Mark the parse callback in the new test noexcept

GCC's -Wnoexcept (part of the ci_test_gcc warning set) rejects the
lambda passed to parse() because it cannot throw but is not declared
noexcept, matching the existing callback in unit-regression2.cpp.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

.github/workflows/ubuntu.yml

@@ -100,7 +100,7 @@ jobs:
     container: ubuntu:focal
     strategy:
       matrix:
-        target: [ci_cmake_flags, ci_test_diagnostics, ci_test_diagnostic_positions, ci_test_noexceptions, ci_test_noimplicitconversions, ci_test_legacycomparison, ci_test_noglobaludls, ci_test_simdutf, ci_test_no_thread_local]
+        target: [ci_cmake_flags, ci_test_diagnostics, ci_test_diagnostic_positions, ci_test_noexceptions, ci_test_noimplicitconversions, ci_test_legacycomparison, ci_test_noglobaludls, ci_test_simdutf]
     steps:
       - name: Install build-essential
         run: apt-get update ; apt-get install -y build-essential unzip wget git libssl-dev

.github/workflows/windows.yml

@@ -158,10 +158,6 @@ jobs:
       # to fit: IMAGE_REL_AMD64_SECREL against `.debug_line'" because the
       # MinGW linker cannot relocate the debug sections this test produces.
       # The tests are only built and run here, so the debug info is not used.
-      # Do not add -O1 here to shrink the objects further: it does make them
-      # link, but the binaries clang 11.0.1 and clang 18.1.8 then produce crash
-      # before doctest prints its first line - 39 of 102 tests on clang 18.
-      # Keep the objects small by splitting the test files instead.
       - name: Run CMake
         run: cmake -S . -B build ^
              -DCMAKE_CXX_COMPILER="C:/Program Files/LLVM/bin/clang++.exe" ^

cmake/ci.cmake

@@ -260,25 +260,6 @@ add_custom_target(ci_test_noglobaludls
     COMMENT "Compile and test with global UDLs disabled"
 )
 
-###############################################################################
-# Disable thread-local storage.
-###############################################################################
-
-# Without thread-local storage, the copy constructor cannot bound its descent
-# and copies every object and array without the call stack. That path is
-# otherwise only reached by values nested deeper than the bound, so this target
-# is what runs the whole test suite through it.
-add_custom_target(ci_test_no_thread_local
-    COMMAND ${CMAKE_COMMAND}
-    -DCMAKE_BUILD_TYPE=Debug -GNinja
-    -DJSON_BuildTests=ON
-    -DCMAKE_CXX_FLAGS=-DJSON_NO_THREAD_LOCAL
-    -S${PROJECT_SOURCE_DIR} -B${PROJECT_BINARY_DIR}/build_no_thread_local
-    COMMAND ${CMAKE_COMMAND} --build ${PROJECT_BINARY_DIR}/build_no_thread_local
-    COMMAND cd ${PROJECT_BINARY_DIR}/build_no_thread_local && ${CMAKE_CTEST_COMMAND} --parallel ${N} --output-on-failure
-    COMMENT "Compile and test without thread-local storage"
-)
-
 ###############################################################################
 # Coverage.
 ###############################################################################

docs/mkdocs/docs/api/basic_json/swap.md

@@ -34,10 +34,14 @@ void swap(typename binary_t::container_type& other);
 ```
 
 1. Exchanges the contents of the JSON value with those of `other`. Does not invoke any move, copy, or swap operations on
-   individual elements. All iterators and references remain valid. The past-the-end iterator is invalidated. 
+   individual elements. All iterators and references remain valid. The past-the-end iterator is invalidated. If macro
+   [`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md) is defined to `#!cpp 1`, the
+   [`start_pos()`](start_pos.md)/[`end_pos()`](end_pos.md) diagnostic positions are exchanged along with the value.
 2. Exchanges the contents of the JSON value from `left` with those of `right`. Does not invoke any move, copy, or swap
    operations on individual elements. All iterators and references remain valid. The past-the-end iterator is
-   invalidated. Implemented as a friend function callable via ADL.
+   invalidated. Implemented as a friend function callable via ADL. If macro
+   [`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md) is defined to `#!cpp 1`, the
+   [`start_pos()`](start_pos.md)/[`end_pos()`](end_pos.md) diagnostic positions are exchanged along with the value.
 3. Exchanges the contents of a JSON array with those of `other`. Does not invoke any move, copy, or swap operations on
    individual elements. All iterators and references remain valid. The past-the-end iterator is invalidated. 
 4. Exchanges the contents of a JSON object with those of `other`. Does not invoke any move, copy, or swap operations on

docs/mkdocs/docs/api/macros/index.md

@@ -22,7 +22,6 @@ header. See also the [macro overview page](../../features/macros.md).
 - [**JSON_HAS_STD_FORMAT**](json_has_std_format.md) - control `std::format`/`std::formatter` support
 - [**JSON_HAS_THREE_WAY_COMPARISON**](json_has_three_way_comparison.md) - control 3-way comparison support
 - [**JSON_NO_IO**](json_no_io.md) - switch off functions relying on certain C++ I/O headers
-- [**JSON_NO_THREAD_LOCAL**](json_no_thread_local.md) - switch off the use of `thread_local` storage
 - [**JSON_SKIP_UNSUPPORTED_COMPILER_CHECK**](json_skip_unsupported_compiler_check.md) - do not warn about unsupported compilers
 - [**JSON_USE_GLOBAL_UDLS**](json_use_global_udls.md) - place user-defined string literals (UDLs) into the global namespace
 - [**JSON_USE_SIMDUTF**](json_use_simdutf.md) - use the simdutf library to accelerate UTF-8 validation

docs/mkdocs/docs/api/macros/json_no_thread_local.md

@@ -1,47 +0,0 @@
-# JSON_NO_THREAD_LOCAL
-
-```cpp
-#define JSON_NO_THREAD_LOCAL
-```
-
-When defined, the library does not use `#!cpp thread_local` storage. This is relevant for the few environments whose
-toolchain does not support it.
-
-The copy constructor copies the first levels of a value by copying the containers, which copy their elements, and
-completes whatever is nested deeper than that without the call stack, so that copying a value cannot exhaust the stack
-however deeply it is nested. It counts the levels it has descended into in a `#!cpp thread_local` variable, as a counter
-shared between threads would be raced.
-
-Without that counter, no descent can be bounded safely, so objects and arrays are copied without the call stack right
-away. Copying keeps working exactly as it does otherwise - the same values come out, and deeply nested values are copied
-just as safely - but copying is slower, because the containers no longer copy themselves. Copying the benchmark
-documents takes 9% (`canada.json`) to 34% (`twitter.json`) longer; values built mostly from objects are affected the
-most.
-
-## Default definition
-
-By default, `#!cpp JSON_NO_THREAD_LOCAL` is not defined.
-
-```cpp
-#undef JSON_NO_THREAD_LOCAL
-```
-
-The library defines it by itself for Clang targeting MinGW, which does not survive the `#!cpp thread_local` storage:
-copying a value segfaults there, with both old and current Clang versions, while GCC targeting MinGW is unaffected.
-
-## Examples
-
-??? example
-
-    The code below forces the library not to use `#!cpp thread_local` storage.
-
-    ```cpp
-    #define JSON_NO_THREAD_LOCAL 1
-    #include <nlohmann/json.hpp>
-
-    ...
-    ```
-
-## Version history
-
-- Added in version 3.12.1.

docs/mkdocs/docs/features/macros.md

@@ -91,13 +91,6 @@ security reasons (e.g., Intel Software Guard Extensions (SGX)).
 
 See [full documentation of `JSON_NO_IO`](../api/macros/json_no_io.md).
 
-## `JSON_NO_THREAD_LOCAL`
-
-When defined, the library does not use `#!cpp thread_local` storage. Copying a value then always avoids the call stack
-rather than descending into a bounded number of levels first, which is slower but yields the same values.
-
-See [full documentation of `JSON_NO_THREAD_LOCAL`](../api/macros/json_no_thread_local.md).
-
 ## `JSON_SKIP_LIBRARY_VERSION_CHECK`
 
 When defined, the library will not create a compiler warning when a different version of the library was already

docs/mkdocs/mkdocs.yml

@@ -291,7 +291,6 @@ nav:
           - 'JSON_HAS_THREE_WAY_COMPARISON': api/macros/json_has_three_way_comparison.md
           - 'JSON_NOEXCEPTION': api/macros/json_noexception.md
           - 'JSON_NO_IO': api/macros/json_no_io.md
-          - 'JSON_NO_THREAD_LOCAL': api/macros/json_no_thread_local.md
           - 'JSON_SKIP_LIBRARY_VERSION_CHECK': api/macros/json_skip_library_version_check.md
           - 'JSON_SKIP_UNSUPPORTED_COMPILER_CHECK': api/macros/json_skip_unsupported_compiler_check.md
           - 'JSON_USE_GLOBAL_UDLS': api/macros/json_use_global_udls.md

include/nlohmann/detail/input/binary_reader.hpp

@@ -1434,7 +1434,7 @@ class binary_reader
                 // a copy, not a reference: it must stay valid across the
                 // pop_back() below, which destroys the container_stack element
                 // it would otherwise alias
-                container_frame top = container_stack.back();
+                const container_frame top = container_stack.back();
                 bool at_end = false;
 
                 if (top.remaining != npos)
@@ -2211,7 +2211,7 @@ class binary_reader
             // would otherwise alias.
             for (;;)
             {
-                container_frame top = container_stack.back();
+                const container_frame top = container_stack.back();
 
                 if (top.remaining != npos)
                 {

include/nlohmann/detail/input/json_sax.hpp

@@ -8,6 +8,7 @@
 
 #pragma once
 
+#include <algorithm> // min
 #include <cstddef>
 #include <string> // string
 #include <type_traits> // enable_if_t
@@ -17,6 +18,7 @@
 #include <nlohmann/detail/exceptions.hpp>
 #include <nlohmann/detail/input/lexer.hpp>
 #include <nlohmann/detail/macro_scope.hpp>
+#include <nlohmann/detail/meta/cpp_future.hpp>
 #include <nlohmann/detail/string_concat.hpp>
 NLOHMANN_JSON_NAMESPACE_BEGIN
 
@@ -150,6 +152,29 @@ constexpr std::size_t unknown_size()
     return (std::numeric_limits<std::size_t>::max)();
 }
 
+/*!
+@brief reserve capacity for @a len elements in array @a arr
+
+Reserving upfront avoids repeated reallocations while the elements are added,
+but the reservation is capped so a bogus/hostile length (which is not bounded
+by max_size(), unlike e.g. std::vector) cannot trigger an oversized allocation
+for a small or truncated input.
+
+The overload below is selected for array types without reserve() (e.g.,
+std::deque), which are then left untouched.
+*/
+template<typename ArrayType>
+auto reserve_array(ArrayType& arr, std::size_t len, priority_tag<1> /*unused*/)
+-> decltype(arr.reserve(len), void())
+{
+    constexpr std::size_t reserve_cap = 16384;
+    arr.reserve((std::min)(len, reserve_cap));
+}
+
+template<typename ArrayType>
+inline void reserve_array(ArrayType& /*arr*/, std::size_t /*len*/, priority_tag<0> /*unused*/)
+{}
+
 /*!
 @brief SAX implementation to create a JSON value from SAX events
 
@@ -305,6 +330,11 @@ class json_sax_dom_parser
             JSON_THROW(out_of_range::create(408, concat("excessive array size: ", std::to_string(len)), ref_stack.back()));
         }
 
+        if (len != detail::unknown_size())
+        {
+            reserve_array(*ref_stack.back()->m_data.m_value.array, len, priority_tag<1> {});
+        }
+
         return true;
     }
 
@@ -683,6 +713,11 @@ class json_sax_dom_callback_parser
             {
                 JSON_THROW(out_of_range::create(408, concat("excessive array size: ", std::to_string(len)), ref_stack.back()));
             }
+
+            if (len != detail::unknown_size())
+            {
+                reserve_array(*ref_stack.back()->m_data.m_value.array, len, priority_tag<1> {});
+            }
         }
 
         return true;

include/nlohmann/detail/macro_scope.hpp

@@ -186,15 +186,6 @@
     #define JSON_NO_UNIQUE_ADDRESS
 #endif
 
-// Clang targeting MinGW does not survive the thread_local storage the copy
-// constructor uses to bound its descent: every test that copies a value
-// segfaults with clang 11.0.1 and clang 18.1.8, while the same tests pass with
-// GCC targeting MinGW and with every other toolchain the library is tested on.
