Compare commits

...
10 Commits
Author SHA1 Message Date
Niels Lohmann f5cb977387 Rename the key index's match callback so Flawfinder doesn't read it as std::equal
Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:42:44 +02:00
Niels Lohmann 4cde5e39b9 Collect the members of ordered_json objects on a stack until they are known
An ordered_json object keeps its members in a vector and appends every new
member to it, so reading an object of unknown size reallocated the vector at
1, 2, 4, 8, ... members, and ordered_map copies all keys on every
reallocation. The DOM SAX parser now collects the members of such objects on
one stack shared by all open objects, as it does the elements of arrays, and
moves them into a vector of exactly the right size when the closing brace
arrives. An object that gets past 32 members switches to appending to its
own vector (and, from 64, to the key index of the earlier commit). The
members are the same, in the same order, and a key that occurs twice keeps
the place of its first occurrence. std::map based objects, objects whose
size the input announces (binary formats), and builds with JSON_DIAGNOSTICS
or JSON_DIAGNOSTIC_POSITIONS are untouched.

What only ordered objects need (the collected members, the key indexes) is
a member of the parser only for them, and the code that stores a value tries
the array that takes its elements directly first. An array that outgrows
the stack now reserves 32 elements, the capacity a vector that grew by
itself has at that point; with the larger reserve before, parsing
floats.json (1 000 000 numbers) was 2 % slower on Apple M1 than before this
series.

ordered_json parse, Linux Ir against develop (GCC / Clang): citm -11.6 % /
-11.7 %, twitter -8.4 % / -8.4 %, jeopardy (12 000 objects of 7 keys)
-10.5 % / -8.0 %, canada -3.9 % / -5.2 %, 124-byte object -20.9 % / -16.5 %.
macOS time (M1, earlier base): citm -19 %, twitter -13 %, jeopardy -13 %,
canada -14 %, 124-byte object -26 %. json parse: steady-state twitter -0.1 %
(see the previous commits), citm -1.2 %.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:28:08 +02:00
Niels Lohmann 284ef38125 Do not convert floats that cannot overflow in accept()
accept() only has to find out whether a float token is finite: json_sax_acceptor
ignores the value, and the parser rejects only a float that overflows. A
float whose digits fit into 19 digits and whose decimal exponent is at most
max_exponent10 - 19 is below 10^max_exponent10 and cannot overflow, so the
lexer, told by accept() that the value is not needed, skips the conversion
for those. A token that may overflow, with more digits, or in another float
format is converted as before, so 1e309 and its neighbours are rejected as
they were (the new test compares accept() with parse() on tokens around the
limit).

json::accept(): canada.json Ir -8.6 % (GCC) / -10.9 % (Clang), 124-byte
object -3.2 %; cycles on the M1 (release benchmarks, LTO): canada -7.8 %,
floats.json -9.0 %. The check tests the accept() flag first, so the parser
that needs the value pays one load and branch: SaxParse canada +0.35 % Ir
(GCC and Clang), DOM canada +0.4 %.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:28:05 +02:00
Niels Lohmann e38bb5d6f8 Test parsing containers of the sizes where the DOM parser changes storage
The DOM parser collects the elements of an array on a stack (up to 16) and
the members of an ordered_json object (up to 32) before it moves them into
the container, and uses a hash table for keys from 64 members on. Parse
arrays and objects of the sizes around these limits, nested into each other,
with repeated keys, and as random trees that are built next to their text,
and compare json and ordered_json against the value built through the API.

