Compare commits

..
Author SHA1 Message Date
Niels Lohmann b98aef8a07 Round-trip BJData ND-array annotations exactly (single precision, key order)
to_bjdata() encoded a JData-annotated object as a BJData ND-array in two
cases where from_bjdata() then returned a different value, breaking the
documented round-trip guarantee:

1. A "single" element that is finite and in range but not exactly
   representable as float (e.g. 0.1) or that underflows to 0 (e.g. 1e-300)
   was silently narrowed instead of falling back to a plain object, unlike
   out-of-range integer elements. write_bjdata_ndarray() now only accepts a
   "single" element if it survives the narrowing to float and back, the
   same criterion write_compact_float() already uses for CBOR/MessagePack.

2. from_bjdata() emitted the annotation keys as _ArraySize_, _ArrayType_,
   _ArrayData_ instead of the documented _ArrayType_, _ArraySize_,
   _ArrayData_, because the size key is written while the dimension vector
   is read, before the type key. For ordered_json, whose comparison takes
   key order into account, this made a round trip of the documented example
   compare unequal. The element type marker is known before the dimension
   vector is read (it precedes '#'), so it is now passed down and the
   "_ArrayType_" key is emitted first.

Fixes #5661.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
2026-09-29 23:52:49 +02:00
7 changed files with 182 additions and 154 deletions
@@ -132,8 +132,14 @@ The library uses the following mapping from JSON values types to BJData types ac
parsed back as a regular array,
- every entry of `"_ArraySize_"` is a positive integer, and their product is representable as a `std::size_t`,
- `"_ArrayData_"` is an array holding exactly that many elements, and
- every element of `"_ArrayData_"` is a number of the kind named by `"_ArrayType_"` (a floating-point number for
`single` and `double`, an integer otherwise).
- every element of `"_ArrayData_"` is a number of the kind named by `"_ArrayType_"`: for the integer types, a
value that fits the named width; for `double`, any value; for `single`, a value that survives narrowing to
`float` and back without change (for instance, `0.1` does not, since it is not exactly representable as
`float`).
An annotated object is always read back with its keys in the order shown above, `"_ArrayType_"`, `"_ArraySize_"`,
`"_ArrayData_"`, regardless of the order the ND-array's header stores them in on the wire. This matters for
`ordered_json`, whose comparison takes key order into account.
The current version of this library does not yet support automatic detection of and conversion from a nested JSON
array input to a BJData ND-array.
+42 -21
View File
@@ -2728,10 +2728,15 @@ class binary_reader
is_ndarray can only return `true` when its initial value
is `false`
@param[in] prefix type marker if already read, otherwise set to 0
@param[in] ndarray_dtype the element type marker of the enclosing bjdata ndarray if
already known (it precedes the dimension vector read here),
otherwise 0; used to emit the "_ArrayType_" annotation key
before "_ArraySize_" if a dimension vector turns out to
describe an ndarray
@return whether size determination completed
*/
bool get_ubjson_size_value(std::size_t& result, bool& is_ndarray, char_int_type prefix = 0)
bool get_ubjson_size_value(std::size_t& result, bool& is_ndarray, char_int_type prefix = 0, char_int_type ndarray_dtype = 0)
{
if (prefix == 0)
{
@@ -2901,8 +2906,37 @@ class binary_reader
}
}
if (JSON_HEDLEY_UNLIKELY(!sax->start_object(3)))
{
return false;
}
// the element type precedes the dimension vector (see get_ubjson_size_type)
// and is passed down as ndarray_dtype; emit it here so the annotation keys
// follow the documented _ArrayType_, _ArraySize_, _ArrayData_ order
if (ndarray_dtype != 0)
{
auto it = std::lower_bound(bjd_types_map.begin(), bjd_types_map.end(), ndarray_dtype, [](const bjd_type & p, char_int_type t)
{