-// Copying works the same way without the counter, only more slowly.
-#if !defined(JSON_NO_THREAD_LOCAL) && defined(__clang__) && defined(__MINGW32__)
-    #define JSON_NO_THREAD_LOCAL 1
-#endif
-
 // disable documentation warnings on clang
 #if defined(__clang__)
     #pragma clang diagnostic push

include/nlohmann/json.hpp

@@ -28,14 +28,14 @@
     #pragma GCC diagnostic ignored "-Wignored-attributes"
 #endif
 
-#include <algorithm> // all_of, find, for_each, none_of
+#include <algorithm> // all_of, find, for_each
 #include <cstddef> // nullptr_t, ptrdiff_t, size_t
 #include <functional> // hash, less
 #include <initializer_list> // initializer_list
 #ifndef JSON_NO_IO
     #include <iosfwd> // istream, ostream
 #endif  // JSON_NO_IO
-#include <iterator> // make_move_iterator, random_access_iterator_tag
+#include <iterator> // random_access_iterator_tag
 #include <memory> // unique_ptr
 #include <string> // string, stoi, to_string
 #include <utility> // declval, forward, move, pair, swap
@@ -822,351 +822,6 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         return j;
     }
 
-#ifndef JSON_NO_THREAD_LOCAL
-    /// the number of levels an operation descends into before it finishes the
-    /// value below it without the call stack
-    static constexpr std::uint8_t nesting_depth_limit()
-    {
-        return 128;
-    }
-
-    /*!
-    @brief how many levels the operation going on in this thread has descended into
-
-    Copying a value and comparing two values share this count. The library never
-    nests one inside the other - copying a value does not compare one, and
-    comparing two values does not copy them - and where user code nests them
-    anyway, sharing the count only ends a descent sooner than it had to, which
-    costs a little speed and is never wrong.
-
-    A byte is enough: the count never exceeds the limit by more than the single
-    level that notices the limit has been reached.
-    */
-    static std::uint8_t& nesting_depth() noexcept
-    {
-        static thread_local std::uint8_t depth = 0; // NOLINT(misc-use-internal-linkage)
-        return depth;
-    }
-#endif
-
-    /*!
-    @brief counts one level of a bounded descent for as long as it runs, and
-           reports whether the descent was still within the limit when it began
-
-    Looks the count up and tests it against the limit itself, rather than
-    leaving that to the caller: either way it is reached exactly once, so
-    there is nothing to be gained by making the caller do it.
-
-    Does nothing and is never @ref okay without thread-local storage, where no
-    descent can be bounded at all: a caller that only descends while this says
-    it may always ends up finishing without the call stack, exactly as if every
-    value were nested past the limit.
-    */
-    class nesting_depth_guard
-    {
-      public:
-        nesting_depth_guard() noexcept
-#ifdef JSON_NO_THREAD_LOCAL
-            : m_okay(false)
-#else
-            : m_okay(nesting_depth() < nesting_depth_limit())
-#endif
-        {
-#ifndef JSON_NO_THREAD_LOCAL
-            ++nesting_depth();
-#endif
-        }
-
-        ~nesting_depth_guard()
-        {
-#ifndef JSON_NO_THREAD_LOCAL
-            --nesting_depth();
-#endif
-        }
-
-        nesting_depth_guard(const nesting_depth_guard&) = delete;
-        nesting_depth_guard& operator=(const nesting_depth_guard&) = delete;
-        nesting_depth_guard(nesting_depth_guard&&) = delete;
-        nesting_depth_guard& operator=(nesting_depth_guard&&) = delete;
-
-        bool okay() const noexcept
-        {
-            return m_okay;
-        }
-
-      private:
-        bool m_okay;
-    };
-
-    /// an entry of the iterative deep copy's worklist: a structured value and
-    /// the value that is to become its copy
-    using copy_worklist_t = std::vector<std::pair<const basic_json*, basic_json*>>;
-
-    /// scratch space to build the key skeleton of an object copy in one go
-    using copy_scratch_t = std::vector<std::pair<typename object_t::key_type, basic_json>>;
-
-    /// @brief copy everything of @a src into @a dst but its type and value
-    static void copy_metadata(const basic_json& src, basic_json& dst)
-    {
-        // a custom base class is only required to be copy-constructible and
-        // move-assignable, so the copy has to go through a temporary
-        static_cast<json_base_class_t&>(dst) = json_base_class_t(static_cast<const json_base_class_t&>(src));
-
-#if JSON_DIAGNOSTIC_POSITIONS
-        dst.start_position = src.start_position;
-        dst.end_position = src.end_position;
-#else
-        static_cast<void>(src);
-        static_cast<void>(dst);
-#endif
-    }
-
-    /*!
-    @brief copy the value of @a src into @a dst, which must not be structured
-
-    Objects and arrays are left alone: creating those is the one thing the copy
-    constructor and @ref copy_shallow do differently from one another, and it is
-    the reason copying a value can descend at all.
-    */
-    /// @note inlined on purpose: both callers have already told an object or an
-    ///       array apart from the rest, and letting the compiler fold that test
-    ///       into this switch is worth a few percent when copying a value made
-    ///       mostly of numbers
-    JSON_HEDLEY_ALWAYS_INLINE
-    static void copy_leaf_value(const basic_json& src, basic_json& dst)
-    {
-        switch (src.m_data.m_type)
-        {
-            case value_t::string:
-            {
-                dst.m_data.m_value = *src.m_data.m_value.string;
-                break;
-            }
-
-            case value_t::binary:
-            {
-                dst.m_data.m_value = *src.m_data.m_value.binary;
-                break;
-            }
-
-            case value_t::boolean:
-            {
-                dst.m_data.m_value = src.m_data.m_value.boolean;
-                break;
-            }
-
-            case value_t::number_integer:
-            {
-                dst.m_data.m_value = src.m_data.m_value.number_integer;
-                break;
-            }
-
-            case value_t::number_unsigned:
-            {
-                dst.m_data.m_value = src.m_data.m_value.number_unsigned;
-                break;
-            }
-
-            case value_t::number_float:
-            {
-                dst.m_data.m_value = src.m_data.m_value.number_float;
-                break;
-            }
-
-            case value_t::object:
-            case value_t::array:
-            case value_t::null:
-            case value_t::discarded:
-            default:
-                break;
-        }
-    }
-
-    /*!
-    @brief copy everything of @a src into the null value @a dst but the children
-
-    Objects and arrays are not copied here; they are appended to @a worklist to
-    be created later by @ref copy_iteratively. Until that happens, @a dst remains
-    a null value, so that a partially built copy can be destroyed at any point
-    without ever violating the class invariants.
-    */
-    static void copy_shallow(const basic_json& src, basic_json& dst, copy_worklist_t& worklist)
-    {
-        copy_metadata(src, dst);
-
-        if (src.m_data.m_type == value_t::object || src.m_data.m_type == value_t::array)
-        {
-            // defer: dst stays a null value until its container exists
-            worklist.emplace_back(&src, &dst);
-            return;
-        }
-
-        copy_leaf_value(src, dst);
-
-        // only now that the value exists may the type be set: had the creation
-        // of the value thrown, dst would have been left as a valid null value
-        dst.m_data.m_type = src.m_data.m_type;
-    }
-
-    /// @brief create the copy of the array @a src in @a dst
-    /// @note structured elements are appended to @a worklist instead
-    static void copy_array_level(const basic_json& src, basic_json& dst, copy_worklist_t& worklist)
-    {
-        const array_t& src_array = *src.m_data.m_value.array;
-
-        // create all elements up front: growing the array afterwards could
-        // invalidate the pointers that are handed to the worklist
-        dst.m_data.m_value.array = create<array_t>(src_array.size(), basic_json());
-
-        auto dst_it = dst.m_data.m_value.array->begin();
-        for (auto src_it = src_array.cbegin(); src_it != src_array.cend(); ++src_it, ++dst_it)
-        {
-            copy_shallow(*src_it, *dst_it, worklist);
-        }
-    }
-
-    /// @brief create the copy of the object @a src in @a dst
-    /// @note structured values are appended to @a worklist instead
-    static void copy_object_level(const basic_json& src, basic_json& dst,
-                                  copy_worklist_t& worklist, copy_scratch_t& scratch)
-    {
-        const object_t& src_object = *src.m_data.m_value.object;
-
-        // build the complete key skeleton and hand it to the object's range
-        // constructor: adding the keys one by one would be quadratic for object
-        // types that are backed by a vector, such as nlohmann::ordered_map
-        scratch.clear();
-        scratch.reserve(src_object.size());
-        for (const auto& element : src_object)
-        {
-            scratch.emplace_back(element.first, basic_json());
-        }
-
-        dst.m_data.m_value.object = create<object_t>(std::make_move_iterator(scratch.begin()),
-                                    std::make_move_iterator(scratch.end()));
-        scratch.clear();
-
-        // pair every value of the copy with its counterpart in the original;
-        // both are enumerated in the same order for every object type with a
-        // deterministic order, so the lookup is only needed for exotic ones
-        auto src_it = src_object.cbegin();
-        for (auto& element : *dst.m_data.m_value.object)
-        {
-            if (JSON_HEDLEY_LIKELY(src_it != src_object.cend() && src_it->first == element.first))
-            {
-                copy_shallow(src_it->second, element.second, worklist);
-                ++src_it;
-            }
-            else
-            {
-                const auto found = src_object.find(element.first);
-                JSON_ASSERT(found != src_object.cend());
-                copy_shallow(found->second, element.second, worklist);
-            }
-        }
-    }
-
-    /*!
-    @brief deep-copy the object or array @a src into this value without recursing
-
-    The values whose copy has not been created yet are kept on an explicit
-    worklist rather than on the call stack. This is only reached for values
-    nested deeper than @ref nesting_depth_limit levels, which is why it copies
-    every container by hand instead of letting the container do it: the fast
-    ways of doing so would descend into the elements and defeat the purpose.
-    */
-    void copy_iteratively(const basic_json& src)
-    {
-        copy_worklist_t worklist;
-        copy_scratch_t scratch;
-
-        const basic_json* src_value = &src;
-        basic_json* dst_value = this;
-
-        for (;;)
-        {
-            if (src_value->m_data.m_type == value_t::array)
-            {
-                copy_array_level(*src_value, *dst_value, worklist);
-            }
-            else
-            {
-                copy_object_level(*src_value, *dst_value, worklist, scratch);
-            }
-
-            // the container is complete and will not be modified again
-            dst_value->set_parents();
-
-            if (worklist.empty())
-            {
-                break;
-            }
-
-            const auto& next = worklist.back();
-            src_value = next.first;
-            dst_value = next.second;
-            worklist.pop_back();
-
-            // the value stops being a null value exactly here
-            dst_value->m_data.m_type = src_value->m_data.m_type;
-        }
-    }
-
-    /*!
-    @brief copy one level of the object or array @a src into this value
-
-    The container copies its own elements, which is the fastest way to fill it.
-    Every element that is structured itself comes back to @ref copy_structured.
-    */
-    void copy_level(const basic_json& src)
-    {
-        if (m_data.m_type == value_t::object)
-        {
-            m_data.m_value = *src.m_data.m_value.object;
-        }
-        else
-        {
-            m_data.m_value = *src.m_data.m_value.array;
-        }
-
-        set_parents();
-    }
-
-    /*!
