Declare the value-initialized iterator in the test const

clang-tidy (misc-const-correctness) flags it, as it is only compared.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01N5NasAgbDGG28MqXjQvLqp
Signed-off-by: Claude <noreply@anthropic.com>

include/nlohmann/detail/exceptions.hpp

@@ -222,7 +222,8 @@ class exception : public std::exception
     template<typename BasicJsonType>
     static std::string get_byte_positions(const BasicJsonType* leaf_element)
     {
-        if ((leaf_element->start_pos() != std::string::npos) && (leaf_element->end_pos() != std::string::npos))
+        // a value-initialized iterator has no value (invalid_iterator.212)
+        if (leaf_element != nullptr && leaf_element->start_pos() != std::string::npos && leaf_element->end_pos() != std::string::npos)
         {
             return concat("(bytes ", std::to_string(leaf_element->start_pos()), "-", std::to_string(leaf_element->end_pos()), ") ");
         }

include/nlohmann/detail/input/binary_reader.hpp

@@ -344,8 +344,8 @@ class binary_reader
     nested document could declare a length that disagrees with where its
     terminator actually falls and quietly hand the bytes in between to the
     enclosing document. A well-formed document is at least 5 bytes (the prefix
-    plus the terminator); open_bson_document() rejects a smaller size before
-    it is stored, so @a document_size is never negative here.
+    plus the terminator); the equality also rejects those impossible sizes,
+    since at least 5 bytes are always consumed.
 
     @param[in] document_start  value of chars_read before the size prefix
     @param[in] document_size   the declared document size
@@ -353,7 +353,7 @@ class binary_reader
     */
     bool check_bson_document_size(const std::size_t document_start, const std::int32_t document_size)
     {
-        if (JSON_HEDLEY_UNLIKELY(static_cast<std::size_t>(document_size) != chars_read - document_start))
+        if (JSON_HEDLEY_UNLIKELY(document_size < 0 || static_cast<std::size_t>(document_size) != chars_read - document_start))
         {
             return bson_document_size_error(document_size, chars_read - document_start);
         }
@@ -3084,13 +3084,16 @@ class binary_reader
         {
             if (size_and_type.second != 0)
             {
-                for (std::size_t i = 0; i < size_and_type.first; ++i)
+                if (size_and_type.second != 'N')
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, size_and_type.second)))
+                    for (std::size_t i = 0; i < size_and_type.first; ++i)
                     {
-                        return false;
+                        if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, size_and_type.second)))
+                        {
+                            return false;
+                        }
+                        dim.push_back(dimlen);
                     }
-                    dim.push_back(dimlen);
                 }
             }
             else
@@ -3292,9 +3295,11 @@ class binary_reader
                     // follow the documented _ArrayType_, _ArraySize_, _ArrayData_ order
                     if (ndarray_dtype != 0)
                     {
-                        // every optimized type marker is_valid_optimized_type() accepts for BJData has a name
                         const char* type_name = bjd_type_name(ndarray_dtype);
-                        JSON_ASSERT(type_name != nullptr);
+                        if (JSON_HEDLEY_UNLIKELY(type_name == nullptr))
+                        {
+                            return unexpected_byte("invalid byte", "type");
+                        }
 