return p.first < t;
});
if (JSON_HEDLEY_UNLIKELY(it == bjd_types_map.end() || it->first != ndarray_dtype))
{
auto last_token = get_token_string();
return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
exception_message(input_format, "invalid byte: 0x" + last_token, "type"), nullptr));
}
string_t type_key = "_ArrayType_";
string_t type = it->second; // sax->string() takes a reference
if (JSON_HEDLEY_UNLIKELY(!sax->key(type_key) || !sax->string(type)))
{
return false;
}
}
string_t key = "_ArraySize_";
if (JSON_HEDLEY_UNLIKELY(!sax->start_object(3) || !sax->key(key) || !sax->start_array(dim.size())))
if (JSON_HEDLEY_UNLIKELY(!sax->key(key) || !sax->start_array(dim.size())))
{
return false;
}
@@ -3003,7 +3037,7 @@ class binary_reader
exception_message(input_format, concat("expected '#' after type information; last byte: 0x", last_token), "size"), nullptr));
}
const bool is_error = get_ubjson_size_value(result.first, is_ndarray);
const bool is_error = get_ubjson_size_value(result.first, is_ndarray, 0, result.second);
// an ndarray was read here only if the flag flipped; when it was
// seeded true, get_ubjson_size_value() already rejected the nested
// dimension vector
@@ -3239,30 +3273,17 @@ class binary_reader
if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
{
size_and_type.second &= ~(static_cast<char_int_type>(1) << 8); // use bit 8 to indicate ndarray, here we remove the bit to restore the type marker
auto it = std::lower_bound(bjd_types_map.begin(), bjd_types_map.end(), size_and_type.second, [](const bjd_type & p, char_int_type t)
{
return p.first < t;
});
string_t key = "_ArrayType_";
if (JSON_HEDLEY_UNLIKELY(it == bjd_types_map.end() || it->first != size_and_type.second))
{
auto last_token = get_token_string();
return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
exception_message(input_format, "invalid byte: 0x" + last_token, "type"), nullptr));
}
string_t type = it->second; // sax->string() takes a reference
if (JSON_HEDLEY_UNLIKELY(!sax->key(key) || !sax->string(type)))
{
return false;
}
// the "_ArrayType_" and "_ArraySize_" annotation keys were already emitted by
// get_ubjson_size_value() (the type marker is known before the dimension vector
// that determines size_and_type.first is read, so it is emitted first there to
// match the documented _ArrayType_, _ArraySize_, _ArrayData_ key order)
if (size_and_type.second == 'C' || size_and_type.second == 'B')
{
size_and_type.second = 'U';
}
key = "_ArrayData_";
string_t key = "_ArrayData_";
if (JSON_HEDLEY_UNLIKELY(!sax->key(key) || !sax->start_array(size_and_type.first) ))
{
return false;
@@ -1991,9 +1991,21 @@ class binary_writer
case 'd':
{
const auto dval = el.template get<double>();
in_range = !std::isfinite(dval) ||
#ifdef __GNUC__
JSON_HEDLEY_DIAGNOSTIC_PUSH
JSON_HEDLEY_PRAGMA(GCC diagnostic ignored "-Wfloat-equal")
#endif
// a value that would be rounded (rather than exactly represented) by the
// narrowing to float is treated like an out-of-range integer element above;
// this is the same criterion write_compact_float() uses for CBOR/MessagePack
in_range = std::isnan(dval) ||
(dval >= static_cast<double>(std::numeric_limits<float>::lowest()) &&
dval <= static_cast<double>((std::numeric_limits<float>::max)()));
dval <= static_cast<double>((std::numeric_limits<float>::max)()) &&
static_cast<double>(static_cast<float>(dval)) == dval) ||
std::isinf(dval);
#ifdef __GNUC__
JSON_HEDLEY_DIAGNOSTIC_POP
#endif
break;
}
default:
+10 -25
View File
@@ -6150,10 +6150,12 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
// the same time, determine whether every added key comes
// after every common key in target's order (a precondition
// for the fast path below, which only ever appends new keys
// at the very end). Both are only needed for an object_t that
// keeps its members in insertion order, such as the one
// backing `ordered_json`; for any other object_t, the fast
// path is always taken and they are not computed.