-    @brief deep-copy the object or array @a src into this value
-
-    Copying a container copies its elements, so a value nested deeply enough
-    used to exhaust the call stack. The descent is bounded here: the first
-    @ref nesting_depth_limit levels are copied by the containers themselves, just
-    as they always were, and anything below that is copied without the call
-    stack by @ref copy_iteratively. Copying a value can therefore no longer
-    exhaust the stack, however deeply it is nested, just like destroying one
-    cannot since #1436.
-
-    Nothing has to be scanned or built by hand to reach that: a value that is
-    not nested deeper than the limit - all but a vanishing minority - is copied
-    exactly as it was before, and this whole detour costs it one counter.
-
-    @sa https://github.com/nlohmann/json/issues/5387
-    */
-    void copy_structured(const basic_json& src)
-    {
-        const nesting_depth_guard guard;
-
-        if (JSON_HEDLEY_LIKELY(guard.okay()))
-        {
-            copy_level(src);
-            return;
-        }
-
-        // Finish this value without descending any further. It is completed
-        // before this returns, so a copy made by a custom base class - or by
-        // anything else that runs while a copy is going on - is unaffected by
-        // the copy it is nested in.
-        copy_iteratively(src);
-    }
-
-
   public:
     //////////////////////////
     // JSON parser callback //
@@ -1546,15 +1201,60 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         // check of passed value is valid
         other.assert_invariant();
 
-        if (m_data.m_type == value_t::object || m_data.m_type == value_t::array)
+        switch (m_data.m_type)
         {
-            // copying the container directly would call this constructor again
-            // for every element, once per nesting level
-            copy_structured(other);
-        }
-        else
-        {
-            copy_leaf_value(other, *this);
+            case value_t::object:
+            {
+                m_data.m_value = *other.m_data.m_value.object;
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_data.m_value = *other.m_data.m_value.array;
+                break;
+            }
+
+            case value_t::string:
+            {
+                m_data.m_value = *other.m_data.m_value.string;
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                m_data.m_value = other.m_data.m_value.boolean;
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                m_data.m_value = other.m_data.m_value.number_integer;
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                m_data.m_value = other.m_data.m_value.number_unsigned;
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                m_data.m_value = other.m_data.m_value.number_float;
+                break;
+            }
+
+            case value_t::binary:
+            {
+                m_data.m_value = *other.m_data.m_value.binary;
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                break;
         }
 
         set_parents();
@@ -3876,6 +3576,11 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         std::swap(m_data.m_type, other.m_data.m_type);
         std::swap(m_data.m_value, other.m_data.m_value);
 
+#if JSON_DIAGNOSTIC_POSITIONS
+        std::swap(start_position, other.start_position);
+        std::swap(end_position, other.end_position);
+#endif
+
         set_parents();
         other.set_parents();
         assert_invariant();

single_include/nlohmann/json.hpp

@@ -28,14 +28,14 @@
     #pragma GCC diagnostic ignored "-Wignored-attributes"
 #endif
 
-#include <algorithm> // all_of, find, for_each, none_of
+#include <algorithm> // all_of, find, for_each
 #include <cstddef> // nullptr_t, ptrdiff_t, size_t
 #include <functional> // hash, less
 #include <initializer_list> // initializer_list
 #ifndef JSON_NO_IO
     #include <iosfwd> // istream, ostream
 #endif  // JSON_NO_IO
-#include <iterator> // make_move_iterator, random_access_iterator_tag
+#include <iterator> // random_access_iterator_tag
 #include <memory> // unique_ptr
 #include <string> // string, stoi, to_string
 #include <utility> // declval, forward, move, pair, swap
@@ -2564,15 +2564,6 @@ JSON_HEDLEY_DIAGNOSTIC_POP
     #define JSON_NO_UNIQUE_ADDRESS
 #endif
 
-// Clang targeting MinGW does not survive the thread_local storage the copy
-// constructor uses to bound its descent: every test that copies a value
-// segfaults with clang 11.0.1 and clang 18.1.8, while the same tests pass with
-// GCC targeting MinGW and with every other toolchain the library is tested on.
-// Copying works the same way without the counter, only more slowly.
-#if !defined(JSON_NO_THREAD_LOCAL) && defined(__clang__) && defined(__MINGW32__)
-    #define JSON_NO_THREAD_LOCAL 1
-#endif
-
 // disable documentation warnings on clang
 #if defined(__clang__)
     #pragma clang diagnostic push
@@ -7901,6 +7892,7 @@ NLOHMANN_JSON_NAMESPACE_END
 
 
 
+#include <algorithm> // min
 #include <cstddef>
 #include <string> // string
 #include <type_traits> // enable_if_t
@@ -10738,6 +10730,8 @@ NLOHMANN_JSON_NAMESPACE_END
 
 // #include <nlohmann/detail/macro_scope.hpp>
 
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+
 // #include <nlohmann/detail/string_concat.hpp>
 
 NLOHMANN_JSON_NAMESPACE_BEGIN
@@ -10872,6 +10866,29 @@ constexpr std::size_t unknown_size()
     return (std::numeric_limits<std::size_t>::max)();
 }
 
+/*!
+@brief reserve capacity for @a len elements in array @a arr
+
+Reserving upfront avoids repeated reallocations while the elements are added,
+but the reservation is capped so a bogus/hostile length (which is not bounded
+by max_size(), unlike e.g. std::vector) cannot trigger an oversized allocation
+for a small or truncated input.
+
+The overload below is selected for array types without reserve() (e.g.,
+std::deque), which are then left untouched.
+*/
+template<typename ArrayType>
+auto reserve_array(ArrayType& arr, std::size_t len, priority_tag<1> /*unused*/)
+-> decltype(arr.reserve(len), void())
+{
+    constexpr std::size_t reserve_cap = 16384;
+    arr.reserve((std::min)(len, reserve_cap));
+}
+
+template<typename ArrayType>
+inline void reserve_array(ArrayType& /*arr*/, std::size_t /*len*/, priority_tag<0> /*unused*/)
+{}
+
 /*!
 @brief SAX implementation to create a JSON value from SAX events
 
@@ -11027,6 +11044,11 @@ class json_sax_dom_parser
             JSON_THROW(out_of_range::create(408, concat("excessive array size: ", std::to_string(len)), ref_stack.back()));
         }
 
+        if (len != detail::unknown_size())
+        {
+            reserve_array(*ref_stack.back()->m_data.m_value.array, len, priority_tag<1> {});
+        }
+
         return true;
     }
 
@@ -11405,6 +11427,11 @@ class json_sax_dom_callback_parser
             {
                 JSON_THROW(out_of_range::create(408, concat("excessive array size: ", std::to_string(len)), ref_stack.back()));
             }
+
+            if (len != detail::unknown_size())
+            {
+                reserve_array(*ref_stack.back()->m_data.m_value.array, len, priority_tag<1> {});
+            }
         }
 
         return true;
@@ -13392,7 +13419,7 @@ class binary_reader
                 // a copy, not a reference: it must stay valid across the
                 // pop_back() below, which destroys the container_stack element
                 // it would otherwise alias
-                container_frame top = container_stack.back();
+                const container_frame top = container_stack.back();
                 bool at_end = false;
 
                 if (top.remaining != npos)
@@ -14169,7 +14196,7 @@ class binary_reader
             // would otherwise alias.
             for (;;)
             {
-                container_frame top = container_stack.back();
+                const container_frame top = container_stack.back();
 
                 if (top.remaining != npos)
                 {
@@ -24564,351 +24591,6 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         return j;
     }
 
-#ifndef JSON_NO_THREAD_LOCAL
-    /// the number of levels an operation descends into before it finishes the
-    /// value below it without the call stack
-    static constexpr std::uint8_t nesting_depth_limit()
-    {
-        return 128;
-    }
-
-    /*!
-    @brief how many levels the operation going on in this thread has descended into
-
-    Copying a value and comparing two values share this count. The library never
-    nests one inside the other - copying a value does not compare one, and
-    comparing two values does not copy them - and where user code nests them
-    anyway, sharing the count only ends a descent sooner than it had to, which
-    costs a little speed and is never wrong.
-
-    A byte is enough: the count never exceeds the limit by more than the single
-    level that notices the limit has been reached.
-    */
-    static std::uint8_t& nesting_depth() noexcept
-    {
-        static thread_local std::uint8_t depth = 0; // NOLINT(misc-use-internal-linkage)
-        return depth;
-    }
-#endif
-
-    /*!
-    @brief counts one level of a bounded descent for as long as it runs, and
-           reports whether the descent was still within the limit when it began
-
-    Looks the count up and tests it against the limit itself, rather than
-    leaving that to the caller: either way it is reached exactly once, so
-    there is nothing to be gained by making the caller do it.
-
-    Does nothing and is never @ref okay without thread-local storage, where no
-    descent can be bounded at all: a caller that only descends while this says
-    it may always ends up finishing without the call stack, exactly as if every
-    value were nested past the limit.
-    */
-    class nesting_depth_guard
-    {
-      public:
-        nesting_depth_guard() noexcept
-#ifdef JSON_NO_THREAD_LOCAL
-            : m_okay(false)
-#else
-            : m_okay(nesting_depth() < nesting_depth_limit())
-#endif
-        {
-#ifndef JSON_NO_THREAD_LOCAL
-            ++nesting_depth();
-#endif
-        }
-
-        ~nesting_depth_guard()
-        {
-#ifndef JSON_NO_THREAD_LOCAL
-            --nesting_depth();
-#endif
-        }
-
-        nesting_depth_guard(const nesting_depth_guard&) = delete;
-        nesting_depth_guard& operator=(const nesting_depth_guard&) = delete;
-        nesting_depth_guard(nesting_depth_guard&&) = delete;
-        nesting_depth_guard& operator=(nesting_depth_guard&&) = delete;
-
-        bool okay() const noexcept
-        {
-            return m_okay;
-        }
-
-      private:
-        bool m_okay;
-    };
-
-    /// an entry of the iterative deep copy's worklist: a structured value and
-    /// the value that is to become its copy
-    using copy_worklist_t = std::vector<std::pair<const basic_json*, basic_json*>>;
-
-    /// scratch space to build the key skeleton of an object copy in one go
-    using copy_scratch_t = std::vector<std::pair<typename object_t::key_type, basic_json>>;
-
-    /// @brief copy everything of @a src into @a dst but its type and value
-    static void copy_metadata(const basic_json& src, basic_json& dst)
-    {
-        // a custom base class is only required to be copy-constructible and
-        // move-assignable, so the copy has to go through a temporary
-        static_cast<json_base_class_t&>(dst) = json_base_class_t(static_cast<const json_base_class_t&>(src));
-
-#if JSON_DIAGNOSTIC_POSITIONS
-        dst.start_position = src.start_position;
-        dst.end_position = src.end_position;
-#else
-        static_cast<void>(src);
-        static_cast<void>(dst);
-#endif
-    }
-
-    /*!
-    @brief copy the value of @a src into @a dst, which must not be structured
-
-    Objects and arrays are left alone: creating those is the one thing the copy
-    constructor and @ref copy_shallow do differently from one another, and it is
-    the reason copying a value can descend at all.
-    */
-    /// @note inlined on purpose: both callers have already told an object or an
-    ///       array apart from the rest, and letting the compiler fold that test
-    ///       into this switch is worth a few percent when copying a value made
-    ///       mostly of numbers
-    JSON_HEDLEY_ALWAYS_INLINE
-    static void copy_leaf_value(const basic_json& src, basic_json& dst)
-    {
-        switch (src.m_data.m_type)
-        {
-            case value_t::string:
-            {
-                dst.m_data.m_value = *src.m_data.m_value.string;
-                break;
-            }
-
-            case value_t::binary:
-            {
-                dst.m_data.m_value = *src.m_data.m_value.binary;
-                break;
-            }
-
-            case value_t::boolean:
-            {
-                dst.m_data.m_value = src.m_data.m_value.boolean;
-                break;
-            }
-
-            case value_t::number_integer:
-            {
-                dst.m_data.m_value = src.m_data.m_value.number_integer;
-                break;
-            }
-
-            case value_t::number_unsigned:
-            {
-                dst.m_data.m_value = src.m_data.m_value.number_unsigned;
-                break;
-            }
-
-            case value_t::number_float:
-            {
-                dst.m_data.m_value = src.m_data.m_value.number_float;
-                break;
-            }
-
-            case value_t::object:
-            case value_t::array:
-            case value_t::null:
-            case value_t::discarded:
-            default:
-                break;
-        }
-    }
-
-    /*!
