input go through validate_one_utf8() as before, so
damaged images still throw type_error.316.

dump(-1, ' ', true) instructions per call, macOS arm64 (develop -> now);
the compact dump without ensure_ascii and dumps of ASCII text are unchanged:

  twitter 5.42M -> 3.57M    citm 3.28M -> 3.24M     jeopardy(4k) 5.78M -> 5.65M
  CJK text, raw 66.5M -> 35.3M    escaped \u 32.7M -> 17.0M
  Latin text with accents 11.5M -> 9.4M    ASCII text 4.40M -> 4.40M

Linux instructions (GCC 13 / Clang 18), dump(-1, ' ', true):

  twitter -9 % / -33 %    CJK text -14 % / -47 %    Latin text -8 % / -18 %
  citm -0.2 % / -0.4 %    ASCII text -0.1 % / -0.3 %

The compact dump without ensure_ascii is unchanged (canada 0.0 %).

The unit test covers every character of the Basic Multilingual Plane, and a
character that needs an escape at each position of strings of 0..40 bytes.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:30:36 +02:00

1145 lines
41 KiB
C++
Raw Blame History

// __ _____ _____ _____
// __| | __| | | | JSON for Modern C++
// | | |__ | | | | | | version 3.12.0
// |_____|_____|_____|_|___| https://github.com/nlohmann/json
//
// SPDX-FileCopyrightText: 2013-2026 Niels Lohmann <https://nlohmann.me>
// SPDX-License-Identifier: MIT
#pragma once
#include <algorithm> // max
#include <array> // array
#include <cmath> // isfinite
#include <cstddef> // size_t
#include <cstdint> // uint8_t, uint32_t
#include <cstring> // memcpy, memset
#include <limits> // numeric_limits
#include <type_traits> // integral_constant
#include <vector> // vector
#include <nlohmann/json.hpp>
#include <nlohmann/detail/view/document_data.hpp>
#include <nlohmann/detail/view/macro_scope.hpp>
#include <nlohmann/detail/view/node.hpp>
#include <nlohmann/detail/view/number.hpp>
#include <nlohmann/detail/view/simd.hpp>
NLOHMANN_JSON_NAMESPACE_BEGIN
namespace detail
{
namespace view
{
/// append-only output buffer: writes through a raw pointer into a string that
/// is resized ahead, and trimmed by finish(). The estimate is reserved, and the
/// string grows in steps of 64 KiB within it: resize() fills the new bytes
/// with zeros (before C++23, a string cannot grow without), and a small step
/// is filled while the writer is about to use it, in the cache, instead of
/// filling the whole estimate in memory first.
template<typename StringType>
class output_buffer
{
public:
output_buffer(StringType& out, std::size_t estimate)
: m_out(sized(out, estimate))
, m_pos(&m_out[0])
, m_end(m_pos + m_out.size())
{}
void finish()
{
const auto size = static_cast<std::size_t>(m_pos - m_out.data());
m_out.resize(size);
}
NLOHMANN_VIEW_ALWAYS_INLINE void reserve(std::size_t n)
{
if (NLOHMANN_VIEW_UNLIKELY(static_cast<std::size_t>(m_end - m_pos) < n))
{
grow(n);
}
}
NLOHMANN_VIEW_ALWAYS_INLINE void put(char c)
{
reserve(1);
*m_pos++ = c;
}
NLOHMANN_VIEW_ALWAYS_INLINE void put(const char* s, std::size_t n)
{
reserve(n);
std::memcpy(m_pos, s, n);
m_pos += n;
}
/// a string literal, without its terminating NUL
template<std::size_t N>
NLOHMANN_VIEW_ALWAYS_INLINE void put(const char (&s)[N]) // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
{
put(s, N - 1);
}
/// the write position and the end of the writable space, for a writer
/// that keeps the position in a local variable (set_cursor() hands it back)
char* cursor() const noexcept
{
return m_pos;
}
char* limit() const noexcept
{
return m_end;
}
void set_cursor(char* p) noexcept
{
m_pos = p;
}
private:
/// the size of a growth step
static constexpr std::size_t step = 0x10000;
static StringType& sized(StringType& out, std::size_t estimate)
{
out.reserve(estimate);
out.resize(detail::clamp(estimate, std::size_t{64}, step));
return out;
}
NLOHMANN_VIEW_NOINLINE void grow(std::size_t n)
{
const auto used = static_cast<std::size_t>(m_pos - m_out.data());
// (a step does not go beyond the reserved estimate, so that a good
// estimate is never copied to a larger allocation; not a clamp: the
// lower bound used + n + 256 can exceed capacity(), and then it must win)
const std::size_t size = (std::max)((std::min)(m_out.size() + step, m_out.capacity()), used + n + 256);
if (size > m_out.capacity())
{
m_out.reserve((std::max)(m_out.capacity() * 2, size));
}
m_out.resize(size);
m_pos = &m_out[0] + used;
m_end = &m_out[0] + m_out.size();
}
StringType& m_out;
char* m_pos;
char* m_end;
};
/// The length of the run at s that dump() writes unchanged without
/// ensure_ascii: all bytes but quotes, backslashes, and control characters.
