CBOR reader: address review of the scalar-run reader

- parse_cbor_scalar_run: set the size explicitly in every case that
  reads a value, instead of relying on a default that every other case
  overwrites.
- Remove cbor_scalar_follows(): parse_cbor_scalar_run already returns on
  the same conditions, and without the pre-check SAX canada needs another
  8.5 % fewer instructions with GCC 13 (0.915 against the previous
  version), with no row more than 0.5 % worse (Mac and GCC).
- Restore the comment on why parse_cbor_internal copies the container
  frame.

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

include/nlohmann/detail/input/binary_reader.hpp

@@ -465,17 +465,15 @@ class binary_reader
             }
 
             const std::size_t element_type_parse_position = chars_read;
-            // an array's elements are named "0", "1", ... in the wire format,
-            // and those names are not passed on: they are skipped unless
-            // error_handler_t::strict must reject ill-formed UTF-8 in them
-            const bool is_object = container_stack.back().is_object;
             key.clear();
-            if (JSON_HEDLEY_UNLIKELY(!get_bson_cstr(key, is_object || error_handler == error_handler_t::strict)))
+            if (JSON_HEDLEY_UNLIKELY(!get_bson_cstr(key)))
             {
                 return false;
             }
 
-            if (is_object && !sax->key(key))
+            // an array's elements are named "0", "1", ... in the wire format,
+            // and those names are not passed on
+            if (container_stack.back().is_object && !sax->key(key))
             {
                 return false;
             }
@@ -491,18 +489,14 @@ class binary_reader
     @brief Parses a C-style string from the BSON input.
     @param[in,out] result  A reference to the string variable where the read
                             string is to be stored.
-    @param[in] store  whether the characters are needed; if not (the names of
-                      an array's elements), they are skipped and @a result may
-                      be left unchanged, and they are not checked against
-                      @ref error_handler
     @return `true` if the \\x00-byte indicating the end of the string was
              encountered before the EOF; false` indicates an unexpected EOF.
     */
-    bool get_bson_cstr(string_t& result, const bool store = true)
+    bool get_bson_cstr(string_t& result)
     {
-        if (get_bson_cstr_bulk(result, store, std::integral_constant<bool, bulk_scan> {}))
+        if (get_bson_cstr_bulk(result, std::integral_constant<bool, bulk_scan> {}))
         {
-            return !store || check_string_utf8(result, "key");
+            return check_string_utf8(result, "key");
         }
 
         auto out = std::back_inserter(result);
@@ -515,7 +509,7 @@ class binary_reader
             }
             if (current == 0x00)
             {
-                return !store || check_string_utf8(result, "key");
+                return check_string_utf8(result, "key");
             }
             *out++ = static_cast<typename string_t::value_type>(current);
         }
@@ -525,11 +519,10 @@ class binary_reader
     @brief read a C-style string from contiguous input in one step
 
     @param[in,out] result  the string to append to
-    @param[in] store  whether to append the characters to @a result
     @return whether the string was read; if the input has no \\x00-byte, nothing
             is read, and @ref get_bson_cstr reports the end of the input
     */
-    bool get_bson_cstr_bulk(string_t& result, const bool store, std::true_type /*bulk*/)
+    bool get_bson_cstr_bulk(string_t& result, std::true_type /*bulk*/)
     {
         const std::size_t remaining = ia.bulk_remaining();
         if (remaining == 0)
@@ -537,15 +530,18 @@ class binary_reader
             return false;
         }
         const auto* const data = reinterpret_cast<const unsigned char*>(ia.bulk_data());
-        const std::size_t length = find_zero_byte(data, remaining);
+        // a plain loop rather than std::memchr: most keys are short (array
+        // indices are keys, too), and the call would cost more than it saves
+        std::size_t length = 0;
+        while (length < remaining && data[length] != 0x00)
+        {
+            ++length;
+        }
         if (length == remaining)
         {
             return false;
         }
-        if (store)
-        {
-            result.append(reinterpret_cast<const typename string_t::value_type*>(data), length);
-        }
+        result.append(reinterpret_cast<const typename string_t::value_type*>(data), length);
         // consume the string and its \x00-byte, as the byte-wise path does
         ia.bulk_skip(length + 1);
         chars_read += length + 1;
@@ -554,7 +550,7 @@ class binary_reader
     }
 
     /// input that is not contiguous: C-style strings are read byte by byte
-    bool get_bson_cstr_bulk(string_t& /*result*/, const bool /*store*/, std::false_type /*bulk*/) const noexcept
+    bool get_bson_cstr_bulk(string_t& /*result*/, std::false_type /*bulk*/) const noexcept
     {
         return false;
     }
@@ -745,11 +741,13 @@ class binary_reader
     bool get_cbor_negative_integer()
     {
         NumberType number{};
-        if (JSON_HEDLEY_UNLIKELY(!get_number(number)))
-        {
-            return false;
-        }
+        return get_number(number) && emit_cbor_negative_integer(number);
+    }
 