-    @brief copy everything of @a src into the null value @a dst but the children
-
-    Objects and arrays are not copied here; they are appended to @a worklist to
-    be created later by @ref copy_iteratively. Until that happens, @a dst remains
-    a null value, so that a partially built copy can be destroyed at any point
-    without ever violating the class invariants.
-    */
-    static void copy_shallow(const basic_json& src, basic_json& dst, copy_worklist_t& worklist)
-    {
-        copy_metadata(src, dst);
-
-        if (src.m_data.m_type == value_t::object || src.m_data.m_type == value_t::array)
-        {
-            // defer: dst stays a null value until its container exists
-            worklist.emplace_back(&src, &dst);
-            return;
-        }
-
-        copy_leaf_value(src, dst);
-
-        // only now that the value exists may the type be set: had the creation
-        // of the value thrown, dst would have been left as a valid null value
-        dst.m_data.m_type = src.m_data.m_type;
-    }
-
-    /// @brief create the copy of the array @a src in @a dst
-    /// @note structured elements are appended to @a worklist instead
-    static void copy_array_level(const basic_json& src, basic_json& dst, copy_worklist_t& worklist)
-    {
-        const array_t& src_array = *src.m_data.m_value.array;
-
-        // create all elements up front: growing the array afterwards could
-        // invalidate the pointers that are handed to the worklist
-        dst.m_data.m_value.array = create<array_t>(src_array.size(), basic_json());
-
-        auto dst_it = dst.m_data.m_value.array->begin();
-        for (auto src_it = src_array.cbegin(); src_it != src_array.cend(); ++src_it, ++dst_it)
-        {
-            copy_shallow(*src_it, *dst_it, worklist);
-        }
-    }
-
-    /// @brief create the copy of the object @a src in @a dst
-    /// @note structured values are appended to @a worklist instead
-    static void copy_object_level(const basic_json& src, basic_json& dst,
-                                  copy_worklist_t& worklist, copy_scratch_t& scratch)
-    {
-        const object_t& src_object = *src.m_data.m_value.object;
-
-        // build the complete key skeleton and hand it to the object's range
-        // constructor: adding the keys one by one would be quadratic for object
-        // types that are backed by a vector, such as nlohmann::ordered_map
-        scratch.clear();
-        scratch.reserve(src_object.size());
-        for (const auto& element : src_object)
-        {
-            scratch.emplace_back(element.first, basic_json());
-        }
-
-        dst.m_data.m_value.object = create<object_t>(std::make_move_iterator(scratch.begin()),
-                                    std::make_move_iterator(scratch.end()));
-        scratch.clear();
-
-        // pair every value of the copy with its counterpart in the original;
-        // both are enumerated in the same order for every object type with a
-        // deterministic order, so the lookup is only needed for exotic ones
-        auto src_it = src_object.cbegin();
-        for (auto& element : *dst.m_data.m_value.object)
-        {
-            if (JSON_HEDLEY_LIKELY(src_it != src_object.cend() && src_it->first == element.first))
-            {
-                copy_shallow(src_it->second, element.second, worklist);
-                ++src_it;
-            }
-            else
-            {
-                const auto found = src_object.find(element.first);
-                JSON_ASSERT(found != src_object.cend());
-                copy_shallow(found->second, element.second, worklist);
-            }
-        }
-    }
-
-    /*!
-    @brief deep-copy the object or array @a src into this value without recursing
-
-    The values whose copy has not been created yet are kept on an explicit
-    worklist rather than on the call stack. This is only reached for values
-    nested deeper than @ref nesting_depth_limit levels, which is why it copies
-    every container by hand instead of letting the container do it: the fast
-    ways of doing so would descend into the elements and defeat the purpose.
-    */
-    void copy_iteratively(const basic_json& src)
-    {
-        copy_worklist_t worklist;
-        copy_scratch_t scratch;
-
-        const basic_json* src_value = &src;
-        basic_json* dst_value = this;
-
-        for (;;)
-        {
-            if (src_value->m_data.m_type == value_t::array)
-            {
-                copy_array_level(*src_value, *dst_value, worklist);
-            }
-            else
-            {
-                copy_object_level(*src_value, *dst_value, worklist, scratch);
-            }
-
-            // the container is complete and will not be modified again
-            dst_value->set_parents();
-
-            if (worklist.empty())
-            {
-                break;
-            }
-
-            const auto& next = worklist.back();
-            src_value = next.first;
-            dst_value = next.second;
-            worklist.pop_back();
-
-            // the value stops being a null value exactly here
-            dst_value->m_data.m_type = src_value->m_data.m_type;
-        }
-    }
-
-    /*!
-    @brief copy one level of the object or array @a src into this value
-
-    The container copies its own elements, which is the fastest way to fill it.
-    Every element that is structured itself comes back to @ref copy_structured.
-    */
-    void copy_level(const basic_json& src)
-    {
-        if (m_data.m_type == value_t::object)
-        {
-            m_data.m_value = *src.m_data.m_value.object;
-        }
-        else
-        {
-            m_data.m_value = *src.m_data.m_value.array;
-        }
-
-        set_parents();
-    }
-
-    /*!
-    @brief deep-copy the object or array @a src into this value
-
-    Copying a container copies its elements, so a value nested deeply enough
-    used to exhaust the call stack. The descent is bounded here: the first
-    @ref nesting_depth_limit levels are copied by the containers themselves, just
-    as they always were, and anything below that is copied without the call
-    stack by @ref copy_iteratively. Copying a value can therefore no longer
-    exhaust the stack, however deeply it is nested, just like destroying one
-    cannot since #1436.
-
-    Nothing has to be scanned or built by hand to reach that: a value that is
-    not nested deeper than the limit - all but a vanishing minority - is copied
-    exactly as it was before, and this whole detour costs it one counter.
-
-    @sa https://github.com/nlohmann/json/issues/5387
-    */
-    void copy_structured(const basic_json& src)
-    {
-        const nesting_depth_guard guard;
-
-        if (JSON_HEDLEY_LIKELY(guard.okay()))
-        {
-            copy_level(src);
-            return;
-        }
-
-        // Finish this value without descending any further. It is completed
-        // before this returns, so a copy made by a custom base class - or by
-        // anything else that runs while a copy is going on - is unaffected by
-        // the copy it is nested in.
-        copy_iteratively(src);
-    }
-
-
   public:
     //////////////////////////
     // JSON parser callback //
@@ -25288,15 +24970,60 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         // check of passed value is valid
         other.assert_invariant();
 
-        if (m_data.m_type == value_t::object || m_data.m_type == value_t::array)
+        switch (m_data.m_type)
         {
-            // copying the container directly would call this constructor again
-            // for every element, once per nesting level
-            copy_structured(other);
-        }
-        else
-        {
-            copy_leaf_value(other, *this);
+            case value_t::object:
+            {
+                m_data.m_value = *other.m_data.m_value.object;
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_data.m_value = *other.m_data.m_value.array;
+                break;
+            }
+
+            case value_t::string:
+            {
+                m_data.m_value = *other.m_data.m_value.string;
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                m_data.m_value = other.m_data.m_value.boolean;
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                m_data.m_value = other.m_data.m_value.number_integer;
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                m_data.m_value = other.m_data.m_value.number_unsigned;
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                m_data.m_value = other.m_data.m_value.number_float;
+                break;
+            }
+
+            case value_t::binary:
+            {
+                m_data.m_value = *other.m_data.m_value.binary;
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                break;
         }
 
         set_parents();
@@ -27618,6 +27345,11 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         std::swap(m_data.m_type, other.m_data.m_type);
         std::swap(m_data.m_value, other.m_data.m_value);
 
+#if JSON_DIAGNOSTIC_POSITIONS
+        std::swap(start_position, other.start_position);
+        std::swap(end_position, other.end_position);
+#endif
+
         set_parents();
         other.set_parents();
         assert_invariant();

tests/CMakeLists.txt

@@ -75,12 +75,7 @@ target_compile_options(test_main PUBLIC
     #                              is annotated JSON_HEDLEY_NO_RETURN (it always throws), which
     #                              makes MSVC flag the code following its call in binary_reader.hpp
     #                              as unreachable for that instantiation, in both Debug and Release
-    #       Disable warning C4503: decorated name length exceeded, name was truncated; the deep
-    #                              copy support added for #5387 pushes the mangled name of
-    #                              std::allocator_traits<...>::construct for the custom-base-class
-    #                              test's map type past VS2015's limit. The name is only used for
-    #                              debug info, so truncation does not affect the build.
-    $<$<CXX_COMPILER_ID:MSVC>:/W4;/wd4566;/wd4996;/wd4702;/wd4503>
+    $<$<CXX_COMPILER_ID:MSVC>:/W4;/wd4566;/wd4996;/wd4702>
     # https://github.com/nlohmann/json/issues/1114
     $<$<CXX_COMPILER_ID:MSVC>:/bigobj> $<$<BOOL:${MINGW}>:-Wa,-mbig-obj>
 

tests/benchmarks/src/benchmarks.cpp

@@ -252,4 +252,323 @@ static void BinaryToCbor(benchmark::State& state)
 }
 BENCHMARK(BinaryToCbor)->RangeMultiplier(2)->Range(8, 8 << 12);
 
+//////////////////////////////////////////////////////////////////////////////
+// parse binary formats
+//////////////////////////////////////////////////////////////////////////////
+
+// Only MessagePack had a read benchmark (FromMsgpack above, left untouched so
+// its numbers stay comparable across releases). The benchmarks below cover the
+// other formats, and read from a contiguous buffer as well as from a FILE*:
+// most callers pass a container, and the two adapters compile to different
+// code. The test data repository ships JSON only, so the input for each is
+// derived at setup time by serializing a parsed test file.
+
+/// binary format to benchmark; the _optimized variants add UBJSON/BJData size
+/// and type annotations, which the readers handle in a separate code path
+enum class binary_format
+{
+    cbor,
+    msgpack,
+    ubjson,
+    ubjson_optimized,
+    bjdata,
+    bjdata_optimized,
+    bson
+};
+
+static std::vector<std::uint8_t> to_binary(const json& j, const binary_format format)
+{
+    switch (format)
+    {
+        case binary_format::cbor:
+            return json::to_cbor(j);
+        case binary_format::msgpack:
+            return json::to_msgpack(j);
+        case binary_format::ubjson:
+            return json::to_ubjson(j);
+        case binary_format::ubjson_optimized:
+            return json::to_ubjson(j, true, true);
+        case binary_format::bjdata:
+            return json::to_bjdata(j);
+        case binary_format::bjdata_optimized:
+            return json::to_bjdata(j, true, true);
+        case binary_format::bson:
+        default:
+            return json::to_bson(j);
+    }
+}
+
+static json from_binary(const std::vector<std::uint8_t>& bytes, const binary_format format)
+{
+    switch (format)
+    {
+        case binary_format::cbor:
+            return json::from_cbor(bytes);
+        case binary_format::msgpack:
+            return json::from_msgpack(bytes);
+        case binary_format::ubjson:
+        case binary_format::ubjson_optimized:
+            return json::from_ubjson(bytes);
+        case binary_format::bjdata:
+        case binary_format::bjdata_optimized:
+            return json::from_bjdata(bytes);
+        case binary_format::bson:
+        default:
+            return json::from_bson(bytes);
+    }
+}
+
+static json from_binary(std::FILE* file, const binary_format format)
+{
+    switch (format)
+    {
+        case binary_format::cbor:
+            return json::from_cbor(file);
+        case binary_format::msgpack:
+            return json::from_msgpack(file);
+        case binary_format::ubjson:
+        case binary_format::ubjson_optimized:
+            return json::from_ubjson(file);
+        case binary_format::bjdata:
+        case binary_format::bjdata_optimized:
+            return json::from_bjdata(file);
+        case binary_format::bson:
+        default:
+            return json::from_bson(file);
+    }
+}
+
+/*!