                         string_t type_key = "_ArrayType_";
                         string_t type = type_name; // sax->string() takes a reference
@@ -3377,6 +3382,12 @@ class binary_reader
         if (current == '$')
         {
             result.second = get();  // must not ignore 'N', because 'N' maybe the type
+            if (input_format == input_format_t::bjdata
+                    && JSON_HEDLEY_UNLIKELY(is_bjd_excluded_optimized_type(result.second)))
+            {
+                return optimized_type_error();
+            }
+
             if (JSON_HEDLEY_UNLIKELY(!unexpect_eof("type")))
             {
                 return false;
@@ -3404,7 +3415,7 @@ class binary_reader
             // an ndarray was read here only if the flag flipped; when it was
             // seeded true, get_ubjson_size_value() already rejected the nested
             // dimension vector
-            if (is_ndarray && !inside_ndarray)
+            if (input_format == input_format_t::bjdata && is_ndarray && !inside_ndarray)
             {
                 result.second |= (1 << 8); // use bit 8 to indicate ndarray, all UBJSON and BJData markers should be ASCII letters
             }
@@ -3414,7 +3425,7 @@ class binary_reader
         if (current == '#')
         {
             const bool is_error = get_ubjson_size_value(result.first, is_ndarray);
-            if (is_ndarray && !inside_ndarray)
+            if (input_format == input_format_t::bjdata && is_ndarray && !inside_ndarray)
             {
                 return last_byte_error(exception_id::unexpected_byte, "ndarray requires both type and size", "size");
             }
@@ -3588,6 +3599,7 @@ class binary_reader
     JSON_HEDLEY_NEVER_INLINE
     bool parse_ubjson_typed_run(std::size_t& remaining, const char_int_type type, std::true_type /*bulk*/)
     {
+        const bool bjdata = input_format == input_format_t::bjdata;
         switch (type)
         {
             case 'U':
@@ -3601,11 +3613,11 @@ class binary_reader
             case 'L':
                 return parse_ubjson_typed_run<std::int64_t>(remaining);
             case 'u':
-                return parse_ubjson_typed_run<std::uint16_t>(remaining);
+                return !bjdata || parse_ubjson_typed_run<std::uint16_t>(remaining);
             case 'm':
-                return parse_ubjson_typed_run<std::uint32_t>(remaining);
+                return !bjdata || parse_ubjson_typed_run<std::uint32_t>(remaining);
             case 'M':
-                return parse_ubjson_typed_run<std::uint64_t>(remaining);
+                return !bjdata || parse_ubjson_typed_run<std::uint64_t>(remaining);
             case 'd':
                 return parse_ubjson_typed_run<float>(remaining);
             case 'D':
@@ -3686,7 +3698,7 @@ class binary_reader
         // if bit-8 of size_and_type.second is set to 1, encode bjdata ndarray as an object in JData annotated array format (https://github.com/NeuroJSON/jdata):
         // {"_ArrayType_" : "typeid", "_ArraySize_" : [n1, n2, ...], "_ArrayData_" : [v1, v2, ...]}
 
-        if ((size_and_type.second & (1 << 8)) != 0)
+        if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
         {
             size_and_type.second &= ~(static_cast<char_int_type>(1) << 8);  // use bit 8 to indicate ndarray, here we remove the bit to restore the type marker
 
@@ -3745,7 +3757,7 @@ class binary_reader
         }
 
         // If BJData type marker is 'B' decode as binary
-        if (size_and_type.first != npos && size_and_type.second == 'B')
+        if (input_format == input_format_t::bjdata && size_and_type.first != npos && size_and_type.second == 'B')
         {
             binary_t result;
             return get_binary(size_and_type.first, result) && sax->binary(result);
@@ -3794,7 +3806,7 @@ class binary_reader
         }
 
         // do not accept ND-array size in objects in BJData
-        if ((size_and_type.second & (1 << 8)) != 0)
+        if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
         {
             return last_byte_error(exception_id::unexpected_byte, "BJData object does not support ND-array size in optimized format", "object");
         }
@@ -5075,9 +5087,8 @@ class binary_reader
   JSON_PRIVATE_UNLESS_TESTED:
     /*!
     @brief check whether @a marker is a type marker that may follow '$'
-    UBJSON: Z N T F i U I l L d D H C S [ {; BJData: the numeric markers and C.
-    The markers BJData excludes (F H N S T Z [ {) are the ones
-    binary_writer::is_bjdata_excluded_type_marker() lists; keep the two in sync.
+    UBJSON: Z N T F i U I l L d D H C S [ {; BJData: the numeric markers and C
+    (BJData's excluded markers are rejected separately).
     */
     bool is_valid_optimized_type(const char_int_type marker) const noexcept
     {
@@ -5112,6 +5123,20 @@ class binary_reader
         }
     }
 