+    /// pass the CBOR negative integer -1 - @a number to the SAX parser
+    template<typename NumberType>
+    bool emit_cbor_negative_integer(const NumberType number)
+    {
         // the value is -1 - number, which fits into number_integer_t
         // whenever number does; the outer cast undoes the integral promotion
         // for number_integer_t types narrower than int
@@ -974,7 +972,7 @@ class binary_reader
             case 0x95:
             case 0x96:
             case 0x97:
-                return enter_array(conditional_static_cast<std::size_t>(static_cast<unsigned int>(current) & 0x1Fu));
+                return enter_cbor_array(conditional_static_cast<std::size_t>(static_cast<unsigned int>(current) & 0x1Fu));
 
             case 0x98: // array (one-byte uint8_t for n follows)
             case 0x99: // array (two-byte uint16_t for n follow)
@@ -983,7 +981,7 @@ class binary_reader
             {
                 std::uint64_t len{};
                 std::size_t size{};
-                return get_cbor_argument(len) && get_cbor_container_size(len, size, "array") && enter_array(size);
+                return get_cbor_argument(len) && get_cbor_container_size(len, size, "array") && enter_cbor_array(size);
             }
 
             case 0x9F: // array (indefinite length)
@@ -1680,6 +1678,157 @@ class binary_reader
         return true;
     }
 
+    /*!
+    @brief read a big-endian number of type @a NumberType at @a data
+    */
+    template<typename NumberType>
+    NumberType load_big_endian(const unsigned char* data) const noexcept
+    {
+        NumberType result{};
+        std::memcpy(&result, data, sizeof(NumberType));
+        if (is_little_endian)
+        {
+            byte_swap(result);
+        }
+        return result;
+    }
+
+    /*!
+    @brief read the integers, floats, booleans, and nulls that follow in a
+           definite-length array, without the per-element bookkeeping of
+           @ref parse_cbor_internal
+
+    Stops at the end of the array, at any other kind of value, and when fewer
+    than 9 bytes (the longest of these values) are left in the input, leaving
+    those to @ref parse_cbor_value, which reads them the same way.
+
+    @param[in,out] remaining  the number of elements of the array not yet read
+    @return whether the SAX parser accepted the values
+    */
+    bool parse_cbor_scalar_run(std::size_t& remaining)
+    {
+        while (remaining != 0 && ia.bulk_remaining() >= 9)
+        {
+            const auto* const data = reinterpret_cast<const unsigned char*>(ia.bulk_data());
+            const unsigned char head = data[0];
+            std::size_t size = 0;
+            bool accepted = true;
+            switch (head)
+            {
+                case 0x18:
+                    size = 2;
+                    accepted = emit_unsigned(data[1]);
+                    break;
+                case 0x19:
+                    size = 3;
+                    accepted = emit_unsigned(load_big_endian<std::uint16_t>(data + 1));
+                    break;
+                case 0x1A:
+                    size = 5;
+                    accepted = emit_unsigned(load_big_endian<std::uint32_t>(data + 1));
+                    break;
+                case 0x1B:
+                    size = 9;
+                    accepted = emit_unsigned(load_big_endian<std::uint64_t>(data + 1));
+                    break;
+                case 0x38:
+                    size = 2;
+                    accepted = emit_cbor_negative_integer(data[1]);
+                    break;
+                case 0x39:
+                    size = 3;
+                    accepted = emit_cbor_negative_integer(load_big_endian<std::uint16_t>(data + 1));
+                    break;
+                case 0x3A:
+                    size = 5;
+                    accepted = emit_cbor_negative_integer(load_big_endian<std::uint32_t>(data + 1));
+                    break;
+                case 0x3B:
+                    size = 9;
+                    accepted = emit_cbor_negative_integer(load_big_endian<std::uint64_t>(data + 1));
+                    break;
+                case 0xF4:
+                    size = 1;
+                    accepted = sax->boolean(false);
+                    break;
+                case 0xF5:
+                    size = 1;
+                    accepted = sax->boolean(true);
+                    break;
+                case 0xF6:
+                    size = 1;
+                    accepted = sax->null();
+                    break;
+                case 0xFA:
+                    size = 5;
+                    accepted = emit_float(load_big_endian<float>(data + 1));
+                    break;
+                case 0xFB:
+                    size = 9;
+                    accepted = emit_float(load_big_endian<double>(data + 1));
+                    break;
+                default:
+                    if (head <= 0x17)
+                    {
+                        size = 1;
+                        accepted = sax->number_unsigned(static_cast<number_unsigned_t>(head));
+                    }
+                    else if (head >= 0x20 && head <= 0x37)
+                    {
+                        size = 1;
+                        accepted = sax->number_integer(static_cast<std::int8_t>(0x20 - 1 - head));
+                    }
+                    else
+                    {
+                        // any other value is read by parse_cbor_value
+                        return true;
+                    }
+                    break;
+            }
+            // as if read by get() and get_number(): current is the head byte
+            ia.bulk_skip(size);
+            chars_read += size;
+            current = head;
+            --remaining;
+            if (JSON_HEDLEY_UNLIKELY(!accepted))
+            {
+                return false;
+            }
+        }
+        return true;
+    }
+
+    /*!
+    @brief open a definite-length array and read the scalars it begins with
+           (see @ref parse_cbor_scalar_run)
+
+    Only the start of an array is tried, which covers arrays of numbers
+    (e.g. coordinates) but costs arrays of other values only one test, not one
+    per element.
+
+    @param[in] len  the number of elements
+    @return whether the SAX parser accepted the array and the values read
+    */
+    bool enter_cbor_array(const std::size_t len)
+    {
+        return enter_array(len) && (len == 0 || parse_cbor_array_start(std::integral_constant<bool, bulk_scan> {}));
+    }
+
+    /// read the scalars the array just opened begins with; out of line, so
+    /// that @ref parse_cbor_value stays small enough to be inlined into
+    /// @ref parse_cbor_internal
+    JSON_HEDLEY_NEVER_INLINE
+    bool parse_cbor_array_start(std::true_type /*bulk*/)
+    {
+        return parse_cbor_scalar_run(container_stack.back().remaining);
+    }
+
+    /// input that is not contiguous: every value is read by parse_cbor_value
+    bool parse_cbor_array_start(std::false_type /*bulk*/) const noexcept
+    {
+        return true;
+    }
+
     /*!
     @brief read a CBOR value and everything nested inside it
 