/// Unlike detail::string_bulk_run(), non-ASCII bytes are not validated: the
/// strings of a document are valid UTF-8 (a damaged image loaded with
/// image_check::bounds can have others, which are then written unchanged).
inline std::size_t plain_output_run(const unsigned char* s, std::size_t n) noexcept
{
constexpr std::uint64_t ones = 0x0101010101010101ull;
constexpr std::uint64_t high = 0x8080808080808080ull;
std::size_t i = 0;
for (; i + 8 <= n; i += 8)
{
const std::uint64_t v = read_eight_bytes(s + i);
const std::uint64_t q = v ^ 0x2222222222222222ull; // '"'
const std::uint64_t b = v ^ 0x5C5C5C5C5C5C5C5Cull; // '\\'
const std::uint64_t stop = (((q - ones) & ~q) | ((b - ones) & ~b) | ((v - 0x2020202020202020ull) & ~v)) & high;
if (stop != 0)
{
// the lowest flagged byte is the first stop: borrows only flag bytes above a true one
return i + (static_cast<std::size_t>(count_trailing_zeros(stop)) / 8);
}
}
for (; i < n; ++i)
{
if (s[i] == '"' || s[i] == '\\' || s[i] < 0x20)
{
return i;
}
}
return n;
}
/// Whether dump() with ensure_ascii writes the string at s unchanged, for a
/// string of the source without escape sequences (which has no quote,
/// backslash, or control character): no byte is DEL (0x7F) or above. Adding 1
/// to each byte sets its high bit exactly for 0x7F, and a carry into the next
/// byte only comes from 0xFF, whose own high bit is set anyway.
inline bool is_ascii_without_delete(const char* s, std::size_t n) noexcept
{
constexpr std::uint64_t ones = 0x0101010101010101ull;
constexpr std::uint64_t high = 0x8080808080808080ull;
std::size_t i = 0;
for (; i + 8 <= n; i += 8)
{
const std::uint64_t v = read_eight_bytes(s + i);
if (((v | (v + ones)) & high) != 0)
{
return false;
}
}
for (; i < n; ++i)
{
if (static_cast<unsigned char>(s[i]) >= 0x7F)
{
return false;
}
}
return true;
}
#if NLOHMANN_VIEW_VECTOR
/// the bytes of the 16 at p that ensure_ascii cannot copy (a quote, a
/// backslash, a control character, DEL, or not ASCII; a signed compare with
/// 0x20 finds control characters and non-ASCII bytes at once), one bit per
/// byte (x86) or one nibble per byte (NEON)
NLOHMANN_VIEW_ALWAYS_INLINE std::uint64_t ascii_stop_mask(const unsigned char* p) noexcept
{
#if NLOHMANN_VIEW_NEON
const uint8x16_t in = vld1q_u8(p);
const uint8x16_t special = vorrq_u8(vorrq_u8(vceqq_u8(in, vdupq_n_u8('"')), vceqq_u8(in, vdupq_n_u8('\\'))),
vorrq_u8(vceqq_u8(in, vdupq_n_u8(0x7F)), vcltq_s8(vreinterpretq_s8_u8(in), vdupq_n_s8(0x20))));
return vget_lane_u64(vreinterpret_u64_u8(vshrn_n_u16(vreinterpretq_u16_u8(special), 4)), 0);
#else
const __m128i in = _mm_loadu_si128(static_cast<const __m128i*>(static_cast<const void*>(p))); // NOLINT(bugprone-casting-through-void)
const __m128i special = _mm_or_si128(_mm_or_si128(_mm_cmpeq_epi8(in, _mm_set1_epi8('"')), _mm_cmpeq_epi8(in, _mm_set1_epi8('\\'))),
_mm_or_si128(_mm_cmpeq_epi8(in, _mm_set1_epi8(0x7F)), _mm_cmplt_epi8(in, _mm_set1_epi8(0x20))));
return static_cast<std::uint64_t>(static_cast<unsigned>(_mm_movemask_epi8(special)));
#endif
}
/// the width in bits of a byte in the mask of ascii_stop_mask()
constexpr unsigned ascii_stop_bits() noexcept
{
return NLOHMANN_VIEW_NEON ? 4u : 1u;
}
#endif
/// As find_ascii_copyable_run(), 16 bytes per step. end is the end of the
/// buffer that holds the n bytes at p (a short run is read as 16 bytes if
/// they are in the buffer).