+@brief serialize a parsed test file to @a format
+
+Returns an empty vector and marks the benchmark as skipped if the file cannot
+be represented in the format, rather than letting the exception escape: BSON
+requires an object at the top level, and several test files are arrays.
+*/
+static std::vector<std::uint8_t> binary_input(benchmark::State& state, const char* filename, const binary_format format)
+{
+    std::ifstream f(filename);
+    std::string const str((std::istreambuf_iterator<char>(f)), std::istreambuf_iterator<char>());
+    const json j = json::parse(str);
+
+    if (format == binary_format::bson && !j.is_object())
+    {
+        state.SkipWithError("BSON requires an object at the top level");
+        return {};
+    }
+
+    return to_binary(j, format);
+}
+
+static void FromBinaryBuffer(benchmark::State& state, const char* filename, const binary_format format)
+{
+    const std::vector<std::uint8_t> bytes = binary_input(state, filename, format);
+    if (bytes.empty())
+    {
+        return;
+    }
+
+    for (auto _ : state)
+    {
+        // the value is destroyed outside the timed section, because destroying
+        // a large DOM is not what this benchmark measures
+        state.PauseTiming();
+        auto* j = new json();
+        state.ResumeTiming();
+
+        *j = from_binary(bytes, format);
+
+        state.PauseTiming();
+        delete j;
+        state.ResumeTiming();
+    }
+
+    state.SetBytesProcessed(state.iterations() * bytes.size());
+}
+
+BENCHMARK_CAPTURE(FromBinaryBuffer, cbor / jeopardy, TEST_DATA_DIRECTORY "/jeopardy/jeopardy.json", binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryBuffer, cbor / canada, TEST_DATA_DIRECTORY "/nativejson-benchmark/canada.json", binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryBuffer, cbor / citm_catalog, TEST_DATA_DIRECTORY "/nativejson-benchmark/citm_catalog.json", binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryBuffer, cbor / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryBuffer, cbor / floats, TEST_DATA_DIRECTORY "/regression/floats.json", binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryBuffer, cbor / signed_ints, TEST_DATA_DIRECTORY "/regression/signed_ints.json", binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryBuffer, msgpack / jeopardy, TEST_DATA_DIRECTORY "/jeopardy/jeopardy.json", binary_format::msgpack);
+BENCHMARK_CAPTURE(FromBinaryBuffer, msgpack / canada, TEST_DATA_DIRECTORY "/nativejson-benchmark/canada.json", binary_format::msgpack);
+BENCHMARK_CAPTURE(FromBinaryBuffer, msgpack / citm_catalog, TEST_DATA_DIRECTORY "/nativejson-benchmark/citm_catalog.json", binary_format::msgpack);
+BENCHMARK_CAPTURE(FromBinaryBuffer, msgpack / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::msgpack);
+BENCHMARK_CAPTURE(FromBinaryBuffer, ubjson / jeopardy, TEST_DATA_DIRECTORY "/jeopardy/jeopardy.json", binary_format::ubjson);
+BENCHMARK_CAPTURE(FromBinaryBuffer, ubjson / canada, TEST_DATA_DIRECTORY "/nativejson-benchmark/canada.json", binary_format::ubjson);
+BENCHMARK_CAPTURE(FromBinaryBuffer, ubjson / citm_catalog, TEST_DATA_DIRECTORY "/nativejson-benchmark/citm_catalog.json", binary_format::ubjson);
+BENCHMARK_CAPTURE(FromBinaryBuffer, ubjson / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::ubjson);
+BENCHMARK_CAPTURE(FromBinaryBuffer, ubjson_optimized / canada, TEST_DATA_DIRECTORY "/nativejson-benchmark/canada.json", binary_format::ubjson_optimized);
+BENCHMARK_CAPTURE(FromBinaryBuffer, ubjson_optimized / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::ubjson_optimized);
+BENCHMARK_CAPTURE(FromBinaryBuffer, bjdata / canada, TEST_DATA_DIRECTORY "/nativejson-benchmark/canada.json", binary_format::bjdata);
+BENCHMARK_CAPTURE(FromBinaryBuffer, bjdata / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::bjdata);
+BENCHMARK_CAPTURE(FromBinaryBuffer, bjdata_optimized / canada, TEST_DATA_DIRECTORY "/nativejson-benchmark/canada.json", binary_format::bjdata_optimized);
+BENCHMARK_CAPTURE(FromBinaryBuffer, bjdata_optimized / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::bjdata_optimized);
+// BSON requires an object at the top level, so the array-rooted test files
+// (jeopardy and the regression files) cannot be captured here
+BENCHMARK_CAPTURE(FromBinaryBuffer, bson / canada, TEST_DATA_DIRECTORY "/nativejson-benchmark/canada.json", binary_format::bson);
+BENCHMARK_CAPTURE(FromBinaryBuffer, bson / citm_catalog, TEST_DATA_DIRECTORY "/nativejson-benchmark/citm_catalog.json", binary_format::bson);
+BENCHMARK_CAPTURE(FromBinaryBuffer, bson / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::bson);
+
+static void FromBinaryFile(benchmark::State& state, const char* filename, const binary_format format)
+{
+    const std::vector<std::uint8_t> bytes = binary_input(state, filename, format);
+    if (bytes.empty())
+    {
+        return;
+    }
+
+    const char* tmp = "benchmark_input.bin";
+    std::ofstream o(tmp, std::ios::binary);
+    o.write(reinterpret_cast<const char*>(bytes.data()), static_cast<std::streamsize>(bytes.size()));
+    o.flush();
+    o.close();
+
+    for (auto _ : state)
+    {
+        state.PauseTiming();
+        auto* j = new json();
+        auto* file = std::fopen(tmp, "rb");
+        state.ResumeTiming();
+
+        *j = from_binary(file, format);
+
+        state.PauseTiming();
+        std::fclose(file);
+        delete j;
+        state.ResumeTiming();
+    }
+
+    state.SetBytesProcessed(state.iterations() * bytes.size());
+}
+
+BENCHMARK_CAPTURE(FromBinaryFile, cbor / canada, TEST_DATA_DIRECTORY "/nativejson-benchmark/canada.json", binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryFile, cbor / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryFile, ubjson / canada, TEST_DATA_DIRECTORY "/nativejson-benchmark/canada.json", binary_format::ubjson);
+BENCHMARK_CAPTURE(FromBinaryFile, ubjson / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::ubjson);
+BENCHMARK_CAPTURE(FromBinaryFile, bjdata / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::bjdata);
+BENCHMARK_CAPTURE(FromBinaryFile, bson / twitter, TEST_DATA_DIRECTORY "/nativejson-benchmark/twitter.json", binary_format::bson);
+
+//////////////////////////////////////////////////////////////////////////////
+// parse binary formats: value shapes
+//////////////////////////////////////////////////////////////////////////////
+
+// The test files above are wide and shallow, but the readers' cost is per
+// container, so these cover the shapes that stress the container handling
+// itself. Every shape is wrapped in an object so that BSON, which requires an
+// object at the top level, measures the same value as the other formats.
+
+/// deeply nested arrays: one container per level, no other work
+static json make_nested()
+{
+    json nested = json::array();
+    json* p = &nested;
+    for (std::size_t i = 1; i < 1000; ++i)
+    {
+        p->push_back(json::array());
+        p = &p->operator[](0);
+    }
+
+    json j = json::object();
+    j["data"] = std::move(nested);
+    return j;
+}
+
+/// many sibling containers: maximum container churn, minimum nesting
+static json make_containers()
+{
+    json data = json::array();
+    for (std::size_t i = 0; i < 100000; ++i)
+    {
+        data.push_back(json::array({1, 2}));
+    }
+
+    json j = json::object();
+    j["data"] = std::move(data);
+    return j;
+}
+
+/// one flat array of numbers: the scalar decoding path, which must not move
+static json make_scalars()
+{
+    json data = json::array();
+    for (std::size_t i = 0; i < 1000000; ++i)
+    {
+        data.push_back(i);
+    }
+
+    json j = json::object();
+    j["data"] = std::move(data);
+    return j;
+}
+
+static void FromBinaryShape(benchmark::State& state, json (*build)(), const binary_format format)
+{
+    const std::vector<std::uint8_t> bytes = to_binary(build(), format);
+
+    for (auto _ : state)
+    {
+        state.PauseTiming();
+        auto* j = new json();
+        state.ResumeTiming();
+
+        *j = from_binary(bytes, format);
+
+        state.PauseTiming();
+        delete j;
+        state.ResumeTiming();
+    }
+
+    state.SetBytesProcessed(state.iterations() * bytes.size());
+}
+
+BENCHMARK_CAPTURE(FromBinaryShape, nested / cbor, make_nested, binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryShape, nested / msgpack, make_nested, binary_format::msgpack);
+BENCHMARK_CAPTURE(FromBinaryShape, nested / ubjson, make_nested, binary_format::ubjson);
+BENCHMARK_CAPTURE(FromBinaryShape, nested / bjdata, make_nested, binary_format::bjdata);
+BENCHMARK_CAPTURE(FromBinaryShape, nested / bson, make_nested, binary_format::bson);
+BENCHMARK_CAPTURE(FromBinaryShape, containers / cbor, make_containers, binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryShape, containers / msgpack, make_containers, binary_format::msgpack);
+BENCHMARK_CAPTURE(FromBinaryShape, containers / ubjson, make_containers, binary_format::ubjson);
+BENCHMARK_CAPTURE(FromBinaryShape, containers / ubjson_optimized, make_containers, binary_format::ubjson_optimized);
+BENCHMARK_CAPTURE(FromBinaryShape, containers / bjdata, make_containers, binary_format::bjdata);
+BENCHMARK_CAPTURE(FromBinaryShape, containers / bson, make_containers, binary_format::bson);
+// BSON names every array element, so a large array measures key generation
+// rather than scalar decoding and is left out here
+BENCHMARK_CAPTURE(FromBinaryShape, scalars / cbor, make_scalars, binary_format::cbor);
+BENCHMARK_CAPTURE(FromBinaryShape, scalars / msgpack, make_scalars, binary_format::msgpack);
+BENCHMARK_CAPTURE(FromBinaryShape, scalars / ubjson, make_scalars, binary_format::ubjson);
+BENCHMARK_CAPTURE(FromBinaryShape, scalars / bjdata, make_scalars, binary_format::bjdata);
+
+/*!
+@brief parse an indefinite-length CBOR string
+
+The writer never emits this form, so the input is assembled by hand: 0x7F
+opens the string, each chunk is a one-character string, and 0xFF closes it.
+*/
+static void FromCborChunkedString(benchmark::State& state, const std::size_t chunks)
+{
+    std::vector<std::uint8_t> bytes;
+    bytes.reserve(2 * chunks + 2);
+    bytes.push_back(0x7F);
+    for (std::size_t i = 0; i < chunks; ++i)
+    {
+        bytes.push_back(0x61); // string of length 1
+        bytes.push_back(0x61); // 'a'
+    }
+    bytes.push_back(0xFF);
+
+    for (auto _ : state)
+    {
+        json j = json::from_cbor(bytes);
+        benchmark::DoNotOptimize(j);
+    }
+
+    state.SetBytesProcessed(state.iterations() * bytes.size());
+}
+
+BENCHMARK_CAPTURE(FromCborChunkedString, 10000 chunks, 10000);
+
 BENCHMARK_MAIN();

tests/src/unit-allocator.cpp

@@ -216,57 +216,6 @@ TEST_CASE("controlled bad_alloc")
             CHECK_THROWS_AS(my_json(s), std::bad_alloc&);
             next_construct_fails = false;
         }
-
-        SECTION("basic_json(const basic_json&) of a deeply nested value (#5387)")
-        {
-            // Copying a value nested deeper than the descent bound builds the
-            // copy from the top down: every value whose own copy has not been
-            // made yet stays a null value until it is. Failing an allocation
-            // part-way through is what proves such a half-built copy can still
-            // be destroyed.