The tests pass on the code from before these changes as well. The random
trees are what finds the mistakes that are easy to make here: while the
collecting change was being written, a collected array that outgrew the
stack at the moment its parent had just reached the limit left its elements
in the parent, and only these tests (142 failed assertions) showed it.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:28:03 +02:00
Niels Lohmann ffa9babc8d Trim the work the DOM parser does per container and per member
Opening and closing an array or object cost the DOM parser more than before
the exact-size array change, which showed on documents of many small
objects (jeopardy.json: +1.4 % Ir with GCC). This takes it back:

- the stack of open containers keeps its first eight entries in the parser
  object instead of in a vector, so it needs no capacity check beyond the
  push itself and no allocation for typical documents;
- a container is created in the place it ends up in, instead of as a
  temporary that is then moved there;
- a new object member, which is null, is constructed over the null instead
  of being swapped with it (only with the default base class of basic_json);
- the key index that big ordered_json objects use is only a part of the
  stack entries of objects that can have one;
- an array that outgrows the stack reserves twice what it has so far, as a
  vector that grew by itself would have.

DOM parse Ir of the series up to here against develop, Linux GCC / Clang:
canada -6.6 % / -7.3 %, citm +1.2 % / +0.2 %, jeopardy +0.4 % / -0.1 %,
124-byte object -6.5 % / -6.2 %. (twitter measures +6.8 % / +6.6 % at this
point in steady state, almost all of it in glibc's malloc/consolidate for the
new allocation sizes; with the last commit of the series it is -0.1 % / 0.0 %.)

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:28:00 +02:00
Niels Lohmann 5b08041308 Look up the keys of big ordered_json objects in a hash table
ordered_map keeps its elements in a vector and searches it linearly for
every key it adds, so parsing an object with n keys into ordered_json took
O(n^2) key comparisons: 33 billion instructions for one object with 30 000
keys, where json needs 245 million. From 64 keys on, the DOM parser now
keeps a hash table of the keys of the open object (hashes and positions
only; the elements stay in the vector) and finds duplicate keys there. The
result is the same, including the position of a key that occurs twice.
Smaller objects, std::map based objects and key types without std::hash
are untouched.

ordered_json parse, macOS instructions: object with 1 000 keys 64.8 M ->
14.6 M, 10 000 keys 5 104 M -> 38.7 M, 30 000 keys 33 358 M -> 89 M.
Linux Ir of an object with 2 000 keys: 43.0 M -> 1.9 M (GCC), 37.9 M ->
1.8 M (Clang).

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:27:58 +02:00
Niels Lohmann eef9aab5ba Remember where the DOM parser stores the next value
The DOM SAX parser looked up the innermost open container for every value:
the end of its frame stack, then the object or array pointer in the frame.
Compilers have to reload all of that after each value is stored, since a
store of a value may alias the frame. The parser now keeps the innermost
container's kind and, for an array that stores directly, the array itself
in members, updated when a container opens or closes.

Flat arrays of numbers are the case that mattered: parsing floats.json
(1 000 000 doubles) into a DOM took 2 % more cycles than before the
previous commit on macOS and now takes 0.4 % fewer than the code before it
(signed_ints.json -0.8 %). Linux Ir is unchanged (canada -6.1 % GCC /
-6.9 % Clang, citm -0.3 % / -0.8 %, twitter +0.2 % / 0.0 %).

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:27:55 +02:00
Niels Lohmann c27f770c07 Reuse one value for the key events of the callback parser
parse() with a callback built a new string value for every key it passed to
the callback (a heap allocation and release per key), and copied the key a
second time into the object. The callback parser now keeps one string value
for key events, which it assigns, and moves the key into the object.

Callback parse, Linux Ir (GCC / Clang): twitter -13.3 % / -11.7 %, citm
-9.5 % / -7.6 %. (Earlier macOS measurement: twitter -19.5 % instructions.)

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:27:53 +02:00
Niels Lohmann 2e454713d0 Build arrays of the DOM parser in one exact-size allocation
Reading an array of unknown length appended its elements to a std::vector
one at a time, so a small array reallocated (and moved its elements) at
sizes 1, 2, 4, ... The DOM SAX parser now collects the elements of arrays
whose length is unknown on one stack shared by all open arrays and moves
them into an array of exactly the right size when the closing bracket
arrives. An array that grows past 16 elements switches to appending to its
own vector, so large arrays pay no second copy. Binary formats, which
announce the length, and builds with JSON_DIAGNOSTICS or
JSON_DIAGNOSTIC_POSITIONS (whose values need their final place from the
start) keep storing into the array directly.