inline std::size_t ascii_copyable_run(const unsigned char* p, std::size_t n, const unsigned char* end) noexcept
{
#if NLOHMANN_VIEW_VECTOR
if (n >= 16)
{
std::size_t i = 0;
for (; i + 16 <= n; i += 16)
{
const std::uint64_t mask = ascii_stop_mask(p + i);
if (mask != 0)
{
return i + (static_cast<std::size_t>(count_trailing_zeros(mask)) / ascii_stop_bits());
}
}
if (i == n)
{
return n;
}
// the last 16 bytes: those before i are known to be copyable
const std::uint64_t mask = ascii_stop_mask(p + n - 16);
return mask != 0 ? n - 16 + (static_cast<std::size_t>(count_trailing_zeros(mask)) / ascii_stop_bits()) : n;
}
if (end - p >= 16)
{
// (the bytes after the run are masked out)
const std::uint64_t mask = ascii_stop_mask(p) & ((std::uint64_t{1} << (n * ascii_stop_bits())) - 1u);
return mask != 0 ? static_cast<std::size_t>(count_trailing_zeros(mask)) / ascii_stop_bits() : n;
}
#else
static_cast<void>(end);
#endif
return find_ascii_copyable_run(p, n);
}
/// A stack that starts in a buffer of the caller (a local array) and moves to
/// the heap (a vector of the caller) only when that is full, so that dumps of
/// shallow documents need no allocation. The top is a pointer, as in
/// std::vector. grow() is static and out of line: it gets the vector and the
/// used range and returns the new storage, so the address of the stack never
/// reaches a call. Otherwise the compiler would have to keep the three
/// pointers in memory and load and store them on every push_back() and
/// pop_back(), not just when growing.
template<typename T>
class small_stack
{
public:
small_stack(T* buffer, std::size_t capacity, std::vector<T>& heap) noexcept
: m_begin(buffer), m_top(buffer), m_end(buffer + capacity), m_heap(heap)
{}
small_stack(const small_stack&) = delete;
small_stack(small_stack&&) = delete;
small_stack& operator=(const small_stack&) = delete;
small_stack& operator=(small_stack&&) = delete;
~small_stack() = default;
NLOHMANN_VIEW_ALWAYS_INLINE void push_back(const T& x)
{
if (NLOHMANN_VIEW_UNLIKELY(m_top == m_end))
{
const std::size_t used = size();
const std::size_t capacity = 2 * static_cast<std::size_t>(m_end - m_begin);
m_begin = grow(m_heap, m_begin, used, capacity);
m_top = m_begin + used;
m_end = m_begin + capacity;
}
*m_top++ = x;
}
NLOHMANN_VIEW_ALWAYS_INLINE T& back() noexcept
{
return m_top[-1];
}
NLOHMANN_VIEW_ALWAYS_INLINE void pop_back() noexcept
{
--m_top;
}
NLOHMANN_VIEW_ALWAYS_INLINE bool empty() const noexcept
{
return m_top == m_begin;
}
NLOHMANN_VIEW_ALWAYS_INLINE std::size_t size() const noexcept
{
return static_cast<std::size_t>(m_top - m_begin);
}
private:
/// the used entries moved to heap storage of the given capacity
NLOHMANN_VIEW_NOINLINE static T* grow(std::vector<T>& heap, const T* begin, std::size_t used, std::size_t capacity)
{
std::vector<T> bigger(capacity);
std::copy(begin, begin + used, bigger.begin());
heap.swap(bigger);
return heap.data();
}
T* m_begin;
T* m_top;
T* m_end;
std::vector<T>& m_heap;
};
/// a string literal at w, without its terminating NUL (after room())
/// @return a pointer past it
template<std::size_t N>
NLOHMANN_VIEW_ALWAYS_INLINE char* put_literal(char* w, const char (&s)[N]) noexcept // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
{
std::memcpy(w, s, N - 1);
return w + N - 1;
}
/// the depth of the innermost open container, counted only for indented output
template<bool Counted>
struct depth_counter
{
std::size_t value = 0;
void open() noexcept
{
++value;
}
std::size_t close() noexcept
{
return --value;
}
std::size_t get() const noexcept
{
return value;
}
};
template<>
struct depth_counter<false>
{
void open() noexcept {}
std::size_t close() noexcept
{
return 0;
}
std::size_t get() const noexcept
{
return 0;
}
};
/// how the view's dump() writes a value
struct dump_style
{
bool pretty = false; ///< indent >= 0
std::size_t indent = 0; ///< characters per level
char indent_char = ' ';
bool ensure_ascii = false;
bool source_numbers = false; ///< copy number tokens from the source
};
/*!