@@ -1709,6 +1858,7 @@ class binary_reader
                 // pop_back() below, which destroys the container_stack element
                 // it would otherwise alias
                 const container_frame top = container_stack.back();
+                const bool is_object = top.is_object;
                 bool at_end = false;
 
                 if (top.remaining != npos)
@@ -1720,7 +1870,7 @@ class binary_reader
                     {
                         // claim the element about to be read
                         --container_stack.back().remaining;
-                        if (top.is_object)
+                        if (is_object)
                         {
                             get();
                         }
@@ -1733,7 +1883,7 @@ class binary_reader
                     // Testing for it consumes a byte, which is the first byte
                     // of the next element when it is not one.
                     at_end = (get() == 0xFF);
-                    fetch = top.is_object;
+                    fetch = is_object;
                 }
 
                 if (at_end)
@@ -1750,7 +1900,7 @@ class binary_reader
                     continue;
                 }
 
-                if (top.is_object)
+                if (is_object)
                 {
                     if (JSON_HEDLEY_UNLIKELY(!skip_cbor_key_tags(tag_handler)))
                     {
@@ -3695,13 +3845,13 @@ class binary_reader
                 return sax->null();
 
             case 0xFB:
-                return sax->number_float(static_cast<number_float_t>(-1.0), "");
+                return sax->number_float(static_cast<number_float_t>(-1.0), no_source);
 
             case 0xFC:
-                return sax->number_float(static_cast<number_float_t>(0.0), "");
+                return sax->number_float(static_cast<number_float_t>(0.0), no_source);
 
             case 0xFD:
-                return sax->number_float(static_cast<number_float_t>(1.0), "");
+                return sax->number_float(static_cast<number_float_t>(1.0), no_source);
 
             case 0xFF: // empty string
             {
@@ -4205,7 +4355,7 @@ class binary_reader
         {
             return out_of_range_error(exception_id::number_overflow, "number overflow", "value");
         }
-        return sax->number_float(result, "");
+        return sax->number_float(result, no_source);
     }
 
     /*!
@@ -4265,7 +4415,7 @@ class binary_reader
         }();
         return sax->number_float((half & 0x8000u) != 0
                                  ? static_cast<number_float_t>(-val)
-                                 : static_cast<number_float_t>(val), "");
+                                 : static_cast<number_float_t>(val), no_source);
     }
 
     /*!
@@ -4590,6 +4740,10 @@ class binary_reader
     /// the SAX parser
     json_sax_t* sax = nullptr;
 
+    /// the (empty) source text of every number_float event: binary formats
+    /// have none, and a string_t built for each float would cost more
+    const string_t no_source{};
+
     /// the containers that have been opened and not closed yet; see @ref container_frame
     std::vector<container_frame> container_stack{};
 

include/nlohmann/detail/input/string_scan.hpp

@@ -90,34 +90,6 @@ JSON_HEDLEY_ALWAYS_INLINE std::size_t find_string_special(const unsigned char* d
     return n;
 }
 