The stacks of open containers are reserved on first use instead of growing
from one element.

The binary readers get a DOM parser of their own, json_sax_dom_binary_parser:
it is the plain parser that stores every value directly in its container
(the code the readers used before). Their containers announce their size, so
the bookkeeping of the collecting parser (frame entries, the destination of a
finished container) was pure overhead for them: with the collecting parser
FromBinaryShape nested was 3-12 % and containers up to 5 % more instructions
(GCC) than before. With the plain parser Linux Ir of FromBinaryShape and
FromBinaryBuffer is within 0.8 % of develop (Clang: identical).

canada.json has 56 000 two-element arrays that took three allocations each
and now take two. DOM parse Ir on Linux against develop (GCC / Clang): canada
-7.0 % / -7.7 %, citm -0.8 % / -1.6 %, 124-byte object -5.4 % / -5.5 %;
accept and SAX are untouched.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:27:51 +02:00
Niels Lohmann 56ea5dd6bf Reserve the parser's nesting stack once
The stack of open arrays and objects was a std::vector<std::uint8_t> that
allocated at the first push_back and then reallocated at 1, 2, 4, 8 bytes,
so a document nested four levels deep made four allocations before it
parsed a value. A document that starts with a container now reserves 32
levels in one allocation; deeper documents grow as before. The loop that
reads the tokens is the code it was, so large documents are not affected
(a stack object with its own storage, and a reserve at every container,
both cost the token loop instructions or cycles).

Fixed cost of parsing a 124-byte object (macOS instructions): accept
-7.6 %, DOM -2.5 %; no change for a single number.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-10-22 09:27:48 +02:00
8 changed files with 2434 additions and 76 deletions

