Merge develop into remove-dead-checks/binary-formats

Signed-off-by: Niels Lohmann <mail@nlohmann.me>

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); the equality also rejects those impossible sizes,
+    plus the terminator); open_bson_document() rejects a smaller size before
-    since at least 5 bytes are always consumed.
+    it is stored, so @a document_size is never negative here.
 
     @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(document_size < 0 || static_cast<std::size_t>(document_size) != chars_read - document_start))
+        if (JSON_HEDLEY_UNLIKELY(static_cast<std::size_t>(document_size) != chars_read - document_start))
         {
             return bson_document_size_error(document_size, chars_read - document_start);
         }
@@ -3092,16 +3092,13 @@ class binary_reader
         {
             if (size_and_type.second != 0)
             {
-                if (size_and_type.second != 'N')
+                for (std::size_t i = 0; i < size_and_type.first; ++i)
                 {
-                    for (std::size_t i = 0; i < size_and_type.first; ++i)
+                    if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, size_and_type.second)))
                     {
-                        if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, size_and_type.second)))
+                        return false;
-                        {
-                            return false;
-                        }
-                        dim.push_back(dimlen);
                     }
+                    dim.push_back(dimlen);
                 }
             }
             else
@@ -3303,11 +3300,9 @@ 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);
-                        if (JSON_HEDLEY_UNLIKELY(type_name == nullptr))
+                        JSON_ASSERT(type_name != nullptr);
-                        {
-                            return unexpected_byte("invalid byte", "type");
-                        }
 
                         string_t type_key = "_ArrayType_";
                         string_t type = type_name; // sax->string() takes a reference
@@ -3390,12 +3385,6 @@ 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;
@@ -3423,7 +3412,7 @@ 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 (input_format == input_format_t::bjdata && is_ndarray && !inside_ndarray)
+            if (is_ndarray && !inside_ndarray)
             {
                 result.second |= (1 << 8); // use bit 8 to indicate ndarray, all UBJSON and BJData markers should be ASCII letters
             }
@@ -3433,7 +3422,7 @@ class binary_reader
         if (current == '#')
         {
             const bool is_error = get_ubjson_size_value(result.first, is_ndarray);
-            if (input_format == input_format_t::bjdata && is_ndarray && !inside_ndarray)
+            if (is_ndarray && !inside_ndarray)
             {
                 return last_byte_error(exception_id::unexpected_byte, "ndarray requires both type and size", "size");
             }
@@ -3607,7 +3596,6 @@ 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':
@@ -3621,11 +3609,11 @@ class binary_reader
             case 'L':
                 return parse_ubjson_typed_run<std::int64_t>(remaining);
             case 'u':
-                return !bjdata || parse_ubjson_typed_run<std::uint16_t>(remaining);
+                return parse_ubjson_typed_run<std::uint16_t>(remaining);
             case 'm':
-                return !bjdata || parse_ubjson_typed_run<std::uint32_t>(remaining);
+                return parse_ubjson_typed_run<std::uint32_t>(remaining);
             case 'M':
-                return !bjdata || parse_ubjson_typed_run<std::uint64_t>(remaining);
+                return parse_ubjson_typed_run<std::uint64_t>(remaining);
             case 'd':
                 return parse_ubjson_typed_run<float>(remaining);
             case 'D':
@@ -3706,7 +3694,7 @@ 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 (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
+        if ((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
 
@@ -3765,7 +3753,7 @@ class binary_reader
         }
 
         // If BJData type marker is 'B' decode as binary
-        if (input_format == input_format_t::bjdata && size_and_type.first != npos && size_and_type.second == 'B')
+        if (size_and_type.first != npos && size_and_type.second == 'B')
         {
             binary_t result;
             return get_binary(size_and_type.first, result) && sax->binary(result);
@@ -3814,7 +3802,7 @@ class binary_reader
         }
 
         // do not accept ND-array size in objects in BJData
-        if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
+        if ((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");
         }
@@ -5095,8 +5083,9 @@ 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
+    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).
+    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.
     */
     bool is_valid_optimized_type(const char_int_type marker) const noexcept
     {
@@ -5131,20 +5120,6 @@ 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

@@ -3126,22 +3126,19 @@ 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 +4054,7 @@ 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_bjd_excluded_optimized_type()).
+    list (binary_reader::is_valid_optimized_type()).
     */
     static constexpr bool is_bjdata_excluded_type_marker(const CharType marker) noexcept
     {

single_include/nlohmann/json.hpp

@@ -18359,8 +18359,8 @@ 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); the equality also rejects those impossible sizes,
+    plus the terminator); open_bson_document() rejects a smaller size before
-    since at least 5 bytes are always consumed.
+    it is stored, so @a document_size is never negative here.
 