// at the very end): for an object_t whose iteration order is
// a pure function of the key set (e.g. the default std::map,
// which always iterates in sorted key order), the order
// check further below is always true and this whole
// mechanism is effectively a no-op; it only matters for a
// reorderable object_t such as the one backing `ordered_json`.
// patch ops for keys that were added (i.e., in target but not
// in source); built here so the fast path below can reuse
// them without a second source.find() per target key. Only
@@ -6177,32 +6179,15 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
}
else
{
#ifdef JSON_HEDLEY_MSVC_VERSION
#pragma warning(push )
#pragma warning(disable : 4127) // ignore warning to replace if with if constexpr
#endif
if (detail::is_ordered_map<object_t>::value)
common_keys_target_order.push_back(it.key());
if (seen_new_key)
{
common_keys_target_order.push_back(it.key());
if (seen_new_key)
{
new_keys_form_suffix = false;
}
new_keys_form_suffix = false;
}
#ifdef JSON_HEDLEY_MSVC_VERSION
#pragma warning( pop )
#endif
}
}
// Only an object type that keeps its members in insertion
// order, such as nlohmann::ordered_map, can need reordering:
// patch() appends a new member at the end of such an object.
// Any other object type places its members itself - std::map
// in key order, a hash map in an order its operator== ignores -
// so a member-by-member diff always reproduces target there.
if (!detail::is_ordered_map<object_t>::value
|| (common_keys_source_order == common_keys_target_order && new_keys_form_suffix))
if (common_keys_source_order == common_keys_target_order && new_keys_form_suffix)
{
// fast path: order of common keys already matches (or the
// object_t's iteration order does not depend on
+66 -48
View File
@@ -15495,10 +15495,15 @@ class binary_reader
is_ndarray can only return `true` when its initial value
is `false`
@param[in] prefix type marker if already read, otherwise set to 0
@param[in] ndarray_dtype the element type marker of the enclosing bjdata ndarray if
already known (it precedes the dimension vector read here),
otherwise 0; used to emit the "_ArrayType_" annotation key
before "_ArraySize_" if a dimension vector turns out to
describe an ndarray
@return whether size determination completed
*/
bool get_ubjson_size_value(std::size_t& result, bool& is_ndarray, char_int_type prefix = 0)
bool get_ubjson_size_value(std::size_t& result, bool& is_ndarray, char_int_type prefix = 0, char_int_type ndarray_dtype = 0)
{
if (prefix == 0)
{
@@ -15668,8 +15673,37 @@ class binary_reader
}
}
if (JSON_HEDLEY_UNLIKELY(!sax->start_object(3)))
{
return false;
}
// the element type precedes the dimension vector (see get_ubjson_size_type)
// and is passed down as ndarray_dtype; emit it here so the annotation keys
// follow the documented _ArrayType_, _ArraySize_, _ArrayData_ order
if (ndarray_dtype != 0)
{
auto it = std::lower_bound(bjd_types_map.begin(), bjd_types_map.end(), ndarray_dtype, [](const bjd_type & p, char_int_type t)
{
return p.first < t;
});
if (JSON_HEDLEY_UNLIKELY(it == bjd_types_map.end() || it->first != ndarray_dtype))
{
auto last_token = get_token_string();
return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
exception_message(input_format, "invalid byte: 0x" + last_token, "type"), nullptr));
}
string_t type_key = "_ArrayType_";
string_t type = it->second; // sax->string() takes a reference
if (JSON_HEDLEY_UNLIKELY(!sax->key(type_key) || !sax->string(type)))
{
return false;
}
}
string_t key = "_ArraySize_";
if (JSON_HEDLEY_UNLIKELY(!sax->start_object(3) || !sax->key(key) || !sax->start_array(dim.size())))
if (JSON_HEDLEY_UNLIKELY(!sax->key(key) || !sax->start_array(dim.size())))
{
return false;
}
@@ -15770,7 +15804,7 @@ class binary_reader
exception_message(input_format, concat("expected '#' after type information; last byte: 0x", last_token), "size"), nullptr));
}
const bool is_error = get_ubjson_size_value(result.first, is_ndarray);
const bool is_error = get_ubjson_size_value(result.first, is_ndarray, 0, result.second);
// an ndarray was read here only if the flag flipped; when it was