No files matched your search

@@ -86,7 +86,7 @@ JSON_INLINE_VARIABLE constexpr std::size_t max_valueless_container_size = 1 << 2
/*! /*!
@brief deserialization of BJData, BON8, BSON, CBOR, MessagePack, and UBJSON values @brief deserialization of BJData, BON8, BSON, CBOR, MessagePack, and UBJSON values
*/ */
template<typename BasicJsonType, typename InputAdapterType, typename SAX = json_sax_dom_parser<BasicJsonType, InputAdapterType>> template<typename BasicJsonType, typename InputAdapterType, typename SAX = json_sax_dom_binary_parser<BasicJsonType, InputAdapterType>>
class binary_reader class binary_reader
{ {
using number_integer_t = typename BasicJsonType::number_integer_t; using number_integer_t = typename BasicJsonType::number_integer_t;
File diff suppressed because it is too large. Load diff
+10
View File
@@ -14,6 +14,7 @@
#include <cstdio> // snprintf #include <cstdio> // snprintf
#include <cstring> // memcmp #include <cstring> // memcmp
#include <initializer_list> // initializer_list #include <initializer_list> // initializer_list
#include <limits> // numeric_limits
#include <string> // char_traits, string #include <string> // char_traits, string
#include <utility> // move #include <utility> // move
#include <vector> // vector #include <vector> // vector
@@ -1816,6 +1817,15 @@ scan_number_done:
/// significand (see scan_number_bulk_contiguous()) /// significand (see scan_number_bulk_contiguous())
bool convert_accumulated_float(const float_significand& s, std::true_type /*native*/) noexcept bool convert_accumulated_float(const float_significand& s, std::true_type /*native*/) noexcept
{ {
// accept() only needs to know that the value is finite (see the parser):
// w < 10^19, so a value below 10^max_exponent10 cannot overflow, and a
// value that underflows is a finite number as well. Only a token that
// may overflow has to be converted to find out.
if (JSON_HEDLEY_UNLIKELY(discard_number_values && s.exponent + 19 <= std::numeric_limits<number_float_t>::max_exponent10))
{
value_float = 0;
return true;
}
value_float = decimal_to_float<number_float_t>(s); value_float = decimal_to_float<number_float_t>(s);
return true; return true;
} }
+7
View File
@@ -250,6 +250,13 @@ class parser
// value to avoid a goto (see comment where set to true) // value to avoid a goto (see comment where set to true)
bool skip_to_state_evaluation = false; bool skip_to_state_evaluation = false;
// a document that starts with a container pushes onto the stack at
// once: allocate it in one step instead of growing it from 1 byte
if (last_token == token_type::begin_object || last_token == token_type::begin_array)
{
states.reserve(32);
}
while (true) while (true)
{ {
if (!skip_to_state_evaluation) if (!skip_to_state_evaluation)
+4 -2
View File
@@ -161,6 +161,8 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
template<typename BasicJsonType, typename InputAdapterType> template<typename BasicJsonType, typename InputAdapterType>
friend class ::nlohmann::detail::json_sax_dom_parser; friend class ::nlohmann::detail::json_sax_dom_parser;
template<typename BasicJsonType, typename InputAdapterType> template<typename BasicJsonType, typename InputAdapterType>
friend class ::nlohmann::detail::json_sax_dom_binary_parser;
template<typename BasicJsonType, typename InputAdapterType>
friend class ::nlohmann::detail::json_sax_dom_callback_parser; friend class ::nlohmann::detail::json_sax_dom_callback_parser;
#if JSON_DIAGNOSTIC_POSITIONS #if JSON_DIAGNOSTIC_POSITIONS
friend struct ::nlohmann::detail::diagnostic_positions; friend struct ::nlohmann::detail::diagnostic_positions;
@@ -6287,7 +6289,7 @@ public:
from_ubjson/from_bjdata/from_bon8/from_bson overloads from_ubjson/from_bjdata/from_bon8/from_bson overloads
Building the @ref binary_reader as a named local, rather than as a Building the @ref binary_reader as a named local, rather than as a
temporary that @ref detail::json_sax_dom_parser::parse is called on in temporary that @ref detail::json_sax_dom_binary_parser::parse is called on in
the same expression, avoids a false-positive cppcheck accessMoved the same expression, avoids a false-positive cppcheck accessMoved
warning in each of the 16 callers. warning in each of the 16 callers.
*/ */