-// return the index of the first 0x00 byte in [data, data+n), or n if there is
-// none; scans 8 bytes at a time (BSON's C-style strings)
-JSON_HEDLEY_ALWAYS_INLINE std::size_t find_zero_byte(const unsigned char* data, std::size_t n) noexcept
-{
-    constexpr std::uint64_t ones = 0x0101010101010101ull;
-    constexpr std::uint64_t high = 0x8080808080808080ull;
-    std::size_t i = 0;
-    for (; i + 8 <= n; i += 8)
-    {
-        const std::uint64_t v = read_eight_bytes(data + i);
-        const std::uint64_t zero = (v - ones) & ~v & high;
-        if (zero != 0)
-        {
-            // the lowest flagged byte is the first zero byte (see
-            // find_string_special())
-            return i + (static_cast<std::size_t>(count_trailing_zeros(zero)) / 8);
-        }
-    }
-    for (; i < n; ++i)
-    {
-        if (data[i] == 0x00)
-        {
-            return i;
-        }
-    }
-    return n;
-}
-
 // classify a byte as one the serializer must NOT copy verbatim when
 // ensure_ascii is requested: the closing quote, an escape, a control character
 // (< 0x20), DEL (0x7F), or any non-ASCII byte (>= 0x80). Everything else -

single_include/nlohmann/json.hpp

@@ -12458,34 +12458,6 @@ JSON_HEDLEY_ALWAYS_INLINE std::size_t find_string_special(const unsigned char* d
     return n;
 }
 
-// return the index of the first 0x00 byte in [data, data+n), or n if there is
-// none; scans 8 bytes at a time (BSON's C-style strings)
-JSON_HEDLEY_ALWAYS_INLINE std::size_t find_zero_byte(const unsigned char* data, std::size_t n) noexcept
-{
-    constexpr std::uint64_t ones = 0x0101010101010101ull;
-    constexpr std::uint64_t high = 0x8080808080808080ull;
-    std::size_t i = 0;
-    for (; i + 8 <= n; i += 8)
-    {
-        const std::uint64_t v = read_eight_bytes(data + i);
-        const std::uint64_t zero = (v - ones) & ~v & high;
-        if (zero != 0)
-        {
-            // the lowest flagged byte is the first zero byte (see
-            // find_string_special())
-            return i + (static_cast<std::size_t>(count_trailing_zeros(zero)) / 8);
-        }
-    }
-    for (; i < n; ++i)
-    {
-        if (data[i] == 0x00)
-        {
-            return i;
-        }
-    }
-    return n;
-}
-
 // classify a byte as one the serializer must NOT copy verbatim when
 // ensure_ascii is requested: the closing quote, an escape, a control character
 // (< 0x20), DEL (0x7F), or any non-ASCII byte (>= 0x80). Everything else -
@@ -18269,17 +18241,15 @@ class binary_reader
             }
 
             const std::size_t element_type_parse_position = chars_read;
-            // an array's elements are named "0", "1", ... in the wire format,
-            // and those names are not passed on: they are skipped unless
-            // error_handler_t::strict must reject ill-formed UTF-8 in them
-            const bool is_object = container_stack.back().is_object;
             key.clear();
-            if (JSON_HEDLEY_UNLIKELY(!get_bson_cstr(key, is_object || error_handler == error_handler_t::strict)))
+            if (JSON_HEDLEY_UNLIKELY(!get_bson_cstr(key)))
             {
                 return false;
             }
 
-            if (is_object && !sax->key(key))
+            // an array's elements are named "0", "1", ... in the wire format,
+            // and those names are not passed on
+            if (container_stack.back().is_object && !sax->key(key))
             {
                 return false;
             }
@@ -18295,18 +18265,14 @@ class binary_reader
     @brief Parses a C-style string from the BSON input.
     @param[in,out] result  A reference to the string variable where the read
                             string is to be stored.
-    @param[in] store  whether the characters are needed; if not (the names of
-                      an array's elements), they are skipped and @a result may
-                      be left unchanged, and they are not checked against
-                      @ref error_handler
     @return `true` if the \\x00-byte indicating the end of the string was
              encountered before the EOF; false` indicates an unexpected EOF.
     */
-    bool get_bson_cstr(string_t& result, const bool store = true)
+    bool get_bson_cstr(string_t& result)
     {
-        if (get_bson_cstr_bulk(result, store, std::integral_constant<bool, bulk_scan> {}))
+        if (get_bson_cstr_bulk(result, std::integral_constant<bool, bulk_scan> {}))
         {
-            return !store || check_string_utf8(result, "key");
+            return check_string_utf8(result, "key");
         }
 
         auto out = std::back_inserter(result);
@@ -18319,7 +18285,7 @@ class binary_reader
             }
             if (current == 0x00)
             {
-                return !store || check_string_utf8(result, "key");
+                return check_string_utf8(result, "key");
             }
             *out++ = static_cast<typename string_t::value_type>(current);
         }
@@ -18329,11 +18295,10 @@ class binary_reader
     @brief read a C-style string from contiguous input in one step
 
