UBJSON/BJData readers: read optimized arrays of numbers in one loop

An array with a type marker for a fixed-size number ($U, $i, $I, $l, $L,
$d, $D, and BJData's $u, $m, $M), including the data of a BJData ndarray,
gives its elements no markers. When the input is contiguous, read the
elements that are in the input in one loop - load, byte swap for the
format's byte order, emit - instead of one trip through
parse_ubjson_internal and get_ubjson_value per element. Elements the
input ends before, other types, and non-contiguous inputs are read as
before.

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

include/nlohmann/detail/input/binary_reader.hpp

@@ -3546,6 +3546,107 @@ class binary_reader
         return unexpected_byte("invalid byte", "value");
     }
 
+    /*!
+    @brief read the elements of an optimized array of numbers (type marker
+           @a type) that are in the input, in one loop, without the
+           per-element bookkeeping of @ref parse_ubjson_internal
+
+    Values are emitted as @ref get_ubjson_value emits them. Types other than
+    the fixed-size integers and floats are left to @ref get_ubjson_value, as
+    are the elements the input ends before.
+
+    @param[in,out] remaining  the number of elements not yet read
+    @param[in] type  the type marker of the elements
+    @return whether the SAX parser accepted the values
+    */
+    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':
+                return parse_ubjson_typed_run<std::uint8_t>(remaining);
+            case 'i':
+                return parse_ubjson_typed_run<std::int8_t>(remaining);
+            case 'I':
+                return parse_ubjson_typed_run<std::int16_t>(remaining);
+            case 'l':
+                return parse_ubjson_typed_run<std::int32_t>(remaining);
+            case 'L':
+                return parse_ubjson_typed_run<std::int64_t>(remaining);
+            case 'u':
+                return !bjdata || parse_ubjson_typed_run<std::uint16_t>(remaining);
+            case 'm':
+                return !bjdata || parse_ubjson_typed_run<std::uint32_t>(remaining);
+            case 'M':
+                return !bjdata || parse_ubjson_typed_run<std::uint64_t>(remaining);
+            case 'd':
+                return parse_ubjson_typed_run<float>(remaining);
+            case 'D':
+                return parse_ubjson_typed_run<double>(remaining);
+            default:
+                return true;
+        }
+    }
+
+    /// input that is not contiguous: every element is read by get_ubjson_value
+    bool parse_ubjson_typed_run(std::size_t& /*remaining*/, const char_int_type /*type*/, std::false_type /*bulk*/) const noexcept
+    {
+        return true;
+    }
+
+    template<typename NumberType>
+    bool parse_ubjson_typed_run(std::size_t& remaining)
+    {
+        constexpr std::size_t size = sizeof(NumberType);
+        const std::size_t count = (std::min)(remaining, ia.bulk_remaining() / size);
+        if (count == 0)
+        {
+            return true;
+        }
+        const auto* const data = reinterpret_cast<const unsigned char*>(ia.bulk_data());
+        // UBJSON stores numbers big-endian, BJData little-endian
+        const bool swap = is_little_endian != (input_format == input_format_t::bjdata);
+        std::size_t read = 0;
+        bool accepted = true;
+        while (read < count && accepted)
+        {
+            NumberType number{};
+            std::memcpy(&number, data + (read * size), size);
+            if (swap)
+            {
+                byte_swap(number);
+            }
+            ++read;
+            accepted = emit_ubjson_number(number);
+        }
+        // as if read by get_number(): current is the last byte read
+        ia.bulk_skip(read * size);
+        chars_read += read * size;
+        current = data[(read * size) - 1];
+        remaining -= read;
+        return accepted;
+    }
+
+    template<typename NumberType, enable_if_t<std::is_floating_point<NumberType>::value, int> = 0>
+    bool emit_ubjson_number(const NumberType number)
+    {
+        return emit_float(number);
+    }
+
+    template < typename NumberType, enable_if_t < std::is_integral<NumberType>::value&& std::is_signed<NumberType>::value, int > = 0 >
+    bool emit_ubjson_number(const NumberType number)
+    {
+        return emit_signed(number);
+    }
+
+    template < typename NumberType, enable_if_t < std::is_integral<NumberType>::value&& !std::is_signed<NumberType>::value, int > = 0 >
+    bool emit_ubjson_number(const NumberType number)
+    {
+        return emit_unsigned(number);
+    }
+
     /*!
     @return whether array creation completed
     */
@@ -3581,7 +3682,15 @@ class binary_reader
                 return false;
             }
 