@@ -6298,7 +6300,7 @@ public:
const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error) const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
{ {
basic_json result; basic_json result;
detail::json_sax_dom_parser<basic_json, InputAdapterType> sdp(result, allow_exceptions); detail::json_sax_dom_binary_parser<basic_json, InputAdapterType> sdp(result, allow_exceptions);
binary_reader<InputAdapterType> reader(std::move(ia), format, error_handler); binary_reader<InputAdapterType> reader(std::move(ia), format, error_handler);
if (!reader.sax_parse(&sdp, strict, tag_handler)) if (!reader.sax_parse(&sdp, strict, tag_handler))
{ {
+1094 -64
View File
File diff suppressed because it is too large. Load diff
+29
View File
@@ -2952,3 +2952,32 @@ TEST_CASE("diagnostic positions: value lifetime, input adapters, and SAX")
} }
} }
#endif #endif
TEST_CASE("accept() agrees with parse() on floats around the overflow limit")
{
// accept() does not convert a float that cannot overflow; the tokens
// around the limit (and beyond it) must still be rejected exactly as
// parse() rejects them
const char* const numbers[] =
{
"1e0", "123.456e-5", "1e-400", "-1e-400", "1e307", "1e308", "1e309", "-1e309",
"10e307", "100e307", "0.1e309", "0.1e310", "1E+308", "1E+309",
"1.7976931348623157e308", "1.7976931348623159e308", "1.8e308", "-1.7976931348623157e308",
"12345678901234567e291", "12345678901234567e292", "123456789012345678e290", "123456789012345678e291",
"1234567890123456789e289", "1234567890123456789e290", "9999999999999999999e289", "9999999999999999999e290",
"0.00000000000000000001e328", "0.00000000000000000001e329"
};
for (const char* number : numbers)
{
for (const std::string text :
{
std::string(number), "[" + std::string(number) + "]", "{\"a\":" + std::string(number) + "}"
})
{
INFO(text);
const bool parses = !json::parse(text, nullptr, false).is_discarded();
CHECK(json::accept(text) == parses);
}
}
}
+271
View File
@@ -237,3 +237,274 @@ TEST_CASE("regression test for issue #5673 - ordered_json::emplace with a non-rv
CHECK(oj.dump() == "{\"a\":1}"); CHECK(oj.dump() == "{\"a\":1}");
} }
} }
namespace
{
// the sizes around which the DOM parser changes how it stores the elements of
// an array (16) and the members of an object in a vector (32, and 64 for the
// key index)
const std::size_t container_sizes[] = {0, 1, 2, 3, 15, 16, 17, 18, 31, 32, 33, 34, 63, 64, 65, 66, 200};
template<typename Json>
std::string array_text(const std::size_t n, const std::size_t depth)
{
std::string text = "[";
for (std::size_t i = 0; i < n; ++i)
{
text += (i == 0 ? "" : ",");
// alternate scalars, empty containers and nested containers
switch (i % 4)
{
case 0:
text += std::to_string(i);
break;
case 1:
text += "\"s" + std::to_string(i) + "\"";
break;
case 2:
text += depth == 0 ? "[]" : array_text<Json>(n / 2, depth - 1);
break;
default:
text += depth == 0 ? "{}" : "{\"a\":" + array_text<Json>(n / 3, depth - 1) + ",\"b\":null}";
break;
}
}
return text + "]";
}
template<typename Json>
std::string object_text(const std::size_t n, const std::size_t depth, const bool duplicates)
{
std::string text = "{";
for (std::size_t i = 0; i < n; ++i)
{
text += (i == 0 ? "" : ",");
text += "\"key" + std::to_string(i) + "\":";
switch (i % 4)
{
case 0:
text += std::to_string(i);
break;
case 1:
text += depth == 0 ? "[]" : array_text<Json>(n / 2, depth - 1);
break;
case 2:
text += depth == 0 ? "{}" : object_text<Json>(n / 2, depth - 1, duplicates);
break;
default:
text += "\"value\"";
break;
}
}
if (duplicates)
{
// keys that are there already, the first and the last one, with other values
text += n == 0 ? "" : ",";
text += "\"key0\":[1,2],\"key" + std::to_string(n - 1) + "\":{\"again\":true}";
}
return text + "}";
}
template<typename Json>
void check_same_dump(const std::string& text)
{
// the text is what dump() writes for the parsed value, and the value
// survives a second round trip: arrays keep their elements in order, an
// ordered_json keeps its members in order of appearance
const Json parsed = Json::parse(text);
const Json again = Json::parse(parsed.dump());
CHECK(parsed == again);
CHECK(parsed.dump() == again.dump());
CHECK(Json::accept(text));
}
} // namespace
TEST_CASE("parsing containers of sizes around the limits of the DOM parser")