     @param[in,out] result  the string to append to
-    @param[in] store  whether to append the characters to @a result
     @return whether the string was read; if the input has no \\x00-byte, nothing
             is read, and @ref get_bson_cstr reports the end of the input
     */
-    bool get_bson_cstr_bulk(string_t& result, const bool store, std::true_type /*bulk*/)
+    bool get_bson_cstr_bulk(string_t& result, std::true_type /*bulk*/)
     {
         const std::size_t remaining = ia.bulk_remaining();
         if (remaining == 0)
@@ -18341,15 +18306,18 @@ class binary_reader
             return false;
         }
         const auto* const data = reinterpret_cast<const unsigned char*>(ia.bulk_data());
-        const std::size_t length = find_zero_byte(data, remaining);
+        // a plain loop rather than std::memchr: most keys are short (array
+        // indices are keys, too), and the call would cost more than it saves
+        std::size_t length = 0;
+        while (length < remaining && data[length] != 0x00)
+        {
+            ++length;
+        }
         if (length == remaining)
         {
             return false;
         }
-        if (store)
-        {
-            result.append(reinterpret_cast<const typename string_t::value_type*>(data), length);
-        }
+        result.append(reinterpret_cast<const typename string_t::value_type*>(data), length);
         // consume the string and its \x00-byte, as the byte-wise path does
         ia.bulk_skip(length + 1);
         chars_read += length + 1;
@@ -18358,7 +18326,7 @@ class binary_reader
     }
 
     /// input that is not contiguous: C-style strings are read byte by byte
-    bool get_bson_cstr_bulk(string_t& /*result*/, const bool /*store*/, std::false_type /*bulk*/) const noexcept
+    bool get_bson_cstr_bulk(string_t& /*result*/, std::false_type /*bulk*/) const noexcept
     {
         return false;
     }
@@ -18549,11 +18517,13 @@ class binary_reader
     bool get_cbor_negative_integer()
     {
         NumberType number{};
-        if (JSON_HEDLEY_UNLIKELY(!get_number(number)))
-        {
-            return false;
-        }
+        return get_number(number) && emit_cbor_negative_integer(number);
+    }
 
+    /// pass the CBOR negative integer -1 - @a number to the SAX parser
+    template<typename NumberType>
+    bool emit_cbor_negative_integer(const NumberType number)
+    {
         // the value is -1 - number, which fits into number_integer_t
         // whenever number does; the outer cast undoes the integral promotion
         // for number_integer_t types narrower than int
@@ -18778,7 +18748,7 @@ class binary_reader
             case 0x95:
             case 0x96:
             case 0x97:
-                return enter_array(conditional_static_cast<std::size_t>(static_cast<unsigned int>(current) & 0x1Fu));
+                return enter_cbor_array(conditional_static_cast<std::size_t>(static_cast<unsigned int>(current) & 0x1Fu));
 
             case 0x98: // array (one-byte uint8_t for n follows)
             case 0x99: // array (two-byte uint16_t for n follow)
@@ -18787,7 +18757,7 @@ class binary_reader
             {
                 std::uint64_t len{};
                 std::size_t size{};
-                return get_cbor_argument(len) && get_cbor_container_size(len, size, "array") && enter_array(size);
+                return get_cbor_argument(len) && get_cbor_container_size(len, size, "array") && enter_cbor_array(size);
             }
 
             case 0x9F: // array (indefinite length)
@@ -19484,6 +19454,157 @@ class binary_reader
         return true;
     }
 