@brief write a view's subtree as basic_json::dump() writes the value
The output of a subtree equals ordered_json::parse(text).dump() of it for
the same arguments (members in document order): strings are escaped by the
same rules, with the library's scanning kernels; floats are written with
the library's conversion; integers are copied from the source, where they
are canonical (except "-0", which parse() reads as 0). The walk is
iterative, so the nesting depth is limited by memory only.
*/
template<typename BasicJsonType, bool Editable>
class view_serializer
{
using nav = navigation<Editable>;
using string_t = typename BasicJsonType::string_t;
using number_float_t = typename BasicJsonType::number_float_t;
/// doubles that are IEEE 754 binary64 are written from their bits (Zmij)
using fast_double = std::integral_constant < bool, std::is_same<number_float_t, double>::value && ::nlohmann::detail::dtoa_impl::is_binary64<double>::value >;
public:
view_serializer(const document_data& d, string_t& out, std::size_t estimate, const dump_style& style)
: m_doc(d), m_out(out, estimate), m_style(style)
{}
void dump(const node* root)
{
if (m_style.ensure_ascii)
{
dump_ascii(root);
}
else if (m_style.pretty)
{
dump_pretty(root);
}
else if (m_style.source_numbers)
{
dump_text<true, false, false>(root);
}
else
{
dump_text<false, false, false>(root);
}
}
private:
/// a string that is not a plain string of the source (decoded, or written
/// by an edit), without ensure_ascii: runs without characters to escape
/// are copied. If validate is set, the string must be valid UTF-8: the
/// runs are found by the lexer's scan, which validates the sequences, and
/// invalid UTF-8 throws what basic_json::dump() throws. (The compact output
/// writes the bytes of a damaged image unchanged, and scans without
/// validating them.)
NLOHMANN_VIEW_NOINLINE void write_decoded(const node& n, bool validate)
{
const auto* const s = reinterpret_cast<const unsigned char*>(m_doc.str(n)); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
m_out.put('"');
for (std::size_t i = 0; i < n.len;)
{
const std::size_t run = validate ? string_bulk_run(s + i, n.len - i) : plain_output_run(s + i, n.len - i);
if (run != 0)
{
m_out.put(reinterpret_cast<const char*>(s + i), run); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
i += run;
continue;
}
if (validate && s[i] >= 0x80)
{
invalid_utf8(s, n.len); // (the scan stopped at ill-formed UTF-8)
}
write_codepoint<false>(s[i], s + i, 1); // a quote, a backslash, or a control character
++i;
}
m_out.put('"');
}
/// dump_text() with indentation, out of line: the compact walk is faster
/// when it is not merged with the indented ones into dump()
NLOHMANN_VIEW_NOINLINE void dump_pretty(const node* root)
{
if (m_style.source_numbers)
{
dump_text<true, true, false>(root);
}
else
{
dump_text<false, true, false>(root);
}
}
/// dump_text() with ensure_ascii, out of line: the walk without it is
/// faster when it is not merged with this one into dump()
NLOHMANN_VIEW_NOINLINE void dump_ascii(const node* root)
{
if (m_style.pretty)
{
if (m_style.source_numbers)
{
dump_text<true, true, true>(root);
}
else
{
dump_text<false, true, true>(root);
}
}
else if (m_style.source_numbers)
{
dump_text<true, false, true>(root);
}
else
{
dump_text<false, false, true>(root);
}
}
/// whether dump_text() copies a string or key from the source: one of
/// the source, and with ensure_ascii one without bytes >= 0x7F (an overload
/// each, so that the walk without ensure_ascii is compiled exactly as
/// before: with GCC, a condition that only folds away later changed its
/// block layout and made it 4-6% slower)
static bool copied(const node& n, const char* /*src*/, std::false_type /*ensure_ascii*/) noexcept
{
return (n.flags & node_flags::storage) == 0;
}
static bool copied(const node& n, const char* src, std::true_type /*ensure_ascii*/) noexcept
{
return (n.flags & node_flags::storage) == 0 && is_ascii_without_delete(src + n.off, n.len);
}
/*!