+    /*!
+    @brief whether @a marker is excluded from BJData's optimized ND-array types
+    @return whether @a marker is one of 'F', 'H', 'N', 'S', 'T', 'Z', '[', '{'
+
+    Mirrors binary_writer's @ref binary_writer::is_bjdata_excluded_type_marker
+    "is_bjdata_excluded_type_marker()`, which encodes the same list the other
+    way; keep the two in sync.
+    */
+    static constexpr bool is_bjd_excluded_optimized_type(const char_int_type marker) noexcept
+    {
+        return marker == '[' || marker == '{' || marker == 'S' || marker == 'H'
+               || marker == 'T' || marker == 'F' || marker == 'N' || marker == 'Z';
+    }
+
     /*!
     @brief look up the ND-array element type name for a BJData dtype marker
     @return the type name ("uint8", "int8", ...), or nullptr if @a marker does

include/nlohmann/detail/output/binary_writer.hpp

@@ -3129,19 +3129,22 @@ class binary_writer
                 // negative fixnum
                 write_number(static_cast<std::int8_t>(j.m_data.m_value.number_integer));
             }
-            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int8_t>::min)())
+            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int8_t>::min)() &&
+                     j.m_data.m_value.number_integer <= (std::numeric_limits<std::int8_t>::max)())
             {
                 // int 8
                 oa.write_character(to_char_type(0xD0));
                 write_number(static_cast<std::int8_t>(j.m_data.m_value.number_integer));
             }
-            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int16_t>::min)())
+            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int16_t>::min)() &&
+                     j.m_data.m_value.number_integer <= (std::numeric_limits<std::int16_t>::max)())
             {
                 // int 16
                 oa.write_character(to_char_type(0xD1));
                 write_number(static_cast<std::int16_t>(j.m_data.m_value.number_integer));
             }
-            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int32_t>::min)())
+            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int32_t>::min)() &&
+                     j.m_data.m_value.number_integer <= (std::numeric_limits<std::int32_t>::max)())
             {
                 // int 32
                 oa.write_character(to_char_type(0xD2));
@@ -4057,7 +4060,7 @@ class binary_writer
     Containers, strings, high-precision numbers, booleans and null cannot be
     declared as the single type of an optimized container in BJData; such a
     container is written unoptimized. The reader rejects them with the same
-    list (binary_reader::is_valid_optimized_type()).
+    list (binary_reader::is_bjd_excluded_optimized_type()).
     */
     static constexpr bool is_bjdata_excluded_type_marker(const CharType marker) noexcept
     {

single_include/nlohmann/json.hpp

@@ -5719,7 +5719,8 @@ class exception : public std::exception
     template<typename BasicJsonType>
     static std::string get_byte_positions(const BasicJsonType* leaf_element)
     {
-        if ((leaf_element->start_pos() != std::string::npos) && (leaf_element->end_pos() != std::string::npos))
+        // a value-initialized iterator has no value (invalid_iterator.212)
+        if (leaf_element != nullptr && leaf_element->start_pos() != std::string::npos && leaf_element->end_pos() != std::string::npos)
         {
             return concat("(bytes ", std::to_string(leaf_element->start_pos()), "-", std::to_string(leaf_element->end_pos()), ") ");
         }
@@ -18351,8 +18352,8 @@ class binary_reader
     nested document could declare a length that disagrees with where its
     terminator actually falls and quietly hand the bytes in between to the
     enclosing document. A well-formed document is at least 5 bytes (the prefix
-    plus the terminator); open_bson_document() rejects a smaller size before
-    it is stored, so @a document_size is never negative here.
+    plus the terminator); the equality also rejects those impossible sizes,
+    since at least 5 bytes are always consumed.
 