+    /*!
+    @brief read a big-endian number of type @a NumberType at @a data
+    */
+    template<typename NumberType>
+    NumberType load_big_endian(const unsigned char* data) const noexcept
+    {
+        NumberType result{};
+        std::memcpy(&result, data, sizeof(NumberType));
+        if (is_little_endian)
+        {
+            byte_swap(result);
+        }
+        return result;
+    }
+
+    /*!
+    @brief read the integers, floats, booleans, and nulls that follow in a
+           definite-length array, without the per-element bookkeeping of
+           @ref parse_cbor_internal
+
+    Stops at the end of the array, at any other kind of value, and when fewer
+    than 9 bytes (the longest of these values) are left in the input, leaving
+    those to @ref parse_cbor_value, which reads them the same way.
+
+    @param[in,out] remaining  the number of elements of the array not yet read
+    @return whether the SAX parser accepted the values
+    */
+    bool parse_cbor_scalar_run(std::size_t& remaining)
+    {
+        while (remaining != 0 && ia.bulk_remaining() >= 9)
+        {
+            const auto* const data = reinterpret_cast<const unsigned char*>(ia.bulk_data());
+            const unsigned char head = data[0];
+            std::size_t size = 0;
+            bool accepted = true;
+            switch (head)
+            {
+                case 0x18:
+                    size = 2;
+                    accepted = emit_unsigned(data[1]);
+                    break;
+                case 0x19:
+                    size = 3;
+                    accepted = emit_unsigned(load_big_endian<std::uint16_t>(data + 1));
+                    break;
+                case 0x1A:
+                    size = 5;
+                    accepted = emit_unsigned(load_big_endian<std::uint32_t>(data + 1));
+                    break;
+                case 0x1B:
+                    size = 9;
+                    accepted = emit_unsigned(load_big_endian<std::uint64_t>(data + 1));
+                    break;
+                case 0x38:
+                    size = 2;
+                    accepted = emit_cbor_negative_integer(data[1]);
+                    break;
+                case 0x39:
+                    size = 3;
+                    accepted = emit_cbor_negative_integer(load_big_endian<std::uint16_t>(data + 1));
+                    break;
+                case 0x3A:
+                    size = 5;
+                    accepted = emit_cbor_negative_integer(load_big_endian<std::uint32_t>(data + 1));
+                    break;
+                case 0x3B:
+                    size = 9;
+                    accepted = emit_cbor_negative_integer(load_big_endian<std::uint64_t>(data + 1));
+                    break;
+                case 0xF4:
+                    size = 1;
+                    accepted = sax->boolean(false);
+                    break;
+                case 0xF5:
+                    size = 1;
+                    accepted = sax->boolean(true);
+                    break;
+                case 0xF6:
+                    size = 1;
+                    accepted = sax->null();
+                    break;
+                case 0xFA:
+                    size = 5;
+                    accepted = emit_float(load_big_endian<float>(data + 1));
+                    break;
+                case 0xFB:
+                    size = 9;
+                    accepted = emit_float(load_big_endian<double>(data + 1));
+                    break;
+                default:
+                    if (head <= 0x17)
+                    {
+                        size = 1;
+                        accepted = sax->number_unsigned(static_cast<number_unsigned_t>(head));
+                    }
+                    else if (head >= 0x20 && head <= 0x37)
+                    {
+                        size = 1;
+                        accepted = sax->number_integer(static_cast<std::int8_t>(0x20 - 1 - head));
+                    }
+                    else
+                    {
+                        // any other value is read by parse_cbor_value
+                        return true;
+                    }
+                    break;
+            }
+            // as if read by get() and get_number(): current is the head byte
+            ia.bulk_skip(size);
+            chars_read += size;
+            current = head;
+            --remaining;
+            if (JSON_HEDLEY_UNLIKELY(!accepted))
+            {
+                return false;
+            }
+        }
+        return true;
+    }
+
+    /*!
+    @brief open a definite-length array and read the scalars it begins with
+           (see @ref parse_cbor_scalar_run)
+
+    Only the start of an array is tried, which covers arrays of numbers
+    (e.g. coordinates) but costs arrays of other values only one test, not one
+    per element.
+
+    @param[in] len  the number of elements
+    @return whether the SAX parser accepted the array and the values read
+    */
+    bool enter_cbor_array(const std::size_t len)
+    {
+        return enter_array(len) && (len == 0 || parse_cbor_array_start(std::integral_constant<bool, bulk_scan> {}));
+    }
+
+    /// read the scalars the array just opened begins with; out of line, so
+    /// that @ref parse_cbor_value stays small enough to be inlined into
+    /// @ref parse_cbor_internal
+    JSON_HEDLEY_NEVER_INLINE
+    bool parse_cbor_array_start(std::true_type /*bulk*/)
+    {
+        return parse_cbor_scalar_run(container_stack.back().remaining);
+    }
+
+    /// input that is not contiguous: every value is read by parse_cbor_value
+    bool parse_cbor_array_start(std::false_type /*bulk*/) const noexcept
+    {
+        return true;
+    }
+
     /*!
     @brief read a CBOR value and everything nested inside it
 
@@ -19513,6 +19634,7 @@ class binary_reader
                 // pop_back() below, which destroys the container_stack element
                 // it would otherwise alias
                 const container_frame top = container_stack.back();
+                const bool is_object = top.is_object;
                 bool at_end = false;
 
                 if (top.remaining != npos)
@@ -19524,7 +19646,7 @@ class binary_reader
                     {
                         // claim the element about to be read
                         --container_stack.back().remaining;
-                        if (top.is_object)
+                        if (is_object)
                         {
                             get();
                         }
@@ -19537,7 +19659,7 @@ class binary_reader
                     // Testing for it consumes a byte, which is the first byte
                     // of the next element when it is not one.
                     at_end = (get() == 0xFF);
-                    fetch = top.is_object;
+                    fetch = is_object;
                 }
 
                 if (at_end)
@@ -19554,7 +19676,7 @@ class binary_reader
                     continue;
                 }
 