@brief a string with ensure_ascii: the characters that are not printable ASCII as \\uXXXX
(a character beyond U+FFFF as a surrogate pair); strings of a document
are valid UTF-8, except in a damaged image loaded with image_check::bounds,
for which this throws what basic_json::dump() throws for the string
*/
NLOHMANN_VIEW_NOINLINE void write_ascii(const node& n)
{
const auto* const s = reinterpret_cast<const unsigned char*>(m_doc.str(n)); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
const std::size_t size = n.len;
char* w = m_out.cursor();
char* lim = m_out.limit();
// room for k bytes
const auto room = [&](std::size_t k)
{
if (NLOHMANN_VIEW_UNLIKELY(static_cast<std::size_t>(lim - w) < k))
{
m_out.set_cursor(w);
m_out.reserve(k);
w = m_out.cursor();
lim = m_out.limit();
}
};
static constexpr const char* hex = "0123456789abcdef";
// \\uXXXX
const auto u_escape = [&](std::uint32_t u)
{
w[0] = '\\';
w[1] = 'u';
w[2] = hex[(u >> 12u) & 0xFu];
w[3] = hex[(u >> 8u) & 0xFu];
w[4] = hex[(u >> 4u) & 0xFu];
w[5] = hex[u & 0xFu];
w += 6;
};
room(1);
*w++ = '"';
std::size_t i = 0;
// (a string of the source starts with a run that has nothing to escape, 16 bytes per step)
const std::size_t prefix = (n.flags & node_flags::storage) == 0 ? ascii_copyable_run(s, size, reinterpret_cast<const unsigned char*>(m_doc.src + m_doc.size)) : 0; // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
if (prefix != 0)
{
room(prefix);
std::memcpy(w, s, prefix);
w += prefix;
i = prefix;
}
while (i < size)
{
// (a character that needs an escape sequence is not worth a call)
const std::size_t run = is_ascii_copyable(s[i]) ? find_ascii_copyable_run(s + i, size - i) : 0;
if (run != 0)
{
room(run);
std::memcpy(w, s + i, run);
w += run;
i += run;
continue;
}
room(12);
if (s[i] >= 0x80)
{
const unsigned char lead = s[i];
// (the two-byte sequences of Latin text and the three-byte sequences of most other
// text, with continuation bytes of any value: the ranges that are restricted (E0,
// ED, and the sequences of four bytes) go through validate_one_utf8())
if (lead >= 0xE1 && lead <= 0xEF && lead != 0xED && size - i >= 3 && (s[i + 1] & 0xC0u) == 0x80 && (s[i + 2] & 0xC0u) == 0x80)
{
u_escape(((lead & 0x0Fu) << 12u) | ((s[i + 1] & 0x3Fu) << 6u) | (s[i + 2] & 0x3Fu));
i += 3;
continue;
}
if (lead >= 0xC2 && lead <= 0xDF && size - i >= 2 && (s[i + 1] & 0xC0u) == 0x80)
{
u_escape(((lead & 0x1Fu) << 6u) | (s[i + 1] & 0x3Fu));
i += 2;
continue;
}
const std::size_t len = validate_one_utf8(s + i, size - i);
if (NLOHMANN_VIEW_UNLIKELY(len == 0))
{
m_out.set_cursor(w);
invalid_utf8(s, size);
return; // LCOV_EXCL_LINE (invalid_utf8 throws)
}
std::uint32_t codepoint = s[i] & (0xFFu >> (len + 1));
for (std::size_t k = 1; k < len; ++k)
{
codepoint = (codepoint << 6u) | (s[i + k] & 0x3Fu);
}
i += len;
if (codepoint <= 0xFFFF)
{
u_escape(codepoint);
}
else
{
u_escape(0xD7C0u + (codepoint >> 10u));
u_escape(0xDC00u + (codepoint & 0x3FFu));
}
continue;
}
// a quote, a backslash, or a control character (or DEL)
const unsigned char c = s[i++];
char escape = 0;
switch (c)
{
case 0x08:
escape = 'b';
break;
case 0x09:
escape = 't';
break;
case 0x0A:
escape = 'n';
break;
case 0x0C:
escape = 'f';
break;
case 0x0D:
escape = 'r';
break;
case 0x22:
escape = '"';
break;
case 0x5C:
escape = '\\';
break;
default:
break;
}
if (escape != 0)
{
w[0] = '\\';
w[1] = escape;
w += 2;
}
else
{
u_escape(c);
}
}
room(1);
*w++ = '"';
m_out.set_cursor(w);
}
/// the copies of dump_text() that are not fixed-size moves (long
/// strings, or near the end of the source); out of line, so that the
/// compiler does not merge the fixed-size moves into this call
NLOHMANN_VIEW_NOINLINE static void copy_long(char* to, const char* from, std::size_t n) noexcept
{
std::memcpy(to, from, n);
}
/// a compile-time option as a runtime condition: testing the template
/// argument directly makes a condition like `Editable && c` constant when
/// the option is off, which MSVC reports as C4127
static NLOHMANN_VIEW_ALWAYS_INLINE bool enabled(bool option) noexcept
{
return option;
}
/*!