     @param[in] document_start  value of chars_read before the size prefix
     @param[in] document_size   the declared document size
@@ -18360,7 +18361,7 @@ class binary_reader
     */
     bool check_bson_document_size(const std::size_t document_start, const std::int32_t document_size)
     {
-        if (JSON_HEDLEY_UNLIKELY(static_cast<std::size_t>(document_size) != chars_read - document_start))
+        if (JSON_HEDLEY_UNLIKELY(document_size < 0 || static_cast<std::size_t>(document_size) != chars_read - document_start))
         {
             return bson_document_size_error(document_size, chars_read - document_start);
         }
@@ -21091,13 +21092,16 @@ class binary_reader
         {
             if (size_and_type.second != 0)
             {
-                for (std::size_t i = 0; i < size_and_type.first; ++i)
+                if (size_and_type.second != 'N')
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, size_and_type.second)))
+                    for (std::size_t i = 0; i < size_and_type.first; ++i)
                     {
-                        return false;
+                        if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, size_and_type.second)))
+                        {
+                            return false;
+                        }
+                        dim.push_back(dimlen);
                     }
-                    dim.push_back(dimlen);
                 }
             }
             else
@@ -21299,9 +21303,11 @@ class binary_reader
                     // follow the documented _ArrayType_, _ArraySize_, _ArrayData_ order
                     if (ndarray_dtype != 0)
                     {
-                        // every optimized type marker is_valid_optimized_type() accepts for BJData has a name
                         const char* type_name = bjd_type_name(ndarray_dtype);
-                        JSON_ASSERT(type_name != nullptr);
+                        if (JSON_HEDLEY_UNLIKELY(type_name == nullptr))
+                        {
+                            return unexpected_byte("invalid byte", "type");
+                        }
 
                         string_t type_key = "_ArrayType_";
                         string_t type = type_name; // sax->string() takes a reference
@@ -21384,6 +21390,12 @@ class binary_reader
         if (current == '$')
         {
             result.second = get();  // must not ignore 'N', because 'N' maybe the type
+            if (input_format == input_format_t::bjdata
+                    && JSON_HEDLEY_UNLIKELY(is_bjd_excluded_optimized_type(result.second)))
+            {
+                return optimized_type_error();
+            }
+
             if (JSON_HEDLEY_UNLIKELY(!unexpect_eof("type")))
             {
                 return false;
@@ -21411,7 +21423,7 @@ class binary_reader
             // an ndarray was read here only if the flag flipped; when it was
             // seeded true, get_ubjson_size_value() already rejected the nested
             // dimension vector
-            if (is_ndarray && !inside_ndarray)
+            if (input_format == input_format_t::bjdata && is_ndarray && !inside_ndarray)
             {
                 result.second |= (1 << 8); // use bit 8 to indicate ndarray, all UBJSON and BJData markers should be ASCII letters
             }
@@ -21421,7 +21433,7 @@ class binary_reader
         if (current == '#')
         {
             const bool is_error = get_ubjson_size_value(result.first, is_ndarray);
-            if (is_ndarray && !inside_ndarray)
+            if (input_format == input_format_t::bjdata && is_ndarray && !inside_ndarray)
             {
                 return last_byte_error(exception_id::unexpected_byte, "ndarray requires both type and size", "size");
             }
@@ -21595,6 +21607,7 @@ class binary_reader
     JSON_HEDLEY_NEVER_INLINE
     bool parse_ubjson_typed_run(std::size_t& remaining, const char_int_type type, std::true_type /*bulk*/)
     {
+        const bool bjdata = input_format == input_format_t::bjdata;
         switch (type)
         {
             case 'U':
@@ -21608,11 +21621,11 @@ class binary_reader
             case 'L':
                 return parse_ubjson_typed_run<std::int64_t>(remaining);
             case 'u':
-                return parse_ubjson_typed_run<std::uint16_t>(remaining);
+                return !bjdata || parse_ubjson_typed_run<std::uint16_t>(remaining);
             case 'm':
-                return parse_ubjson_typed_run<std::uint32_t>(remaining);
+                return !bjdata || parse_ubjson_typed_run<std::uint32_t>(remaining);
             case 'M':
-                return parse_ubjson_typed_run<std::uint64_t>(remaining);
+                return !bjdata || parse_ubjson_typed_run<std::uint64_t>(remaining);
             case 'd':
                 return parse_ubjson_typed_run<float>(remaining);
             case 'D':
@@ -21693,7 +21706,7 @@ class binary_reader
         // if bit-8 of size_and_type.second is set to 1, encode bjdata ndarray as an object in JData annotated array format (https://github.com/NeuroJSON/jdata):
         // {"_ArrayType_" : "typeid", "_ArraySize_" : [n1, n2, ...], "_ArrayData_" : [v1, v2, ...]}
 