// seeded true, get_ubjson_size_value() already rejected the nested
// dimension vector
@@ -16006,30 +16040,17 @@ class binary_reader
if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
{
size_and_type.second &= ~(static_cast<char_int_type>(1) << 8); // use bit 8 to indicate ndarray, here we remove the bit to restore the type marker
auto it = std::lower_bound(bjd_types_map.begin(), bjd_types_map.end(), size_and_type.second, [](const bjd_type & p, char_int_type t)
{
return p.first < t;
});
string_t key = "_ArrayType_";
if (JSON_HEDLEY_UNLIKELY(it == bjd_types_map.end() || it->first != size_and_type.second))
{
auto last_token = get_token_string();
return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
exception_message(input_format, "invalid byte: 0x" + last_token, "type"), nullptr));
}
string_t type = it->second; // sax->string() takes a reference
if (JSON_HEDLEY_UNLIKELY(!sax->key(key) || !sax->string(type)))
{
return false;
}
// the "_ArrayType_" and "_ArraySize_" annotation keys were already emitted by
// get_ubjson_size_value() (the type marker is known before the dimension vector
// that determines size_and_type.first is read, so it is emitted first there to
// match the documented _ArrayType_, _ArraySize_, _ArrayData_ key order)
if (size_and_type.second == 'C' || size_and_type.second == 'B')
{
size_and_type.second = 'U';
}
key = "_ArrayData_";
string_t key = "_ArrayData_";
if (JSON_HEDLEY_UNLIKELY(!sax->key(key) || !sax->start_array(size_and_type.first) ))
{
return false;
@@ -22321,9 +22342,21 @@ class binary_writer
case 'd':
{
const auto dval = el.template get<double>();
in_range = !std::isfinite(dval) ||
#ifdef __GNUC__
JSON_HEDLEY_DIAGNOSTIC_PUSH
JSON_HEDLEY_PRAGMA(GCC diagnostic ignored "-Wfloat-equal")
#endif
// a value that would be rounded (rather than exactly represented) by the
// narrowing to float is treated like an out-of-range integer element above;
// this is the same criterion write_compact_float() uses for CBOR/MessagePack
in_range = std::isnan(dval) ||
(dval >= static_cast<double>(std::numeric_limits<float>::lowest()) &&
dval <= static_cast<double>((std::numeric_limits<float>::max)()));
dval <= static_cast<double>((std::numeric_limits<float>::max)()) &&
static_cast<double>(static_cast<float>(dval)) == dval) ||
std::isinf(dval);
#ifdef __GNUC__
JSON_HEDLEY_DIAGNOSTIC_POP
#endif
break;
}
default:
@@ -32231,10 +32264,12 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
// the same time, determine whether every added key comes
// after every common key in target's order (a precondition
// for the fast path below, which only ever appends new keys
// at the very end). Both are only needed for an object_t that
// keeps its members in insertion order, such as the one
// backing `ordered_json`; for any other object_t, the fast
// path is always taken and they are not computed.
// at the very end): for an object_t whose iteration order is
// a pure function of the key set (e.g. the default std::map,
// which always iterates in sorted key order), the order
// check further below is always true and this whole
// mechanism is effectively a no-op; it only matters for a
// reorderable object_t such as the one backing `ordered_json`.
// patch ops for keys that were added (i.e., in target but not
// in source); built here so the fast path below can reuse
// them without a second source.find() per target key. Only
@@ -32258,32 +32293,15 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
}
else
{
#ifdef JSON_HEDLEY_MSVC_VERSION
#pragma warning(push )
#pragma warning(disable : 4127) // ignore warning to replace if with if constexpr
#endif
if (detail::is_ordered_map<object_t>::value)
common_keys_target_order.push_back(it.key());
if (seen_new_key)
{
common_keys_target_order.push_back(it.key());
if (seen_new_key)
{
new_keys_form_suffix = false;
}
new_keys_form_suffix = false;
}
#ifdef JSON_HEDLEY_MSVC_VERSION
#pragma warning( pop )
#endif
}
}
// Only an object type that keeps its members in insertion
// order, such as nlohmann::ordered_map, can need reordering:
// patch() appends a new member at the end of such an object.