@brief the output, compact (the default dump()) or indented, with or without ensure_ascii
The same walk as dump(), with the write position in a local variable
(stores through char pointers would otherwise force a reload of the
buffer's members after each one), and with strings and number tokens of
the source copied by fixed-size moves of 32 bytes where the source has
that many bytes left, instead of a library call per token. The buffer
keeps 64 bytes of slack for the overshoot. With ensure_ascii (Ascii), a
string of the source is copied only if it has nothing to escape (see
copied()); any other string is written by write_ascii().
*/
template<bool SourceNumbers, bool Pretty, bool Ascii>
void dump_text(const node* root)
{
struct frame
{
const node* pos; ///< (editable documents) next element, or key of the next member
const node* end;
bool object;
};
std::array<frame, 32> buffer; // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
std::vector<frame> heap;
small_stack<frame> stack(buffer.data(), buffer.size(), heap);
const char* const src = m_doc.src;
const char* const src_end = src + m_doc.size;
char* w = m_out.cursor();
char* lim = m_out.limit();
// (indented output) a line break and 63 indentation characters, from
// which the line break of a level of up to 63 characters is a copy
std::array<char, 64> ruler; // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): filled below if used
if (Pretty)
{
ruler.fill(m_style.indent_char);
ruler[0] = '\n';
}
const std::size_t indent = m_style.indent;
depth_counter<Pretty> level; // depth of the innermost open container (indented output)
// room for n bytes and the slack
const auto room = [&](std::size_t n)
{
if (NLOHMANN_VIEW_UNLIKELY(static_cast<std::size_t>(lim - w) < n + 64))
{
m_out.set_cursor(w);
m_out.reserve(n + 64);
w = m_out.cursor();
lim = m_out.limit();
}
};
// copy n bytes of the source (after room(n))
const auto copy = [&](const char* from, std::size_t n)
{
if (n <= 32 && static_cast<std::size_t>(src_end - from) >= 32)
{
std::memcpy(w, from, 32);
}
else if (n <= 256 && static_cast<std::size_t>(src_end - from) >= n + 32)
{
for (std::size_t i = 0; i < n; i += 32)
{
std::memcpy(w + i, from + i, 32);
}
}
else
{
copy_long(w, from, n);
}
w += n;
};
// (indented output) the line break and the indentation of a level
const auto newline = [&](std::size_t depth)
{
const std::size_t count = 1 + (depth * indent);
room(count);
if (count <= ruler.size())
{
std::memcpy(w, ruler.data(), ruler.size());
}
else
{
*w = '\n';
std::memset(w + 1, m_style.indent_char, count - 1);
}
w += count;
};
// a string that is not a plain string of the source (out of line, so
// that the cursor stays in a register here)
const auto escaped = [&](const node & n)
{
m_out.set_cursor(w);
if (enabled(Ascii))
{
write_ascii(n);
}
else
{
write_decoded(n, Pretty);
}
w = m_out.cursor();
lim = m_out.limit();
};
// Read-only documents: the elements of a container follow it in the
// node array, so the walk goes through the array in order, and a
// frame only needs the end of its container. Editable documents: the
// elements of a moved container live elsewhere, so a frame keeps the
// position of the next element (see navigation).
// The innermost open container is kept in registers (cur; end ==
// nullptr: none), the stack holds the ones around it.