-        if ((size_and_type.second & (1 << 8)) != 0)
+        if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
         {
             size_and_type.second &= ~(static_cast<char_int_type>(1) << 8);  // use bit 8 to indicate ndarray, here we remove the bit to restore the type marker
 
@@ -21752,7 +21765,7 @@ class binary_reader
         }
 
         // If BJData type marker is 'B' decode as binary
-        if (size_and_type.first != npos && size_and_type.second == 'B')
+        if (input_format == input_format_t::bjdata && size_and_type.first != npos && size_and_type.second == 'B')
         {
             binary_t result;
             return get_binary(size_and_type.first, result) && sax->binary(result);
@@ -21801,7 +21814,7 @@ class binary_reader
         }
 
         // do not accept ND-array size in objects in BJData
-        if ((size_and_type.second & (1 << 8)) != 0)
+        if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
         {
             return last_byte_error(exception_id::unexpected_byte, "BJData object does not support ND-array size in optimized format", "object");
         }
@@ -23082,9 +23095,8 @@ class binary_reader
   JSON_PRIVATE_UNLESS_TESTED:
     /*!
     @brief check whether @a marker is a type marker that may follow '$'
-    UBJSON: Z N T F i U I l L d D H C S [ {; BJData: the numeric markers and C.
-    The markers BJData excludes (F H N S T Z [ {) are the ones
-    binary_writer::is_bjdata_excluded_type_marker() lists; keep the two in sync.
+    UBJSON: Z N T F i U I l L d D H C S [ {; BJData: the numeric markers and C
+    (BJData's excluded markers are rejected separately).
     */
     bool is_valid_optimized_type(const char_int_type marker) const noexcept
     {
@@ -23119,6 +23131,20 @@ class binary_reader
         }
     }
 
+    /*!
+    @brief whether @a marker is excluded from BJData's optimized ND-array types
+    @return whether @a marker is one of 'F', 'H', 'N', 'S', 'T', 'Z', '[', '{'
+
+    Mirrors binary_writer's @ref binary_writer::is_bjdata_excluded_type_marker
+    "is_bjdata_excluded_type_marker()`, which encodes the same list the other
+    way; keep the two in sync.
+    */
+    static constexpr bool is_bjd_excluded_optimized_type(const char_int_type marker) noexcept
+    {
+        return marker == '[' || marker == '{' || marker == 'S' || marker == 'H'
+               || marker == 'T' || marker == 'F' || marker == 'N' || marker == 'Z';
+    }
+
     /*!
     @brief look up the ND-array element type name for a BJData dtype marker
     @return the type name ("uint8", "int8", ...), or nullptr if @a marker does
@@ -31076,19 +31102,22 @@ class binary_writer
                 // negative fixnum
                 write_number(static_cast<std::int8_t>(j.m_data.m_value.number_integer));
             }
-            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int8_t>::min)())
+            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int8_t>::min)() &&
+                     j.m_data.m_value.number_integer <= (std::numeric_limits<std::int8_t>::max)())
             {
                 // int 8
                 oa.write_character(to_char_type(0xD0));
                 write_number(static_cast<std::int8_t>(j.m_data.m_value.number_integer));
             }
-            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int16_t>::min)())
+            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int16_t>::min)() &&
+                     j.m_data.m_value.number_integer <= (std::numeric_limits<std::int16_t>::max)())
             {
                 // int 16
                 oa.write_character(to_char_type(0xD1));
                 write_number(static_cast<std::int16_t>(j.m_data.m_value.number_integer));
             }
-            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int32_t>::min)())
+            else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int32_t>::min)() &&
+                     j.m_data.m_value.number_integer <= (std::numeric_limits<std::int32_t>::max)())
             {
                 // int 32
                 oa.write_character(to_char_type(0xD2));
@@ -32004,7 +32033,7 @@ class binary_writer
     Containers, strings, high-precision numbers, booleans and null cannot be
     declared as the single type of an optimized container in BJData; such a
     container is written unoptimized. The reader rejects them with the same
-    list (binary_reader::is_valid_optimized_type()).
+    list (binary_reader::is_bjd_excluded_optimized_type()).
     */
     static constexpr bool is_bjdata_excluded_type_marker(const CharType marker) noexcept
     {