-            //
-            // Which path the failure lands in depends on the build: the first
-            // allocation of a copy belongs to the outermost level, so here it
-            // is the descending one. Built with JSON_NO_THREAD_LOCAL - as the
-            // ci_test_no_thread_local target builds the whole suite - no
-            // descent is made at all and the very same failure lands in the
-            // iterative path instead, part-way through its worklist.
-            const auto check_deep_copy = [](bool objects)
-            {
-                CAPTURE(objects);
-
-                next_construct_fails = false;
-
-                // deeper than the 128 levels the copy constructor descends into
-                const std::size_t depth = 300;
-
-                my_json j = 1;
-                for (std::size_t i = 0; i < depth; ++i)
-                {
-                    if (objects)
-                    {
-                        my_json wrapper = my_json::object();
-                        wrapper["a"] = std::move(j);
-                        j = std::move(wrapper);
-                    }
-                    else
-                    {
-                        j = my_json::array({std::move(j)});
-                    }
-                }
-
-                // NOLINTNEXTLINE(performance-unnecessary-copy-initialization): the copy is what is tested
-                CHECK_NOTHROW(my_json(j));
-
-                next_construct_fails = true;
-                // NOLINTNEXTLINE(performance-unnecessary-copy-initialization): the copy is what is tested
-                CHECK_THROWS_AS(my_json(j), std::bad_alloc&);
-                next_construct_fails = false;
-            };
-
-            check_deep_copy(false);
-            check_deep_copy(true);
-        }
     }
 }
 

tests/src/unit-bjdata.cpp

@@ -3551,6 +3551,111 @@ TEST_CASE("BJData")
     }
 }
 
+TEST_CASE("issue #5405 - array reserve for definite-length BJData arrays")
+{
+#if !defined(JSON_NOEXCEPTION)
+    // this SECTION relies on catching a thrown exception to distinguish
+    // which of two acceptable, bounded rejections a hostile header took;
+    // under JSON_NOEXCEPTION, JSON_THROW never produces a catchable C++
+    // exception (it aborts instead), so this cannot be tested that way here
+    SECTION("a huge claimed length with no element data must not over-allocate")
+    {
+        // optimized form [$type#count: type 'i' (int8), count as a four-byte
+        // little-endian 'l' (int32) of 0x7FFFFFFF (2147483647), but no
+        // element data at all. max_size() for a std::vector is far larger
+        // than this count, so it does not reject the header outright; the
+        // (capped) reservation must not attempt to allocate space for
+        // billions of elements before the missing data is detected.
+        json _;
+        const std::vector<uint8_t> input = {'[', '$', 'i', '#', 'l', 0xFF, 0xFF, 0xFF, 0x7F};
+        // On a platform where std::vector<json>::max_size() is smaller than
+        // the claimed count (e.g. 32-bit, where max_size() is bounded by a
+        // 32-bit SIZE_MAX divided by sizeof(json)), the SAX consumer's own
+        // check rejects the header outright (out_of_range.408, with the
+        // claimed count in the message) instead of accepting it and only
+        // finding it short of data once the (capped) reservation looks for
+        // element bytes that were never provided (parse_error.110). Either
+        // is an acceptable, bounded rejection of the hostile header -- the
+        // property under test is that no path attempts to allocate space
+        // for billions of elements.
+        bool threw = false;
+        try
+        {
+            _ = json::from_bjdata(input);
+        }
+        catch (const json::parse_error& e)
+        {
+            threw = true;
+            CHECK(e.id == 110);
+            CHECK(std::string(e.what()) == "[json.exception.parse_error.110] parse error at byte 10: syntax error while parsing BJData number: unexpected end of input");
+        }
+        catch (const json::out_of_range& e)
+        {
+            threw = true;
+            CHECK(e.id == 408);
+            CHECK(std::string(e.what()).find("excessive array size") != std::string::npos);
+        }
+        CHECK(threw);
+
+        // json_sax_dom_parser::start_array()'s max_size() check (unlike the
+        // scanner's own parse_error path) throws unconditionally via
+        // JSON_THROW rather than going through sax->parse_error(), so it is
+        // not gated by allow_exceptions=false on a platform where this
+        // header hits that check (e.g. 32-bit, see above) -- allow either
+        // a discarded result or the same out_of_range it throws with
+        // exceptions enabled.
+        try
+        {
+            CHECK(json::from_bjdata(input, true, false).is_discarded());
+        }
+        catch (const json::out_of_range& e)
+        {
+            CHECK(e.id == 408);
+        }
+    }
+#endif
+
+    SECTION("arrays of various sizes decode to the same value as before the reserve optimization")
+    {
+        for (const auto size :
+                {
+                    std::size_t{0}, std::size_t{1}, std::size_t{5}, // small
+                    std::size_t{16384},                             // exactly at the reserve cap
+                    std::size_t{20000}                              // above the reserve cap
+                })
+        {
+            CAPTURE(size)
+            json j = json::array();
+            for (std::size_t i = 0; i < size; ++i)
+            {
+                j.push_back(static_cast<int>(i % 1000));
+            }
+
+            // exercise both the plain and the optimized [$type#count encoding
+            const auto packed_plain = json::to_bjdata(j);
+            CHECK(json::from_bjdata(packed_plain) == j);
+
+            const auto packed_optimized = json::to_bjdata(j, true, true);
+            CHECK(json::from_bjdata(packed_optimized) == j);
+        }
+    }
+
+    SECTION("a user-defined SAX consumer is unaffected by the internal DOM reserve optimization")
+    {
+        // the reserve() call is local to json_sax_dom_parser / json_sax_dom_callback_parser;
+        // a custom SAX consumer that does not touch a DOM array sees identical events
+        json j = json::array();
+        for (int i = 0; i < 100; ++i)
+        {
+            j.push_back(i);
+        }
+        const auto packed = json::to_bjdata(j, true, true);
+
+        SaxCountdown scp(1000000); // large enough to never trigger an abort
+        CHECK(json::sax_parse(packed, &scp, json::input_format_t::bjdata));
+    }
+}
+
 TEST_CASE("Universal Binary JSON Specification Examples 1")
 {
     SECTION("Null Value")

tests/src/unit-cbor.cpp

@@ -2174,6 +2174,92 @@ TEST_CASE("CBOR indefinite-length strings do not recurse per chunk")
     }
 }
 
+TEST_CASE("issue #5405 - array reserve for definite-length CBOR arrays")
+{
+#if !defined(JSON_NOEXCEPTION)
+    // this SECTION relies on catching a thrown exception to distinguish
+    // which of two acceptable, bounded rejections a hostile header took;
+    // under JSON_NOEXCEPTION, JSON_THROW never produces a catchable C++
+    // exception (it aborts instead), so this cannot be tested that way here
+    SECTION("a huge claimed length with no element data must not over-allocate")
+    {
+        // 0x9A: array with a four-byte length; claims 0xFFFFFFFF (4294967295)
+        // elements but provides none. max_size() for a std::vector is far
+        // larger than this count, so it does not reject the header outright;
+        // the (capped) reservation must not attempt to allocate space for
+        // billions of elements before the missing data is detected.
+        json _;
+        const std::vector<uint8_t> input = {0x9A, 0xFF, 0xFF, 0xFF, 0xFF};
+        // On a platform where std::size_t is narrower than 64 bits (e.g.
+        // 32-bit), the claimed count 0xFFFFFFFF coincides with that
+        // platform's detail::unknown_size() sentinel (SIZE_MAX), so the
+        // format-level size check rejects it outright (out_of_range.408,
+        // "excessive ... size") before the SAX consumer's own max_size()
+        // check would even run; on a 64-bit platform it passes both of
+        // those checks and is only found short of data once the (capped)
+        // reservation looks for element bytes that were never provided
+        // (parse_error.110). Either is an acceptable, bounded rejection of
+        // the hostile header -- the property under test is that no path
+        // attempts to allocate space for billions of elements.
+        bool threw = false;
+        try
+        {
+            _ = json::from_cbor(input);
+        }
+        catch (const json::parse_error& e)
+        {
+            threw = true;
+            CHECK(e.id == 110);
+            CHECK(std::string(e.what()) == "[json.exception.parse_error.110] parse error at byte 6: syntax error while parsing CBOR value: unexpected end of input");
+        }
+        catch (const json::out_of_range& e)
+        {
+            threw = true;
+            CHECK(e.id == 408);
+            CHECK(std::string(e.what()).find("excessive") != std::string::npos);
+        }
+        CHECK(threw);
+        CHECK(json::from_cbor(input, true, false).is_discarded());
+    }
+#endif
+
+    SECTION("arrays of various sizes decode to the same value as before the reserve optimization")
+    {
+        for (const auto size :
+                {
+                    std::size_t{0}, std::size_t{1}, std::size_t{5}, // small
+                    std::size_t{16384},                             // exactly at the reserve cap
+                    std::size_t{20000}                              // above the reserve cap
+                })
+        {
+            CAPTURE(size)
+            json j = json::array();
+            for (std::size_t i = 0; i < size; ++i)
+            {
+                j.push_back(static_cast<int>(i % 1000));
+            }
+
+            const auto packed = json::to_cbor(j);
+            CHECK(json::from_cbor(packed) == j);
+        }
+    }
+
+    SECTION("a user-defined SAX consumer is unaffected by the internal DOM reserve optimization")
+    {
+        // the reserve() call is local to json_sax_dom_parser / json_sax_dom_callback_parser;
+        // a custom SAX consumer that does not touch a DOM array sees identical events
+        json j = json::array();
+        for (int i = 0; i < 100; ++i)
+        {
+            j.push_back(i);
+        }
+        const auto packed = json::to_cbor(j);
+
+        SaxCountdown scp(1000000); // large enough to never trigger an abort
+        CHECK(json::sax_parse(packed, &scp, json::input_format_t::cbor));
+    }
+}
+
 TEST_CASE("CBOR roundtrips" * doctest::skip())
 {
     SECTION("input from flynn")

tests/src/unit-class_parser.cpp

@@ -2259,6 +2259,86 @@ TEST_CASE("parser class")
 #endif
 }
 
+#if JSON_DIAGNOSTIC_POSITIONS
+
+TEST_CASE("diagnostic positions: value lifetime")
+{
+    SECTION("copy constructor copies positions, recursively")
+    {
+        const std::string s = R"({"a":1,"b":[1,2,3]})";
+        const json a = json::parse(s);
+        const json b = a; // NOLINT(performance-unnecessary-copy-initialization)
+
+        CHECK(b.start_pos() == a.start_pos());
+        CHECK(b.end_pos() == a.end_pos());
+        CHECK(b["b"].start_pos() == a["b"].start_pos());
+        CHECK(b["b"].end_pos() == a["b"].end_pos());
+    }
+
+    SECTION("move constructor resets the moved-from value to npos")
+    {
+        const std::string s = R"({"a":1,"b":[1,2,3]})";
+        json a = json::parse(s);
+        const auto a_start = a.start_pos();
+        const auto a_end = a.end_pos();
+
+        const json b(std::move(a));
+
+        CHECK(b.start_pos() == a_start);
+        CHECK(b.end_pos() == a_end);
+
+        CHECK(a.start_pos() == std::string::npos); // NOLINT(bugprone-use-after-move,clang-analyzer-cplusplus.Move)
+        CHECK(a.end_pos() == std::string::npos); // NOLINT(bugprone-use-after-move,clang-analyzer-cplusplus.Move)
+    }
+
+    SECTION("swap() exchanges positions along with the values")
+    {
+        // basic_json::swap() (and the friend swap() that forwards to it) used
+        // to swap only m_data.m_type/m_data.m_value, leaving
+        // start_position/end_position untouched -- unlike copy-assignment's
+        // operator=(basic_json), which swaps positions as part of its
+        // copy-and-swap implementation. After swap(a, b), each value ended up
+        // with the *other* value's content but its *own* original position.
+        // This is now fixed so that swap() is consistent with copy-assignment.