frame cur{nullptr, nullptr, false};
const node* n = root;
for (;;)
{
// write the value at n (read-only documents: and advance n)
bool opened = false;
// false positive: n comes from nav::value(), which never returns null for a valid index
// @infer-ignore NULLPTR_DEREFERENCE
switch (static_cast<value_t>(n->kind))
{
case value_t::string:
if (copied(*n, src, std::integral_constant<bool, Ascii>()))
{
room(n->len + 2);
*w++ = '"';
copy(src + n->off, n->len);
*w++ = '"';
}
else
{
escaped(*n);
}
break;
case value_t::number_integer:
case value_t::number_unsigned:
{
const std::uint32_t len = number_length(*n);
room(len);
if (enabled(Editable) && (n->flags & node_flags::storage) != 0)
{
copy_long(w, m_doc.str(*n), len); // a canonical token written by an edit
w += len;
break;
}
const char* const token = src + n->off;
if (!enabled(SourceNumbers) && NLOHMANN_VIEW_UNLIKELY(len == 2 && token[0] == '-' && token[1] == '0'))
{
*w++ = '0'; // parse() reads -0 as the integer 0
}
else
{
copy(token, len);
}
break;
}
case value_t::number_float:
if (enabled(SourceNumbers) && (n->flags & node_flags::storage) != node_flags::edited)
{
room(n->len);
copy(src + n->off, n->len);
}
else if (enabled(fast_double::value))
{
room(64);
w = write_double_at(w, *n);
}
else
{
m_out.set_cursor(w);
write_float_node(*n);
w = m_out.cursor();
lim = m_out.limit();
}
break;
case value_t::boolean:
room(8);
if ((n->flags & node_flags::is_true) != 0)
{
w = put_literal(w, "true");
}
else
{
w = put_literal(w, "false");
}
break;
case value_t::object:
case value_t::array:
{
const bool object = n->kind == static_cast<std::uint8_t>(value_t::object);
room(8);
if (n->len == 0)
{
w = object ? put_literal(w, "{}") : put_literal(w, "[]");
}
else
{
*w++ = object ? '{' : '[';
stack.push_back(cur);
level.open();
if (Editable)
{
cur = frame{nav::first(m_doc, n), nav::end(m_doc, n), object};
}
else
{
cur = frame{nullptr, n + n->next, object};
}
opened = true;
}
break;
}
case value_t::null:
room(8);
w = put_literal(w, "null");
break;
case value_t::binary: // LCOV_EXCL_LINE (not in a document)
case value_t::discarded: // LCOV_EXCL_LINE
default: // LCOV_EXCL_LINE
break; // LCOV_EXCL_LINE
}
if (!Editable)
{
++n; // the next node: the first element of an opened container, or the node after a scalar
}
// go to the next value: close finished containers, then separate
// (a container just opened has an element)
if (!opened)
{
for (;;)
{
if (cur.end == nullptr)
{
m_out.set_cursor(w);
m_out.finish();
return;
}
if ((Editable ? cur.pos : n) != cur.end)
{
break;
}
if (Pretty)
{
newline(level.close());
}
room(1);
*w++ = cur.object ? '}' : ']';
cur = stack.back();
stack.pop_back();
}
room(1);
*w++ = ',';
}
if (Pretty)
{
newline(level.get());
}
const node* const at = Editable ? cur.pos : n;
if (cur.object)
{
const node& key = *at;
if (copied(key, src, std::integral_constant<bool, Ascii>()))
{
room(key.len + (Pretty ? 4 : 3));
*w++ = '"';
copy(src + key.off, key.len);
w[0] = '"';
w[1] = ':';
w += 2;
}
else
{
escaped(key);
room(Pretty ? 2 : 1);
*w++ = ':';
}
if (Pretty)
{
*w++ = ' ';
}
if (Editable)
{
n = nav::value(at + 1);
cur.pos = document_data::after(at + 1);
}
else
{
++n;
}
}
else if (Editable)
{
n = nav::value(at);
cur.pos = document_data::after(at);
}
}
}
/// a float node as dump() writes it
void write_float_node(const node& n)
{
write_float_node(n, fast_double {});
}
void write_float_node(const node& n, std::false_type /*other*/)
{
write_float(float_value<number_float_t>(m_doc, n));
}
void write_float_node(const node& n, std::true_type /*double*/)
{
m_out.reserve(64);
m_out.set_cursor(write_double_at(m_out.cursor(), n));
}
/*!
@brief (doubles) the float at n as dump() writes it, at w (64 bytes of room)
A token of at most 15 significant digits is written from its digits,
without a conversion: two decimals of at most 15 digits are farther
apart than the rounding interval of a (normal) double (the argument
behind DBL_DIG), so the token's digits are the shortest ones of its
double, which the library's conversion writes (Zmij). Other tokens are
converted from the digits already read.