// Any other object type places its members itself - std::map
// in key order, a hash map in an order its operator== ignores -
// so a member-by-member diff always reproduces target there.
if (!detail::is_ordered_map<object_t>::value
|| (common_keys_source_order == common_keys_target_order && new_keys_form_suffix))
if (common_keys_source_order == common_keys_target_order && new_keys_form_suffix)
{
// fast path: order of common keys already matches (or the
// object_t's iteration order does not depend on
+42 -4
View File
@@ -11,6 +11,7 @@
#define JSON_TESTS_PRIVATE
#include <nlohmann/json.hpp>
using nlohmann::json;
using ordered_json = nlohmann::ordered_json;
#include <algorithm>
#include <climits>
@@ -2294,29 +2295,33 @@ TEST_CASE("BJData")
SECTION("start_array() in ndarray _ArraySize_")
{
// _ArrayType_ (2 events: key + string) is now emitted before
// _ArraySize_ (see GitHub issue #5661), which shifts the events
// below later by the same 2 events
std::vector<uint8_t> const v = {'[', '$', 'i', '#', '[', '$', 'i', '#', 'i', 2, 2, 1, 1, 2};
SaxCountdown scp(2);
SaxCountdown scp(4);
CHECK_FALSE(json::sax_parse(v, &scp, json::input_format_t::bjdata));
}
SECTION("number_integer() in ndarray _ArraySize_")
{
std::vector<uint8_t> const v = {'[', '$', 'U', '#', '[', '$', 'i', '#', 'i', 2, 2, 1, 1, 2};
SaxCountdown scp(3);
SaxCountdown scp(5);
CHECK_FALSE(json::sax_parse(v, &scp, json::input_format_t::bjdata));
}
SECTION("key() in ndarray _ArrayType_")
{
// _ArrayType_ is emitted right after start_object(), before _ArraySize_
std::vector<uint8_t> const v = {'[', '$', 'U', '#', '[', '$', 'U', '#', 'i', 2, 2, 2, 1, 2, 3, 4};
SaxCountdown scp(6);
SaxCountdown scp(1);
CHECK_FALSE(json::sax_parse(v, &scp, json::input_format_t::bjdata));
}
SECTION("string() in ndarray _ArrayType_")
{
std::vector<uint8_t> const v = {'[', '$', 'U', '#', '[', '$', 'U', '#', 'i', 2, 2, 2, 1, 2, 3, 4};
SaxCountdown scp(7);
SaxCountdown scp(2);
CHECK_FALSE(json::sax_parse(v, &scp, json::input_format_t::bjdata));
}
@@ -2919,6 +2924,22 @@ TEST_CASE("BJData")
CHECK(out_single.at(0) == '{');
CHECK(json::from_bjdata(out_single) == j_single);
// a double element that is finite and within the range of "single"
// but is not exactly representable as a float, so narrowing it would
// silently round it (0.1 is read back as 0.10000000149011612); this,
// like the overflow case above, falls back to a plain object (see
// GitHub issue #5661)
json const j_single_rounded = json({{"_ArrayType_", "single"}, {"_ArraySize_", {2, 1}}, {"_ArrayData_", {1.5, 0.1}}});
const auto out_single_rounded = json::to_bjdata(j_single_rounded);
CHECK(out_single_rounded.at(0) == '{');
CHECK(json::from_bjdata(out_single_rounded) == j_single_rounded);
// a double element that underflows to 0 when narrowed to "single"
json const j_single_underflow = json({{"_ArrayType_", "single"}, {"_ArraySize_", {2, 1}}, {"_ArrayData_", {1.5, 1e-300}}});
const auto out_single_underflow = json::to_bjdata(j_single_underflow);
CHECK(out_single_underflow.at(0) == '{');
CHECK(json::from_bjdata(out_single_underflow) == j_single_underflow);
// in-range boundary values still use the compact ndarray encoding
json const j_uint8_ok = json({{"_ArrayType_", "uint8"}, {"_ArraySize_", {2, 1}}, {"_ArrayData_", {0, 255}}});
CHECK(json::to_bjdata(j_uint8_ok) == std::vector<uint8_t>({'[', '$', 'U', '#', '[', 'i', 2, 'i', 1, ']', 0, 255}));