     @param[in] document_start  value of chars_read before the size prefix
     @param[in] document_size   the declared document size
@@ -18368,7 +18368,7 @@ class binary_reader
     */
     bool check_bson_document_size(const std::size_t document_start, const std::int32_t document_size)
     {
-        if (JSON_HEDLEY_UNLIKELY(document_size < 0 || static_cast<std::size_t>(document_size) != chars_read - document_start))
+        if (JSON_HEDLEY_UNLIKELY(static_cast<std::size_t>(document_size) != chars_read - document_start))
         {
             return bson_document_size_error(document_size, chars_read - document_start);
         }
@@ -21107,16 +21107,13 @@ class binary_reader
         {
             if (size_and_type.second != 0)
             {
-                if (size_and_type.second != 'N')
+                for (std::size_t i = 0; i < size_and_type.first; ++i)
                 {
-                    for (std::size_t i = 0; i < size_and_type.first; ++i)
+                    if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, size_and_type.second)))
                     {
-                        if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, size_and_type.second)))
+                        return false;
-                        {
-                            return false;
-                        }
-                        dim.push_back(dimlen);
                     }
+                    dim.push_back(dimlen);
                 }
             }
             else
@@ -21318,11 +21315,9 @@ 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);
-                        if (JSON_HEDLEY_UNLIKELY(type_name == nullptr))
+                        JSON_ASSERT(type_name != nullptr);
-                        {
-                            return unexpected_byte("invalid byte", "type");
-                        }
 
                         string_t type_key = "_ArrayType_";
                         string_t type = type_name; // sax->string() takes a reference
@@ -21405,12 +21400,6 @@ 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;
@@ -21438,7 +21427,7 @@ 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 (input_format == input_format_t::bjdata && is_ndarray && !inside_ndarray)
+            if (is_ndarray && !inside_ndarray)
             {
                 result.second |= (1 << 8); // use bit 8 to indicate ndarray, all UBJSON and BJData markers should be ASCII letters
             }
@@ -21448,7 +21437,7 @@ class binary_reader
         if (current == '#')
         {
             const bool is_error = get_ubjson_size_value(result.first, is_ndarray);
-            if (input_format == input_format_t::bjdata && is_ndarray && !inside_ndarray)
+            if (is_ndarray && !inside_ndarray)
             {
                 return last_byte_error(exception_id::unexpected_byte, "ndarray requires both type and size", "size");
             }
@@ -21622,7 +21611,6 @@ 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':
@@ -21636,11 +21624,11 @@ class binary_reader
             case 'L':
                 return parse_ubjson_typed_run<std::int64_t>(remaining);
             case 'u':
-                return !bjdata || parse_ubjson_typed_run<std::uint16_t>(remaining);
+                return parse_ubjson_typed_run<std::uint16_t>(remaining);
             case 'm':
-                return !bjdata || parse_ubjson_typed_run<std::uint32_t>(remaining);
+                return parse_ubjson_typed_run<std::uint32_t>(remaining);
             case 'M':
-                return !bjdata || parse_ubjson_typed_run<std::uint64_t>(remaining);
+                return parse_ubjson_typed_run<std::uint64_t>(remaining);
             case 'd':
                 return parse_ubjson_typed_run<float>(remaining);
             case 'D':
@@ -21721,7 +21709,7 @@ 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 (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
+        if ((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
 
@@ -21780,7 +21768,7 @@ class binary_reader
         }
 
         // If BJData type marker is 'B' decode as binary
-        if (input_format == input_format_t::bjdata && size_and_type.first != npos && size_and_type.second == 'B')
+        if (size_and_type.first != npos && size_and_type.second == 'B')
         {
             binary_t result;
             return get_binary(size_and_type.first, result) && sax->binary(result);
@@ -21829,7 +21817,7 @@ class binary_reader
         }
 
         // do not accept ND-array size in objects in BJData
-        if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
+        if ((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");
         }
@@ -23110,8 +23098,9 @@ 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
+    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).
+    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.
     */
     bool is_valid_optimized_type(const char_int_type marker) const noexcept
     {
@@ -23146,20 +23135,6 @@ 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
@@ -31114,22 +31089,19 @@ 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));
@@ -32045,7 +32017,7 @@ 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_bjd_excluded_optimized_type()).
+    list (binary_reader::is_valid_optimized_type()).
     */
     static constexpr bool is_bjdata_excluded_type_marker(const CharType marker) noexcept
     {

tests/src/unit-bjdata.cpp

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