-            for (std::size_t i = 0; i < size_and_type.first; ++i)
+            // the numbers that are in the input, in one loop; the rest (if
+            // the input ends early) one by one below
+            std::size_t remaining = size_and_type.first;
+            if (!nd_is_char && JSON_HEDLEY_UNLIKELY(!parse_ubjson_typed_run(remaining, size_and_type.second, std::integral_constant<bool, bulk_scan> {})))
+            {
+                return false;
+            }
+
+            for (std::size_t i = size_and_type.first - remaining; i < size_and_type.first; ++i)
             {
                 if (nd_is_char)
                 {
@@ -3641,7 +3750,8 @@ class binary_reader
                 container_stack.back().remaining = 0;
             }
 
-            return true;
+            return size_and_type.second == 0
+                   || parse_ubjson_typed_run(container_stack.back().remaining, size_and_type.second, std::integral_constant<bool, bulk_scan> {});
         }
 
         return enter_array(detail::unknown_size());

single_include/nlohmann/json.hpp

@@ -21323,6 +21323,107 @@ class binary_reader
         return unexpected_byte("invalid byte", "value");
     }
 
+    /*!
+    @brief read the elements of an optimized array of numbers (type marker
+           @a type) that are in the input, in one loop, without the
+           per-element bookkeeping of @ref parse_ubjson_internal
+
+    Values are emitted as @ref get_ubjson_value emits them. Types other than
+    the fixed-size integers and floats are left to @ref get_ubjson_value, as
+    are the elements the input ends before.
+
+    @param[in,out] remaining  the number of elements not yet read
+    @param[in] type  the type marker of the elements
+    @return whether the SAX parser accepted the values
+    */
+    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':
+                return parse_ubjson_typed_run<std::uint8_t>(remaining);
+            case 'i':
+                return parse_ubjson_typed_run<std::int8_t>(remaining);
+            case 'I':
+                return parse_ubjson_typed_run<std::int16_t>(remaining);
+            case 'l':
+                return parse_ubjson_typed_run<std::int32_t>(remaining);
+            case 'L':
+                return parse_ubjson_typed_run<std::int64_t>(remaining);
+            case 'u':
+                return !bjdata || parse_ubjson_typed_run<std::uint16_t>(remaining);
+            case 'm':
+                return !bjdata || parse_ubjson_typed_run<std::uint32_t>(remaining);
+            case 'M':
+                return !bjdata || parse_ubjson_typed_run<std::uint64_t>(remaining);
+            case 'd':
+                return parse_ubjson_typed_run<float>(remaining);
+            case 'D':
+                return parse_ubjson_typed_run<double>(remaining);
+            default:
+                return true;
+        }
+    }
+
+    /// input that is not contiguous: every element is read by get_ubjson_value
+    bool parse_ubjson_typed_run(std::size_t& /*remaining*/, const char_int_type /*type*/, std::false_type /*bulk*/) const noexcept
+    {
+        return true;
+    }
+
+    template<typename NumberType>
+    bool parse_ubjson_typed_run(std::size_t& remaining)
+    {
+        constexpr std::size_t size = sizeof(NumberType);
+        const std::size_t count = (std::min)(remaining, ia.bulk_remaining() / size);
+        if (count == 0)
+        {
+            return true;
+        }
+        const auto* const data = reinterpret_cast<const unsigned char*>(ia.bulk_data());
+        // UBJSON stores numbers big-endian, BJData little-endian
+        const bool swap = is_little_endian != (input_format == input_format_t::bjdata);
+        std::size_t read = 0;
+        bool accepted = true;
+        while (read < count && accepted)
+        {
+            NumberType number{};
+            std::memcpy(&number, data + (read * size), size);
+            if (swap)
+            {
+                byte_swap(number);
+            }
+            ++read;
+            accepted = emit_ubjson_number(number);
+        }
+        // as if read by get_number(): current is the last byte read
+        ia.bulk_skip(read * size);
+        chars_read += read * size;
+        current = data[(read * size) - 1];
+        remaining -= read;
+        return accepted;
+    }
+
+    template<typename NumberType, enable_if_t<std::is_floating_point<NumberType>::value, int> = 0>
+    bool emit_ubjson_number(const NumberType number)
+    {
+        return emit_float(number);
+    }
+
+    template < typename NumberType, enable_if_t < std::is_integral<NumberType>::value&& std::is_signed<NumberType>::value, int > = 0 >
+    bool emit_ubjson_number(const NumberType number)
+    {
+        return emit_signed(number);
+    }
+
+    template < typename NumberType, enable_if_t < std::is_integral<NumberType>::value&& !std::is_signed<NumberType>::value, int > = 0 >
+    bool emit_ubjson_number(const NumberType number)
+    {
+        return emit_unsigned(number);
+    }
+
     /*!
     @return whether array creation completed
     */
@@ -21358,7 +21459,15 @@ class binary_reader
                 return false;
             }
 