@@ -2932,6 +2953,23 @@ TEST_CASE("BJData")
CHECK(json::from_bjdata(out_single_ok) == json({{"_ArrayType_", "single"}, {"_ArraySize_", {2, 1}}, {"_ArrayData_", {1.5f, -1.5f}}}));
}
SECTION("ndarray annotation keys are read back in the documented order")
{
// from_bjdata() must emit the annotation object's keys in the order
// used throughout the documentation, _ArrayType_, _ArraySize_,
// _ArrayData_: the type marker precedes the dimension vector on the
// wire (see get_ubjson_size_type()), so it is known, and emitted,
// before _ArraySize_. For a plain json this key order is invisible
// (its comparison ignores it), but for an ordered_json it is not (see
// GitHub issue #5661).
const ordered_json o = ordered_json::parse(R"({"_ArrayType_":"uint8","_ArraySize_":[2,2],"_ArrayData_":[1,2,3,4]})");
const auto packed = ordered_json::to_bjdata(o);
CHECK(packed.at(0) == '[');
const ordered_json o_back = ordered_json::from_bjdata(packed);
CHECK(o_back == o);
CHECK(o_back.dump() == o.dump());
}
SECTION("ndarray that would not be read back as an annotated object stays as object")
{
// the reader only restores an annotated object from an ND-array
-52
View File
@@ -1752,58 +1752,6 @@ TEST_CASE("JSON patch - diff emits array removals in descending index order")
}
}
TEST_CASE("JSON patch - diff() takes the fast path for non-reorderable object types (regression #5639)")
{
// #5465 added an order check to diff()'s object handling so a
// member-by-member diff is only used when it would also reproduce
// target's member *order* -- needed for ordered_json, whose object_t
// keeps insertion order and whose patch() "add" op appends a new
// member at the end. For json's default object_t (std::map, which
// orders members by key regardless of insertion history), that check
// could still fail: a new key that sorts before an existing common key
// makes target's iteration interleave the new key between common keys,
// even though nothing else about the object changed. That sent the
// whole object through the slow (remove-every-member,
// re-add-every-member) path instead of the minimal one.
SECTION("json: added key sorts before an existing common key")
{
const json source = {{"a", 1}, {"c", {{"x", 1}, {"y", 2}}}};
const json target = {{"a", 1}, {"b", 0}, {"c", {{"x", 1}, {"y", 2}}}};
const json patch = json::diff(source, target);
// only the new key is added; "a" and "c" are left alone instead of
// being removed and re-added
const json expected = R"([{"op": "add", "path": "/b", "value": 0}])"_json;
CHECK(patch == expected);
CHECK(source.patch(patch) == target);
}
SECTION("ordered_json: reordering behavior from #5465 is unchanged")
{
using nlohmann::ordered_json;
// same key/value shape as the json case above, but for ordered_json
// the *target*'s member order must be reproduced, so the slow path
// is still required here.
ordered_json source;
source["a"] = 1;
source["c"] = ordered_json{{"x", 1}, {"y", 2}};
ordered_json target;
target["a"] = 1;
target["b"] = 0;
target["c"] = ordered_json{{"x", 1}, {"y", 2}};
const ordered_json patch = ordered_json::diff(source, target);
// unlike the json case: every member is still removed and re-added
// so the result ends up in target's order (2 removes + 3 adds)
CHECK(patch.size() == 5);
CHECK(source.patch(patch) == target);
}
}
TEST_CASE("JSON patch - every operation on ordered_json")
{
using nlohmann::ordered_json;