-                if (top.is_object)
+                if (is_object)
                 {
                     if (JSON_HEDLEY_UNLIKELY(!skip_cbor_key_tags(tag_handler)))
                     {
@@ -21499,13 +21621,13 @@ class binary_reader
                 return sax->null();
 
             case 0xFB:
-                return sax->number_float(static_cast<number_float_t>(-1.0), "");
+                return sax->number_float(static_cast<number_float_t>(-1.0), no_source);
 
             case 0xFC:
-                return sax->number_float(static_cast<number_float_t>(0.0), "");
+                return sax->number_float(static_cast<number_float_t>(0.0), no_source);
 
             case 0xFD:
-                return sax->number_float(static_cast<number_float_t>(1.0), "");
+                return sax->number_float(static_cast<number_float_t>(1.0), no_source);
 
             case 0xFF: // empty string
             {
@@ -22009,7 +22131,7 @@ class binary_reader
         {
             return out_of_range_error(exception_id::number_overflow, "number overflow", "value");
         }
-        return sax->number_float(result, "");
+        return sax->number_float(result, no_source);
     }
 
     /*!
@@ -22069,7 +22191,7 @@ class binary_reader
         }();
         return sax->number_float((half & 0x8000u) != 0
                                  ? static_cast<number_float_t>(-val)
-                                 : static_cast<number_float_t>(val), "");
+                                 : static_cast<number_float_t>(val), no_source);
     }
 
     /*!
@@ -22394,6 +22516,10 @@ class binary_reader
     /// the SAX parser
     json_sax_t* sax = nullptr;
 
+    /// the (empty) source text of every number_float event: binary formats
+    /// have none, and a string_t built for each float would cost more
+    const string_t no_source{};
+
     /// the containers that have been opened and not closed yet; see @ref container_frame
     std::vector<container_frame> container_stack{};
 

tests/src/unit-bson.cpp

@@ -1164,13 +1164,7 @@ TEST_CASE("BSON keys from contiguous and stream input")
     // contiguous input reads a key up to its \x00-byte in one step, a stream
     // reads it byte by byte; both must give the same value or error for the
     // complete document and for every truncation of it
-    json j = {{"", true}, {"k", {1, 2, 3}}, {std::string(40, 'x'), {{"nested key", "value"}}}};
-    // keys and array element names ("0" to "11") that end at every position of
-    // an 8-byte word, which contiguous input scans at a time
-    for (std::size_t length = 1; length <= 17; ++length)
-    {
-        j[std::string(length, static_cast<char>('a' + length))] = json::array({0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, {11}});
-    }
+    const json j = {{"", true}, {"k", {1, 2, 3}}, {std::string(40, 'x'), {{"nested key", "value"}}}};
     const std::vector<std::uint8_t> bson = json::to_bson(j);
     CHECK(json::from_bson(bson) == j);
 
@@ -1190,7 +1184,7 @@ TEST_CASE("BSON keys from contiguous and stream input")
         }
         try
         {
-            std::istringstream stream(std::string(reinterpret_cast<const char*>(input.data()), input.size()));
+            std::istringstream stream(std::string(input.begin(), input.end()));
             from_stream = json::from_bson(stream).dump();
         }
         catch (const json::parse_error& e)
@@ -1200,33 +1194,6 @@ TEST_CASE("BSON keys from contiguous and stream input")
         CHECK(from_vector == from_stream);
     }
 }
-
-TEST_CASE("BSON array element names")
-{
-    // an array's element names are not part of the value, so they are not
-    // passed on; error_handler_t::strict still rejects ill-formed UTF-8 in them
-    const std::vector<std::uint8_t> bson =
-    {
-        0x1D, 0x00, 0x00, 0x00, // size (little endian)
-        0x04, 'a', 0x00,        // entry: array "a"
-        0x15, 0x00, 0x00, 0x00, // size (little endian)
-        0x08, 0xFF, 0x00, 0x01, // entry: boolean named "\xFF", true
-        0x08, '1', '2', '3', '4', '5', '6', '7', '8', '9', 0x00, 0x00, // entry: boolean named "123456789", false
-        0x00,                   // end of the array
-        0x00                    // end of the document
-    };
-    const json expected = {{"a", {true, false}}};
-
-    CHECK(json::from_bson(bson) == expected);
-    CHECK(json::from_bson(bson, true, true, json::error_handler_t::replace) == expected);
-    CHECK(json::from_bson(bson, true, true, json::error_handler_t::ignore) == expected);
-
-    json _;
-    CHECK_THROWS_WITH_AS(_ = json::from_bson(bson, true, true, json::error_handler_t::strict), "[json.exception.parse_error.113] parse error at byte 14: syntax error while parsing BSON key: invalid string: ill-formed UTF-8 byte", json::parse_error&);
-
-    std::istringstream stream(std::string(reinterpret_cast<const char*>(bson.data()), bson.size()));
-    CHECK(json::from_bson(stream) == expected);
-}
 #endif
 
 TEST_CASE("Negative size of binary value")

tests/src/unit-cbor.cpp

@@ -29,7 +29,9 @@ using nlohmann::json;
 #include "round_trip_corpus.hpp"
 #include "test_utils.hpp"
 #include "sax_countdown.hpp"
+#include "sax_event_loggers.hpp"
 using utils::SaxCountdown;
+using utils::SaxEventLogger;
 