-            for (std::size_t i = 0; i < size_and_type.first; ++i)
+            // the numbers that are in the input, in one loop; the rest (if
+            // the input ends early) one by one below
+            std::size_t remaining = size_and_type.first;
+            if (!nd_is_char && JSON_HEDLEY_UNLIKELY(!parse_ubjson_typed_run(remaining, size_and_type.second, std::integral_constant<bool, bulk_scan> {})))
+            {
+                return false;
+            }
+
+            for (std::size_t i = size_and_type.first - remaining; i < size_and_type.first; ++i)
             {
                 if (nd_is_char)
                 {
@@ -21418,7 +21527,8 @@ class binary_reader
                 container_stack.back().remaining = 0;
             }
 
-            return true;
+            return size_and_type.second == 0
+                   || parse_ubjson_typed_run(container_stack.back().remaining, size_and_type.second, std::integral_constant<bool, bulk_scan> {});
         }
 
         return enter_array(detail::unknown_size());

tests/src/unit-bjdata.cpp

@@ -26,11 +26,14 @@ using nlohmann::ordered_json;
 #include <limits>
 #include <iostream>
 #include <fstream>
+#include <list>
 #include <set>
 #include "make_test_data_available.hpp"
 #include "test_utils.hpp"
 #include "sax_countdown.hpp"
+#include "sax_event_loggers.hpp"
 using utils::SaxCountdown;
+using utils::SaxEventLogger;
 
 
 // trait_test_arg and the "value_in_range_of trait" TEST_CASE_TEMPLATE_DEFINE
@@ -4746,3 +4749,75 @@ TEST_CASE("BJData optimized container type marker is validated for count 0")
         CHECK(json::from_bjdata(vb) == json::binary({}));
     }
 }
+
+TEST_CASE("BJData optimized arrays from contiguous and non-contiguous input")
+{
+    // contiguous input reads the numbers of an array with a type marker in
+    // one loop, other input number by number; 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
+    json document =
+    {
+        {"u8", {0, 200, 255}},
+        {"i8", {-5, 100, -128}},
+        {"i16", {-30000, 1000, 30000}},
+        {"i32", {-2000000000, 5, 70000}},
+        {"i64", {-9000000000000000000LL, 1, 5000000000LL}},
+        {"f64", {1.5, -0.1, 2.5}},
+        {"pairs", {{1.25, -3.75}, {1.25, -3.75}, json::array()}},
+        {"mixed", {1, "a", 2.5, nullptr}},
+        {"u16", {40000, 1, 65535}},
+        {"u32", {4000000000U, 1, 70000}},
+        {"u64", {18446744073709551615ULL, 1, 5000000000ULL}},
+        {"nd", {{"_ArrayType_", "double"}, {"_ArraySize_", {2, 2}}, {"_ArrayData_", {1.5, -2.5, 3.5, -4.5}}}},
+        {"nd8", {{"_ArrayType_", "uint8"}, {"_ArraySize_", {3}}, {"_ArrayData_", {1, 2, 3}}}},
+    };
+    const std::vector<std::uint8_t> encoded = json::to_bjdata(document, true, true);
+    CHECK(json::from_bjdata(encoded) == document);
+
+    for (std::size_t length = 0; length <= encoded.size(); ++length)
+    {
+        CAPTURE(length)
+        const std::vector<std::uint8_t> contiguous(encoded.begin(), encoded.begin() + static_cast<std::ptrdiff_t>(length));
+        const std::list<std::uint8_t> bytewise(contiguous.begin(), contiguous.end());
+
+        CHECK(json::from_bjdata(contiguous, true, false).dump() == json::from_bjdata(bytewise.begin(), bytewise.end(), true, false).dump());
+#if !defined(JSON_NOEXCEPTION)
+        std::string contiguous_error;
+        std::string bytewise_error;
+        try
+        {
+            static_cast<void>(json::from_bjdata(contiguous));
+        }
+        catch (const json::exception& e)
+        {
+            contiguous_error = e.what();
+        }
+        try
+        {
+            static_cast<void>(json::from_bjdata(bytewise.begin(), bytewise.end()));
+        }
+        catch (const json::exception& e)
+        {
+            bytewise_error = e.what();
+        }
+        CHECK(contiguous_error == bytewise_error);
+#endif
+
+        SaxEventLogger from_contiguous;
+        SaxEventLogger from_bytewise;
+        CHECK(json::sax_parse(contiguous, &from_contiguous, json::input_format_t::bjdata)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::bjdata));
+        CHECK(from_contiguous.events == from_bytewise.events);
+    }
+
+    for (int events = 0; events <= 80; ++events)
+    {
+        CAPTURE(events)
+        const std::list<std::uint8_t> bytewise(encoded.begin(), encoded.end());
+        SaxCountdown from_contiguous(events);
+        SaxCountdown from_bytewise(events);
+        CHECK(json::sax_parse(encoded, &from_contiguous, json::input_format_t::bjdata)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::bjdata));
+    }
+}