+        json a = json::parse(R"({"a":1})");
+        json b = json::parse(R"([1,2,3,4,5])");
+        const auto a_start = a.start_pos();
+        const auto a_end = a.end_pos();
+        const auto b_start = b.start_pos();
+        const auto b_end = b.end_pos();
+        // lengths (and thus end positions) differ, which is enough to tell
+        // after the swap whether positions actually moved with the values
+        CHECK(a_end != b_end);
+
+        using std::swap;
+        swap(a, b);
+
+        CHECK(a == json::parse(R"([1,2,3,4,5])"));
+        CHECK(b == json::parse(R"({"a":1})"));
+
+        CHECK(a.start_pos() == b_start);
+        CHECK(a.end_pos() == b_end);
+        CHECK(b.start_pos() == a_start);
+        CHECK(b.end_pos() == a_end);
+
+        // member swap() behaves the same as the free function
+        json c = json::parse(R"({"a":1})");
+        json d = json::parse(R"([1,2,3,4,5])");
+        const auto c_start = c.start_pos();
+        const auto c_end = c.end_pos();
+        const auto d_start = d.start_pos();
+        const auto d_end = d.end_pos();
+
+        c.swap(d);
+
+        CHECK(c.start_pos() == d_start);
+        CHECK(c.end_pos() == d_end);
+        CHECK(d.start_pos() == c_start);
+        CHECK(d.end_pos() == c_end);
+    }
+}
+#endif
+
 // this test relies on parse errors being thrown, so it is skipped when
 // exceptions are disabled (json::parse aborts instead of throwing there)
 #if !defined(JSON_NOEXCEPTION)

tests/src/unit-diagnostic-positions.cpp

@@ -75,72 +75,6 @@ TEST_CASE("Better diagnostics with positions")
         CHECK(j.end_pos() == root.size());
     }
 
-    SECTION("copying keeps the positions of nested values (#5387)")
-    {
-        // Values nested deeper than the copy constructor's descent bound are
-        // copied without the call stack, on a path that has to carry the
-        // positions over itself; shallower ones copy their containers, which
-        // bring the positions along. Both sides of the bound are checked here.
-        const auto check_copy = [](std::size_t depth, bool objects)
-        {
-            CAPTURE(depth)
-            CAPTURE(objects)
-
-            const std::string opening = objects ? R"({"a":)" : "[";
-            const std::string closing = objects ? "}" : "]";
-
-            std::string text;
-            for (std::size_t i = 0; i < depth; ++i)
-            {
-                text += opening;
-            }
-            text += "12";
-            for (std::size_t i = 0; i < depth; ++i)
-            {
-                text += closing;
-            }
-
-            const json original = json::parse(text);
-            const json copy(original); // NOLINT(performance-unnecessary-copy-initialization)
-
-            const json* o = &original;
-            const json* c = &copy;
-            for (std::size_t level = 0; level <= depth; ++level)
-            {
-                CAPTURE(level)
-                REQUIRE(c->start_pos() == o->start_pos());
-                REQUIRE(c->end_pos() == o->end_pos());
-
-                if (level < depth)
-                {
-                    o = objects ? &o->at("a") : &o->at(0);
-                    c = objects ? &c->at("a") : &c->at(0);
-                }
-            }
-        };
-
-        const auto check_arrays = [&check_copy](std::size_t depth)
-        {
-            check_copy(depth, false);
-        };
-        const auto check_objects = [&check_copy](std::size_t depth)
-        {
-            check_copy(depth, true);
-        };
-
-        check_arrays(1);
-        check_arrays(127);
-        check_arrays(128);
-        check_arrays(129);
-        check_arrays(300);
-
-        check_objects(1);
-        check_objects(127);
-        check_objects(128);
-        check_objects(129);
-        check_objects(300);
-    }
-
     SECTION("JSON patch add to primitive parent (#4292)")
     {
         // the JSON Patch "add" target /foo/bar/baz has a string parent

tests/src/unit-diagnostics.cpp

@@ -273,62 +273,5 @@ TEST_CASE("Regression tests for extended diagnostics")
         CHECK(j1["numbers"]["two"] == 2);
         CHECK(j1["string"] == "t");
     }
-
-    SECTION("Regression test for issue #5387 - copying keeps the parents of nested values")
-    {
-        // A value nested deeper than the copy constructor's descent bound is
-        // copied without the call stack. Every container that path creates has
-        // to have the parents of its children set, or the JSON Pointer in the
-        // diagnostic is cut short.
-        const std::size_t depth = 300;
-
-        SECTION("objects")
-        {
-            json j = "not a number";
-            std::string pointer;
-            for (std::size_t i = 0; i < depth; ++i)
-            {
-                j = json{{"a", j}};
-                pointer += "/a";
-            }
-
-            json const copy(j); // NOLINT(performance-unnecessary-copy-initialization)
-
-            const json* inner = &copy;
-            for (std::size_t i = 0; i < depth; ++i)
-            {
-                inner = &inner->at("a");
-            }
-
-            std::string const expected = "[json.exception.type_error.302] (" + pointer + ") type must be number, but is string";
-            int i = 0;
-            CHECK_THROWS_WITH_AS(i = inner->get<int>(), expected.c_str(), json::type_error);
-            CHECK(i == 0);
-        }
-
-        SECTION("arrays")
-        {
-            json j = "not a number";
-            std::string pointer;
-            for (std::size_t i = 0; i < depth; ++i)
-            {
-                j = json::array({j});
-                pointer += "/0";
-            }
-
-            json const copy(j); // NOLINT(performance-unnecessary-copy-initialization)
-
-            const json* inner = &copy;
-            for (std::size_t i = 0; i < depth; ++i)
-            {
-                inner = &inner->at(0);
-            }
-
-            std::string const expected = "[json.exception.type_error.302] (" + pointer + ") type must be number, but is string";
-            int i = 0;
-            CHECK_THROWS_WITH_AS(i = inner->get<int>(), expected.c_str(), json::type_error);
-            CHECK(i == 0);
-        }
-    }
 }
 

tests/src/unit-large_json.cpp

@@ -12,7 +12,6 @@
 using nlohmann::json;
 
 #include <algorithm>
-#include <string>
 
 TEST_CASE("tests on very large JSONs")
 {
@@ -28,153 +27,3 @@ TEST_CASE("tests on very large JSONs")
     }
 }
 
-namespace
-{
-
-// Descend a chain of single-element containers and return the value at its end,
-// reporting the number of levels traversed in @a depth.
-//
-// The values in the test case below are nested far deeper than the call stack
-// can follow, so they must not be inspected with operator== or dump(): both are
-// still recursive and would overflow the stack themselves.
-const json* innermost_value(const json& j, std::size_t& depth)
-{
-    const json* current = &j;
-    depth = 0;
-
-    while ((current->is_array() || current->is_object()) && !current->empty())
-    {
-        current = current->is_array()
-                  ? &current->front()
-                  : &current->begin().value();
-        ++depth;
-    }
-
-    return current;
-}
-
-} // namespace
-
-TEST_CASE("tests on deeply nested JSONs")
-{
-    // deep enough to exhaust the call stack, but small enough to stay cheap:
-    // parsing is iterative, so building the values below costs little
-    const std::size_t depth = 100000;
-
-    SECTION("issue #5387 - stack overflow in the copy constructor")
-    {
-        SECTION("array")
-        {
-            const json j = json::parse(std::string(depth, '[') + '0' + std::string(depth, ']'));
-
-            const json copy(j); // NOLINT(performance-unnecessary-copy-initialization): the copy is what is tested
-
-            std::size_t copy_depth = 0;
-            CHECK(*innermost_value(copy, copy_depth) == 0);
-            CHECK(copy_depth == depth);
-        }
-
-        SECTION("object")
-        {
-            std::string s;
-            s.reserve((6 * depth) + 1);
-            for (std::size_t i = 0; i < depth; ++i)
-            {
-                s += "{\"a\":";
-            }
-            s += '1';
-            s.append(depth, '}');
-
-            const json j = json::parse(s);
-
-            const json copy(j); // NOLINT(performance-unnecessary-copy-initialization): the copy is what is tested
-
-            std::size_t copy_depth = 0;
-            CHECK(*innermost_value(copy, copy_depth) == 1);
-            CHECK(copy_depth == depth);
-        }
-
-        SECTION("copy assignment")
-        {
-            // operator=(basic_json) takes its argument by value, so the deep
-            // copy happens in the copy constructor
-            const json j = json::parse(std::string(depth, '[') + '0' + std::string(depth, ']'));
-
-            json target;
-            target = j;
-
-            std::size_t target_depth = 0;
-            CHECK(*innermost_value(target, target_depth) == 0);
-            CHECK(target_depth == depth);
-        }
-
-        SECTION("depths around the bound of the recursive descent")
-        {
-            // The copy constructor descends into a bounded number of levels and
-            // completes whatever is below that without the call stack. Cover
-            // every depth around that bound, so that the two ways of copying
-            // are known to meet cleanly - wherever the bound is set.
-            for (std::size_t d = 1; d <= 300; ++d)
-            {
-                CAPTURE(d);
-
-                const json array = json::parse(std::string(d, '[') + '0' + std::string(d, ']'));
-                const json array_copy(array); // NOLINT(performance-unnecessary-copy-initialization): the copy is what is tested
-                std::size_t array_depth = 0;
-                CHECK(*innermost_value(array_copy, array_depth) == 0);
-                CHECK(array_depth == d);
-
-                std::string object_text;
-                for (std::size_t i = 0; i < d; ++i)
-                {
-                    object_text += "{\"a\":";
-                }
-                object_text += '1';
-                object_text.append(d, '}');
-
-                const json object = json::parse(object_text);
-                const json object_copy(object); // NOLINT(performance-unnecessary-copy-initialization): the copy is what is tested
-                std::size_t object_depth = 0;
-                CHECK(*innermost_value(object_copy, object_depth) == 1);
-                CHECK(object_depth == d);
-            }
-        }
-
-        SECTION("a value that is deep in one place only")
-        {
-            json j = json::object();
-            j["shallow"] = 1;
-            j["deep"] = json::parse(std::string(depth, '[') + '0' + std::string(depth, ']'));
-            j["also_shallow"] = json::array({1, 2, 3});
-
-            const json copy(j);
-
-            CHECK(copy["shallow"] == 1);
-            CHECK(copy["also_shallow"] == json::array({1, 2, 3}));
-
-            std::size_t deep_depth = 0;
-            CHECK(*innermost_value(copy["deep"], deep_depth) == 0);
-            CHECK(deep_depth == depth);
-        }
-
-        SECTION("the copy is independent of the original")
-        {
-            const json j = json::parse(std::string(depth, '[') + '0' + std::string(depth, ']'));
-
-            json copy(j);
-
-            // reach the innermost value without recursing and replace it
-            json* current = &copy;
-            while (current->is_array() && !current->empty())
-            {
-                current = &current->front();
-            }
-            *current = 42;
-
-            std::size_t unused = 0;
-            CHECK(*innermost_value(copy, unused) == 42);
-            CHECK(*innermost_value(j, unused) == 0);
-        }
-    }
-}
-

tests/src/unit-msgpack.cpp

@@ -1597,6 +1597,91 @@ TEST_CASE("MessagePack")
     }
 }
 
+TEST_CASE("issue #5405 - array reserve for definite-length MessagePack arrays")
+{
+#if !defined(JSON_NOEXCEPTION)
+    // this SECTION relies on catching a thrown exception to distinguish
+    // which of two acceptable, bounded rejections a hostile header took;
+    // under JSON_NOEXCEPTION, JSON_THROW never produces a catchable C++
+    // exception (it aborts instead), so this cannot be tested that way here
+    SECTION("a huge claimed length with no element data must not over-allocate")
+    {
+        // 0xdd: array 32 (four-byte length); claims 0xFFFFFFFF (4294967295)
+        // elements but provides none. max_size() for a std::vector is far
+        // larger than this count, so it does not reject the header outright;
+        // the (capped) reservation must not attempt to allocate space for
+        // billions of elements before the missing data is detected.
+        json _;
+        const std::vector<uint8_t> input = {0xdd, 0xFF, 0xFF, 0xFF, 0xFF};
+        // On a platform where std::size_t is narrower than 64 bits (e.g.