 
 TEST_CASE("CBOR")
@@ -2446,6 +2448,78 @@ TEST_CASE("CBOR SAX parsing stops at every event")
     CHECK(count_events(std::vector<std::uint8_t>({0xBF, 0x61, 'a', 0x9F, 0x01, 0xFF, 0xFF})) == 6);
 }
 
+TEST_CASE("CBOR arrays of scalars from contiguous and non-contiguous input")
+{
+    // contiguous input reads the integers, floats, booleans, and nulls an
+    // array begins with in one loop, other input value by value; both must
+    // give the same events, value, or error for the complete document and for
+    // every truncation of it, and stop at every event a SAX handler rejects
+    const std::vector<std::uint8_t> cbor =
+    {
+        0xA3,                                                       // map(3)
+        0x61, 'a',                                                  // "a"
+        0x98, 0x16,                                                 // array(22)
+        0x00, 0x17, 0x18, 0xFF, 0x19, 0x01, 0x00,                   // 0, 23, 255, 256
+        0x1A, 0x00, 0x01, 0x00, 0x00,                               // 65536
+        0x1B, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,       // 18446744073709551615
+        0x20, 0x37, 0x38, 0xFF, 0x39, 0x01, 0x00,                   // -1, -24, -256, -257
+        0x3A, 0x00, 0x01, 0x00, 0x00,                               // -65537
+        0x3B, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,       // -18446744073709551616 (a float)
+        0x3B, 0x7F, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,       // -9223372036854775808
+        0xF4, 0xF5, 0xF6,                                           // false, true, null
+        0xFA, 0x3F, 0xC0, 0x00, 0x00,                               // 1.5 (float32)
+        0xFB, 0xBF, 0xB9, 0x99, 0x99, 0x99, 0x99, 0x99, 0x9A,       // -0.1
+        0xF9, 0x3C, 0x00,                                           // 1.0 (float16, ends the loop)
+        0x01, 0x61, 'a',                                            // 1, "a"
+        0xFB, 0x40, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       // 2.5
+        0x61, 'b',                                                  // "b"
+        0x83,                                                       // array(3)
+        0x82, 0xFB, 0x3F, 0xF4, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // [1.25,
+        0xFB, 0xC0, 0x0E, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       //  -3.75]
+        0x82, 0xFB, 0x3F, 0xF4, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // [1.25,
+        0xFB, 0xC0, 0x0E, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       //  -3.75]
+        0x80,                                                       // []
+        0x61, 'c',                                                  // "c"
+        0x84, 0xA0, 0x01, 0xC1, 0x02, 0x81, 0x03                    // [{}, 1, tag(1) 2, [3]]
+    };
+    const auto tags = json::cbor_tag_handler_t::ignore;
+
+    const json j = json::from_cbor(cbor, true, true, tags);
+    CHECK(j["a"].size() == 22);
+    CHECK(j["a"][5] == 18446744073709551615u);
+    CHECK(j["a"][11].is_number_float());
+    CHECK(j["a"][11] == -18446744073709551616.0);
+    CHECK(j["a"][12] == (std::numeric_limits<std::int64_t>::min)());
+    CHECK(j["a"][17] == -0.1);
+    CHECK(j["b"] == json::parse("[[1.25, -3.75], [1.25, -3.75], []]"));
+    CHECK(j["c"] == json::parse("[{}, 1, 2, [3]]"));
+
+    for (std::size_t length = 0; length <= cbor.size(); ++length)
+    {
+        CAPTURE(length)
+        const std::vector<std::uint8_t> contiguous(cbor.begin(), cbor.begin() + static_cast<std::ptrdiff_t>(length));
+        const std::list<std::uint8_t> bytewise(contiguous.begin(), contiguous.end());
+
+        CHECK(json::from_cbor(contiguous, true, false, tags).dump() == json::from_cbor(bytewise.begin(), bytewise.end(), true, false, tags).dump());
+
+        SaxEventLogger from_contiguous;
+        SaxEventLogger from_bytewise;
+        CHECK(json::sax_parse(contiguous, &from_contiguous, json::input_format_t::cbor, true, false, false, tags)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::cbor, true, false, false, tags));
+        CHECK(from_contiguous.events == from_bytewise.events);
+    }
+
+    for (int events = 0; events <= 60; ++events)
+    {
+        CAPTURE(events)
+        const std::list<std::uint8_t> bytewise(cbor.begin(), cbor.end());
+        SaxCountdown from_contiguous(events);
+        SaxCountdown from_bytewise(events);
+        CHECK(json::sax_parse(cbor, &from_contiguous, json::input_format_t::cbor, true, false, false, tags)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::cbor, true, false, false, tags));
+    }
+}
+
 TEST_CASE("CBOR indefinite-length strings do not recurse per chunk")
 {
     // Reading an indefinite-length string or byte array used to call itself