tests/src/unit-ubjson.cpp

@@ -21,12 +21,15 @@ using nlohmann::json;
 
 #include <iostream>
 #include <fstream>
+#include <list>
 #include <set>
 #include "make_test_data_available.hpp"
 #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("UBJSON")
@@ -3400,3 +3403,70 @@ TEST_CASE("UBJSON optimized container type marker is validated for count 0")
         }
     }
 }
+
+TEST_CASE("UBJSON optimized arrays from contiguous and non-contiguous input")
+{
+    // contiguous input reads the numbers of an array with a type marker in
+    // one loop, other input number by number; 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
+    json document =
+    {
+        {"u8", {0, 200, 255}},
+        {"i8", {-5, 100, -128}},
+        {"i16", {-30000, 1000, 30000}},
+        {"i32", {-2000000000, 5, 70000}},
+        {"i64", {-9000000000000000000LL, 1, 5000000000LL}},
+        {"f64", {1.5, -0.1, 2.5}},
+        {"pairs", {{1.25, -3.75}, {1.25, -3.75}, json::array()}},
+        {"mixed", {1, "a", 2.5, nullptr}},
+    };
+    const std::vector<std::uint8_t> encoded = json::to_ubjson(document, true, true);
+    CHECK(json::from_ubjson(encoded) == document);
+
+    for (std::size_t length = 0; length <= encoded.size(); ++length)
+    {
+        CAPTURE(length)
+        const std::vector<std::uint8_t> contiguous(encoded.begin(), encoded.begin() + static_cast<std::ptrdiff_t>(length));
+        const std::list<std::uint8_t> bytewise(contiguous.begin(), contiguous.end());
+
+        CHECK(json::from_ubjson(contiguous, true, false).dump() == json::from_ubjson(bytewise.begin(), bytewise.end(), true, false).dump());
+#if !defined(JSON_NOEXCEPTION)
+        std::string contiguous_error;
+        std::string bytewise_error;
+        try
+        {
+            static_cast<void>(json::from_ubjson(contiguous));
+        }
+        catch (const json::exception& e)
+        {
+            contiguous_error = e.what();
+        }
+        try
+        {
+            static_cast<void>(json::from_ubjson(bytewise.begin(), bytewise.end()));
+        }
+        catch (const json::exception& e)
+        {
+            bytewise_error = e.what();
+        }
+        CHECK(contiguous_error == bytewise_error);
+#endif
+
+        SaxEventLogger from_contiguous;
+        SaxEventLogger from_bytewise;
+        CHECK(json::sax_parse(contiguous, &from_contiguous, json::input_format_t::ubjson)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::ubjson));
+        CHECK(from_contiguous.events == from_bytewise.events);
+    }
+
+    for (int events = 0; events <= 80; ++events)
+    {
+        CAPTURE(events)
+        const std::list<std::uint8_t> bytewise(encoded.begin(), encoded.end());
+        SaxCountdown from_contiguous(events);
+        SaxCountdown from_bytewise(events);
+        CHECK(json::sax_parse(encoded, &from_contiguous, json::input_format_t::ubjson)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::ubjson));
+    }
+}