tests/src/unit-bjdata.cpp

@@ -78,25 +78,23 @@ TEST_CASE("BJData")
 {
     SECTION("binary_reader BJData lookup tables")
     {
-        // bjd_type_name() is static; is_valid_optimized_type() needs a BJData reader
+        // both lookups are static member functions
         std::vector<std::uint8_t> const data;
         using reader_t = nlohmann::detail::binary_reader<json, decltype(nlohmann::detail::input_adapter(data))>;
 
-        // the markers BJData excludes as optimized types must match
-        // binary_writer's is_bjdata_excluded_type_marker(), which encodes the
-        // same 8 markers
-        reader_t reader(nlohmann::detail::input_adapter(data), json::input_format_t::bjdata);
+        // the excluded optimized-type markers must match binary_writer's
+        // is_bjdata_excluded_type_marker(), which encodes the same 8 markers
         for (const char marker :
-                {'[', '{', 'S', 'H', 'T', 'F', 'N', 'Z', 'x'
+                {'[', '{', 'S', 'H', 'T', 'F', 'N', 'Z'
                 })
         {
-            CHECK(!reader.is_valid_optimized_type(static_cast<unsigned char>(marker)));
+            CHECK(reader_t::is_bjd_excluded_optimized_type(static_cast<unsigned char>(marker)));
         }
         for (const char marker :
-                {'U', 'i', 'u', 'I', 'm', 'l', 'M', 'L', 'd', 'D', 'C', 'B', 'h'
+                {'U', 'i', 'u', 'I', 'm', 'l', 'M', 'L', 'd', 'D', 'C', 'B', 'x'
                 })
         {
-            CHECK(reader.is_valid_optimized_type(static_cast<unsigned char>(marker)));
+            CHECK(!reader_t::is_bjd_excluded_optimized_type(static_cast<unsigned char>(marker)));
         }
 
         // every dtype marker must round-trip to its ND-array type name

tests/src/unit-diagnostic-positions.cpp

@@ -243,3 +243,14 @@ TEST_CASE("values read from a binary format have no positions")
         check_no_positions(j);
     }
 }
+
+TEST_CASE("comparing a value-initialized iterator throws invalid_iterator.212")
+{
+    // such an iterator has no value to take the byte positions from; the
+    // comparison used to dereference the null pointer instead of throwing
+    json j = {1, 2};
+    const json::iterator it{};
+    CHECK_THROWS_WITH_AS(it == j.begin(), "[json.exception.invalid_iterator.212] cannot compare iterators of different containers", json::invalid_iterator);
+    CHECK_THROWS_WITH_AS(it < j.begin(), "[json.exception.invalid_iterator.212] cannot compare iterators of different containers", json::invalid_iterator);
+    CHECK_THROWS_WITH_AS(j.begin() == it, "[json.exception.invalid_iterator.212] cannot compare iterators of different containers", json::invalid_iterator);
+}