{
SECTION("arrays")
{
for (const std::size_t n : container_sizes)
{
const std::string text = array_text<json>(n, 2);
INFO("size ", n);
const json parsed = json::parse(text);
CHECK(parsed.is_array());
CHECK(parsed.size() == n);
for (std::size_t i = 0; i < n; i += 4)
{
CHECK(parsed[i] == i);
}
check_same_dump<json>(text);
check_same_dump<ordered_json>(text);
CHECK(ordered_json::parse(text).dump() == parsed.dump());
}
}
SECTION("objects")
{
for (const bool duplicates :
{
false, true
})
{
for (const std::size_t n : container_sizes)
{
const std::string text = object_text<json>(n, 2, duplicates);
INFO("size ", n, " duplicates ", duplicates);
const json parsed = json::parse(text);
const ordered_json ordered = ordered_json::parse(text);
CHECK(parsed.is_object());
CHECK(ordered.is_object());
CHECK(parsed.size() == ordered.size());
// the same members, in the order of their first appearance
std::size_t i = 0;
for (auto it = ordered.begin(); it != ordered.end(); ++it, ++i)
{
if (i < n)
{
CHECK(it.key() == "key" + std::to_string(i));
}
CHECK(parsed.at(it.key()) == json(it.value()));
}
if (duplicates && n > 1)
{
// the later value replaced the earlier one in its place
CHECK(ordered.at("key0") == ordered_json::parse("[1,2]"));
CHECK(ordered.at("key" + std::to_string(n - 1)) == ordered_json::parse("{\"again\":true}"));
CHECK(ordered.begin().key() == "key0");
CHECK(ordered.size() == n);
}
check_same_dump<json>(text);
check_same_dump<ordered_json>(text);
}
}
}
SECTION("objects with many different and repeated keys")
{
// 300 members of which a third repeat an earlier key
std::string text = "{";
for (std::size_t i = 0; i < 300; ++i)
{
text += (i == 0 ? "" : ",");
text += "\"k" + std::to_string(i % 200) + "\":" + std::to_string(i);
}
text += "}";
const ordered_json ordered = ordered_json::parse(text);
const json parsed = json::parse(text);
CHECK(ordered.size() == 200);
CHECK(parsed.size() == 200);
std::size_t i = 0;
for (auto it = ordered.begin(); it != ordered.end(); ++it, ++i)
{
CHECK(it.key() == "k" + std::to_string(i));
// the last value written
CHECK(it.value() == (i < 100 ? i + 200 : i));
CHECK(parsed.at(it.key()) == json(it.value()));
}
}
}
namespace
{
// a deterministic source of structure for parsing tests: builds a JSON text
// and, at the same time, the value it has to parse to
class tree_generator
{
public:
explicit tree_generator(std::uint32_t seed) : state(seed) {}
template<typename Json>
Json make(std::string& text, const std::size_t depth)
{
switch (next(depth == 0 ? 5 : 9))
{
case 0:
text += "null";
return Json(nullptr);
case 1:
text += "true";
return Json(true);
case 2:
{
const int n = static_cast<int>(next(2000)) - 1000;
text += std::to_string(n);
return Json(n);
}
case 3:
text += "\"text\"";
return Json("text");
case 4:
text += "2.5";
return Json(2.5);
case 5:
case 6:
case 7:
{
Json array = Json::array();
text += "[";
const std::size_t n = size();
for (std::size_t i = 0; i < n; ++i)
{
text += (i == 0 ? "" : ",");
array.push_back(make<Json>(text, depth - 1));
}
text += "]";
return array;
}
default:
{
Json object = Json::object();
text += "{";
const std::size_t n = size();
for (std::size_t i = 0; i < n; ++i)
{
const std::string key = "k" + std::to_string(i);
text += (i == 0 ? "\"" : ",\"") + key + "\":";
object[key] = make<Json>(text, depth - 1);
}
text += "}";
return object;
}
}
}
private:
std::uint32_t next(const std::uint32_t bound)
{
state = state * 1664525U + 1013904223U;
return (state >> 8) % bound;
}
std::size_t size()
{
return container_sizes[next(sizeof(container_sizes) / sizeof(*container_sizes))];
}
std::uint32_t state;
};
} // namespace
TEST_CASE("parsing random trees of arrays and objects")
{
for (std::uint32_t seed = 1; seed <= 150; ++seed)
{
INFO("seed ", seed);
std::string text;
std::string ordered_text;
const json expected = tree_generator(seed).make<json>(text, 3);
const ordered_json ordered_expected = tree_generator(seed).make<ordered_json>(ordered_text, 3);
CHECK(text == ordered_text);
CHECK(json::parse(text) == expected);
CHECK(json::accept(text));
const ordered_json ordered = ordered_json::parse(text);
CHECK(ordered == ordered_expected);
CHECK(ordered.dump() == ordered_expected.dump());
CHECK(json::parse(text).dump() == expected.dump());
}
}