+        // 32-bit), the claimed count 0xFFFFFFFF coincides with that
+        // platform's SIZE_MAX, which some size-narrowing checks treat the
+        // same as detail::unknown_size(); it may then be rejected before
+        // the SAX consumer's own max_size() check (out_of_range.408) rather
+        // than being accepted and only found short of data once the
+        // (capped) reservation looks for element bytes that were never
+        // provided (parse_error.110). Either is an acceptable, bounded
+        // rejection of the hostile header -- the property under test is
+        // that no path attempts to allocate space for billions of elements.
+        bool threw = false;
+        try
+        {
+            _ = json::from_msgpack(input);
+        }
+        catch (const json::parse_error& e)
+        {
+            threw = true;
+            CHECK(e.id == 110);
+            CHECK(std::string(e.what()) == "[json.exception.parse_error.110] parse error at byte 6: syntax error while parsing MessagePack value: unexpected end of input");
+        }
+        catch (const json::out_of_range& e)
+        {
+            threw = true;
+            CHECK(e.id == 408);
+            CHECK(std::string(e.what()).find("excessive") != std::string::npos);
+        }
+        CHECK(threw);
+        CHECK(json::from_msgpack(input, true, false).is_discarded());
+    }
+#endif
+
+    SECTION("arrays of various sizes decode to the same value as before the reserve optimization")
+    {
+        for (const auto size :
+                {
+                    std::size_t{0}, std::size_t{1}, std::size_t{5}, // small
+                    std::size_t{16384},                             // exactly at the reserve cap
+                    std::size_t{20000}                              // above the reserve cap
+                })
+        {
+            CAPTURE(size)
+            json j = json::array();
+            for (std::size_t i = 0; i < size; ++i)
+            {
+                j.push_back(static_cast<int>(i % 1000));
+            }
+
+            const auto packed = json::to_msgpack(j);
+            CHECK(json::from_msgpack(packed) == j);
+        }
+    }
+
+    SECTION("a user-defined SAX consumer is unaffected by the internal DOM reserve optimization")
+    {
+        // the reserve() call is local to json_sax_dom_parser / json_sax_dom_callback_parser;
+        // a custom SAX consumer that does not touch a DOM array sees identical events
+        json j = json::array();
+        for (int i = 0; i < 100; ++i)
+        {
+            j.push_back(i);
+        }
+        const auto packed = json::to_msgpack(j);
+
+        SaxCountdown scp(1000000); // large enough to never trigger an abort
+        CHECK(json::sax_parse(packed, &scp, json::input_format_t::msgpack));
+    }
+}
+
 // use this testcase outside [hide] to run it with Valgrind
 TEST_CASE("MessagePack nesting does not consume the call stack")
 {

tests/src/unit-ordered_json.cpp

@@ -81,37 +81,3 @@ TEST_CASE("regression test for issue #3732 - iteration_proxy_value<iter_impl<ord
     };
     static_cast<void>(fn);
 }
-
-TEST_CASE("copying an ordered_json with nested values")
-{
-    // ordered_map is backed by a vector, so copying an object that has
-    // structured values takes a different route than copying a std::map-backed
-    // one; see https://github.com/nlohmann/json/issues/5387
-    ordered_json oj;
-    oj["z"] = 1;
-    oj["a"]["y"] = 2;
-    oj["a"]["b"]["x"] = 3;
-    oj["m"] = {1, 2, {{"w", 4}}};
-
-    const ordered_json copy(oj);
-
-    SECTION("the copy is equal to the original")
-    {
-        CHECK(copy == oj);
-        CHECK(copy.dump() == oj.dump());
-    }
-
-    SECTION("the key order is preserved at every level")
-    {
-        CHECK(copy.dump() == R"({"z":1,"a":{"y":2,"b":{"x":3}},"m":[1,2,{"w":4}]})");
-    }
-
-    SECTION("the copy is independent of the original")
-    {
-        ordered_json mutated(oj);
-        mutated["a"]["b"]["x"] = 99;
-
-        CHECK(oj["a"]["b"]["x"] == 3);
-        CHECK(mutated["a"]["b"]["x"] == 99);
-    }
-}

tests/src/unit-regression3.cpp

@@ -27,6 +27,7 @@ using ordered_json = nlohmann::ordered_json;
 #endif
 
 #include <cstdio>
+#include <deque>
 #include <list>
 #include <type_traits>
 #include <utility>
@@ -896,4 +897,49 @@ TEST_CASE("issue #5402 - update(merge_objects=true) overwrites a primitive with
 }
 
 
+TEST_CASE("regression test #5476 - array type without reserve()")
+{
+    // the capacity reserved for definite-length arrays must not require the
+    // array type to have a reserve() member function
+    using deque_json = nlohmann::basic_json<std::map, std::deque>;
+
+    SECTION("std::deque")
+    {
+        const auto j = deque_json::parse(R"({"a":[1,[2,3]],"b":[]})");
+        CHECK(j.dump() == R"({"a":[1,[2,3]],"b":[]})");
+
+        // the binary formats pass a definite length to start_array()
+        CHECK(deque_json::from_cbor(deque_json::to_cbor(j)) == j);
+        CHECK(deque_json::from_msgpack(deque_json::to_msgpack(j)) == j);
+
+        // parse() instantiates the callback parser as well, which reserves too
+        const auto with_callback = deque_json::parse(R"([1,2,3])", [](int /*depth*/, deque_json::parse_event_t /*event*/, deque_json& /*parsed*/) noexcept
+        {
+            return true;
+        });
+        CHECK(with_callback == deque_json({1, 2, 3}));
+    }
+
+    SECTION("std::vector still reserves")
+    {
+        json array = json::array();
+        for (int i = 0; i < 100; ++i)
+        {
+            array.push_back(i);
+        }
+
+        const auto j = json::from_cbor(json::to_cbor(array));
+        CHECK(j == array);
+        CHECK(j.get_ref<const json::array_t&>().capacity() >= 100);
+    }
+
+    SECTION("the reservation stays capped")
+    {
+        // CBOR array announcing 2^32-1 elements, but truncated right after the
+        // header: the input must be rejected without reserving that capacity
+        const std::vector<std::uint8_t> truncated = {0x9A, 0xFF, 0xFF, 0xFF, 0xFF};
+        CHECK(json::from_cbor(truncated, true, false).is_discarded());
+    }
+}
+
 DOCTEST_CLANG_SUPPRESS_WARNING_POP

tests/src/unit-serialization.cpp

@@ -469,7 +469,7 @@ TEST_CASE("serialization of strings (bulk fast path)")
     SECTION("invalid UTF-8 handling is unaffected by the fast path")
     {
         const json j = std::string("valid\xff" "more");
-        CHECK_THROWS_WITH_AS(j.dump(), "[json.exception.type_error.316] invalid UTF-8 byte at index 5: 0xFF", json::type_error&);
+        CHECK_THROWS_WITH_AS(utils::ignore_return_value(j.dump()), "[json.exception.type_error.316] invalid UTF-8 byte at index 5: 0xFF", json::type_error&);
         CHECK(j.dump(-1, ' ', false, json::error_handler_t::replace) == "\"valid\xef\xbf\xbd" "more\"");
         CHECK(j.dump(-1, ' ', true, json::error_handler_t::replace) == "\"valid\\ufffdmore\"");
         CHECK(j.dump(-1, ' ', false, json::error_handler_t::ignore) == "\"validmore\"");

tests/src/unit-ubjson.cpp

@@ -2315,6 +2315,112 @@ TEST_CASE("UBJSON optimized arrays of a valueless type are bounded")
     }
 }
 
+TEST_CASE("issue #5405 - array reserve for definite-length UBJSON arrays")
+{
+#if !defined(JSON_NOEXCEPTION)
+    // this SECTION relies on catching a thrown exception to distinguish
+    // which of two acceptable, bounded rejections a hostile header took;
+    // under JSON_NOEXCEPTION, JSON_THROW never produces a catchable C++
+    // exception (it aborts instead), so this cannot be tested that way here
+    SECTION("a huge claimed length with no element data must not over-allocate")
+    {
+        // optimized form [$type#count: type 'i' (int8), count as a four-byte
+        // 'l' (int32) of 0x7FFFFFFF (2147483647), but no element data at all.
+        // max_size() for a std::vector is far larger than this count, so it
+        // does not reject the header outright; the (capped) reservation must
+        // not attempt to allocate space for billions of elements before the
+        // missing data is detected.
+        json _;
+        const std::vector<uint8_t> input = {'[', '$', 'i', '#', 'l', 0x7F, 0xFF, 0xFF, 0xFF};
+        // On a platform where std::vector<json>::max_size() is smaller than
+        // the claimed count (e.g. 32-bit, where max_size() is bounded by a
+        // 32-bit SIZE_MAX divided by sizeof(json)), the SAX consumer's own
+        // check rejects the header outright (out_of_range.408, with the
+        // claimed count in the message) instead of accepting it and only
+        // finding it short of data once the (capped) reservation looks for
+        // element bytes that were never provided (parse_error.110). Either
+        // is an acceptable, bounded rejection of the hostile header -- the
+        // property under test is that no path attempts to allocate space
+        // for billions of elements.
+        bool threw = false;
+        try
+        {
+            _ = json::from_ubjson(input);
+        }
+        catch (const json::parse_error& e)
+        {
+            threw = true;
+            CHECK(e.id == 110);
+            CHECK(std::string(e.what()) == "[json.exception.parse_error.110] parse error at byte 10: syntax error while parsing UBJSON number: unexpected end of input");
+        }
+        catch (const json::out_of_range& e)
+        {
+            threw = true;
+            CHECK(e.id == 408);
+            CHECK(std::string(e.what()).find("excessive array size") != std::string::npos);
+        }
+        CHECK(threw);
+
+        // json_sax_dom_parser::start_array()'s max_size() check (unlike the
+        // scanner's own parse_error path) throws unconditionally via
+        // JSON_THROW rather than going through sax->parse_error(), so it is
+        // not gated by allow_exceptions=false on a platform where this
+        // header hits that check (e.g. 32-bit, see above) -- allow either
+        // a discarded result or the same out_of_range it throws with
+        // exceptions enabled.
+        try
+        {
+            CHECK(json::from_ubjson(input, true, false).is_discarded());
+        }
+        catch (const json::out_of_range& e)
+        {
+            CHECK(e.id == 408);
+        }
+    }
+#endif
+
+    SECTION("arrays of various sizes decode to the same value as before the reserve optimization")
+    {
+        for (const auto size :
+                {
+                    std::size_t{0}, std::size_t{1}, std::size_t{5}, // small
+                    std::size_t{16384},                             // exactly at the reserve cap
+                    std::size_t{20000}                              // above the reserve cap
+                })
+        {
+            CAPTURE(size)
+            json j = json::array();
+            for (std::size_t i = 0; i < size; ++i)
+            {
+                j.push_back(static_cast<int>(i % 1000));
+            }
+
+            // exercise both the plain and the optimized [$type#count encoding
+            const auto packed_plain = json::to_ubjson(j);
+            CHECK(json::from_ubjson(packed_plain) == j);
+
+            const auto packed_optimized = json::to_ubjson(j, true, true);
+            CHECK(json::from_ubjson(packed_optimized) == j);
+        }
+    }
+
+    SECTION("a user-defined SAX consumer is unaffected by the internal DOM reserve optimization")
+    {
+        // the reserve() call is local to json_sax_dom_parser / json_sax_dom_callback_parser;
+        // a custom SAX consumer that does not touch a DOM array sees identical events
+        json j = json::array();
+        for (int i = 0; i < 100; ++i)
+        {
+            j.push_back(i);
+        }
+        const auto packed = json::to_ubjson(j, true, true);
+
+        SaxCountdown scp(1000000); // large enough to never trigger an abort
+        CHECK(json::sax_parse(packed, &scp, json::input_format_t::ubjson));
+    }
+}
+
+
 TEST_CASE("Universal Binary JSON Specification Examples 1")
 {
     SECTION("Null Value")