*/
char* write_double_at(char* w, const node& n)
{
const unsigned int_digits = n.extra & 0xFFu;
const unsigned frac_digits = n.extra >> 8u;
if ((n.flags & node_flags::storage) != node_flags::edited && int_digits + frac_digits <= 19)
{
const auto* const first = reinterpret_cast<const unsigned char*>(m_doc.src + n.off); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
const float_significand d = layout_decimal(first, first + n.len, int_digits, frac_digits, reinterpret_cast<const unsigned char*>(m_doc.src + m_doc.size)); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
// (the exponent keeps the value far from subnormals and overflow)
if (d.w != 0 && d.w < 1000000000000000u && d.exponent >= -290 && d.exponent <= 290)
{
*w = '-';
w += d.negative ? 1 : 0;
// (without leading zeros, all digits of the token count; the
// check also keeps an image that was only checked for bounds,
// whose token may not be made of digits, from the counted
// overload)
const auto& powers = ::nlohmann::detail::dtoa_impl::powers_of_ten();
const unsigned count = int_digits + frac_digits;
return count - 1u < 15u && d.w >= powers[count - 1u] && d.w < powers[count]
? ::nlohmann::detail::dtoa_impl::write_short_decimal(w, d.w, static_cast<int>(count), static_cast<int>(d.exponent))
: ::nlohmann::detail::dtoa_impl::write_short_decimal(w, d.w, static_cast<int>(d.exponent));
}
return write_double_value_at(w, decimal_to_float<double>(d)); // (without reading the token again)
}
return write_double_value_at(w, static_cast<double>(float_value<number_float_t>(m_doc, n)));
}
/// n bytes of text at w
static char* write_text_at(char* w, const char* text, std::size_t n) noexcept
{
std::memcpy(w, text, n);
return w + n;
}
/// a double as dump() writes it, at w (64 bytes of room)
static char* write_double_value_at(char* w, double x)
{
// (from the bits: without the checks of to_chars())
auto bits = bit_cast<std::uint64_t>(x);
if (NLOHMANN_VIEW_UNLIKELY((bits & 0x7FF0000000000000u) == 0x7FF0000000000000u))
{
return write_text_at(w, "null", 4);
}
*w = '-';
w += bits >> 63u;
bits &= ~(std::uint64_t{1} << 63u);
if (bits == 0)
{
return write_text_at(w, "0.0", 3);
}
return ::nlohmann::detail::dtoa_impl::write_shortest(w, ::nlohmann::detail::zmij::to_shortest(bits));
}
/// as serializer::dump_float()
void write_float(number_float_t x)
{
if (!std::isfinite(x))
{
m_out.put("null");
return;
}
static constexpr bool is_ieee_single_or_double
= std::numeric_limits<number_float_t>::is_iec559
&& ((std::numeric_limits<number_float_t>::digits == 24 && std::numeric_limits<number_float_t>::max_exponent == 128)
|| (std::numeric_limits<number_float_t>::digits == 53 && std::numeric_limits<number_float_t>::max_exponent == 1024));
write_float(x, std::integral_constant<bool, is_ieee_single_or_double>());
}
void write_float(number_float_t x, std::true_type /*is_ieee_single_or_double*/)
{
std::array<char, 64> buf{};
const char* const end = ::nlohmann::detail::to_chars(buf.data(), buf.data() + buf.size(), x);
m_out.put(buf.data(), static_cast<std::size_t>(end - buf.data()));
}
void write_float(number_float_t x, std::false_type /*is_ieee_single_or_double*/)
{
// other types (e.g. long double) are rare: the library writes them
const string_t s = BasicJsonType(x).dump();
m_out.put(s.data(), s.size());
}
/// throw what basic_json::dump() throws for a string that is not valid UTF-8
NLOHMANN_VIEW_NOINLINE static void invalid_utf8(const unsigned char* s, std::size_t n)
{
const string_t dumped = BasicJsonType(string_t(reinterpret_cast<const char*>(s), n)).dump(); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
static_cast<void>(dumped);
}
template<bool EnsureAscii>
void write_codepoint(std::uint32_t codepoint, const unsigned char* bytes, std::size_t len)
{
switch (codepoint)
{
case 0x08:
m_out.put("\\b");
return;
case 0x09:
m_out.put("\\t");
return;
case 0x0A:
m_out.put("\\n");
return;
case 0x0C:
m_out.put("\\f");
return;
case 0x0D:
m_out.put("\\r");
return;
case 0x22:
m_out.put("\\\"");
return;
case 0x5C:
m_out.put("\\\\");
return;
default:
break;
}
if (codepoint <= 0x1F || (EnsureAscii && codepoint >= 0x7F))
{
if (codepoint <= 0xFFFF)
{
write_u_escape(codepoint);
}
else
{
write_u_escape(0xD7C0u + (codepoint >> 10u));
write_u_escape(0xDC00u + (codepoint & 0x3FFu));
}
return;
}
m_out.put(reinterpret_cast<const char*>(bytes), len); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast) LCOV_EXCL_LINE (printable characters are copied in runs)
}
void write_u_escape(std::uint32_t u)
{
static constexpr const char* hex = "0123456789abcdef";
const std::array<char, 6> e = {{'\\', 'u', hex[(u >> 12u) & 0xFu], hex[(u >> 8u) & 0xFu], hex[(u >> 4u) & 0xFu], hex[u & 0xFu]}};
m_out.put(e.data(), e.size());
}
const document_data& m_doc;
output_buffer<string_t> m_out;
const dump_style m_style;
};
} // namespace view
} // namespace detail
NLOHMANN_JSON_NAMESPACE_END
Reference in new issue View Git Blame Copy Permalink