accept() checks strings without copying them

json::accept() parses through json_sax_acceptor, which ignores the
values of strings and object keys, but the lexer still copied every
string token into its token buffer. That copy was 18 % of the
instructions of json::accept() on twitter.json (x86, GCC 16).

The lexer's flag for accept(), renamed from discard_number_values to
discard_values, now also selects a second instantiation of
scan_string() that performs the same checks (escapes, surrogates,
UTF-8, control characters, with the same error messages) without
writing to the token buffer. scan() chooses the instantiation once per
string: a run-time check inside the string loop made number-heavy
files up to 3 % slower on Apple M1.

json::accept() over 19 files, time after/before (geomean): 0.876 on x86
with GCC 16.2, 0.884 with Clang 23, 0.932 on Apple M1; up to 0.74 on
single files; files with few strings (canada, mesh, numbers) unchanged.

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

include/nlohmann/detail/input/lexer.hpp

@@ -268,10 +268,10 @@ class lexer : public lexer_base<BasicJsonType>
   public:
     using token_type = typename lexer_base<BasicJsonType>::token_type;
 
-    explicit lexer(InputAdapterType&& adapter, bool ignore_comments_ = false, bool discard_number_values_ = false) noexcept
+    explicit lexer(InputAdapterType&& adapter, bool ignore_comments_ = false, bool discard_values_ = false) noexcept
         : ia(std::move(adapter))
         , ignore_comments(ignore_comments_)
-        , discard_number_values(discard_number_values_)
+        , discard_values(discard_values_)
     {
         // the token string of an error before the first token is everything
         // read so far
@@ -388,6 +388,8 @@ class lexer : public lexer_base<BasicJsonType>
     checks if it is inside the range. If a violation was detected, set up an
     error message and return false. Otherwise, return true.
 
+    @param[in] keep    whether the bytes are added to token_buffer (false for
+                       accept(), see scan())
     @param[in] ranges  list of integers; interpreted as list of pairs of
                        inclusive lower and upper bound, respectively
 
@@ -396,17 +398,18 @@ class lexer : public lexer_base<BasicJsonType>
 
     @return true if and only if no range violation was detected
     */
-    bool next_byte_in_range(std::initializer_list<char_int_type> ranges)
+    template<bool Keep>
+    bool next_byte_in_range(std::integral_constant<bool, Keep> keep, std::initializer_list<char_int_type> ranges)
     {
         JSON_ASSERT(ranges.size() == 2 || ranges.size() == 4 || ranges.size() == 6);
-        add(current);
+        add_string_char(keep, current);
 
         for (auto range = ranges.begin(); range != ranges.end(); ++range)
         {
             get();
             if (JSON_HEDLEY_LIKELY(*range <= current && current <= *(++range))) // NOLINT(bugprone-inc-dec-in-conditions)
             {
-                add(current);
+                add_string_char(keep, current);
             }
             else
             {
@@ -419,14 +422,16 @@ class lexer : public lexer_base<BasicJsonType>
     }
 
     /// contiguous input: bulk-append the run of ordinary characters and complete
-    /// well-formed UTF-8 sequences starting at the current read position, then
+    /// well-formed UTF-8 sequences starting at the current read position (only
+    /// skip it if @a keep is false), then
     /// read the first byte that needs individual handling (the closing quote, an
     /// escape, a control character, or an ill-formed UTF-8 byte) like get().
     /// That byte is taken from the buffer that was just scanned. Through get(),
     /// Apple Clang merged this read with the other get() calls of the scan
     /// function, which load the read position again right after it was stored
     /// (3% of json::accept() on jeopardy.json on Apple M1).
-    char_int_type get_after_string_run(std::true_type /*bulk*/)
+    template<bool Keep>
+    char_int_type get_after_string_run(std::true_type /*bulk*/, std::integral_constant<bool, Keep> keep)
     {
         // a pending unget must be consumed through the normal path first
         if (next_unget)
@@ -444,7 +449,7 @@ class lexer : public lexer_base<BasicJsonType>
         const std::size_t pos = lexer_string_bulk_run(data, remaining);
         if (pos != 0)
         {
-            token_buffer.append(reinterpret_cast<const typename string_t::value_type*>(data), pos);
+            append_string_run(keep, data, pos);
             bulk_consume(reinterpret_cast<const char*>(data), pos);
         }
         if (pos == remaining)
@@ -488,7 +493,8 @@ class lexer : public lexer_base<BasicJsonType>
     }
 
     /// streaming input: no bulk fast path
-    char_int_type get_after_string_run(std::false_type /*bulk*/)
+    template<bool Keep>
+    char_int_type get_after_string_run(std::false_type /*bulk*/, std::integral_constant<bool, Keep> /*keep*/)
     {
         return get();
     }
@@ -502,13 +508,18 @@ class lexer : public lexer_base<BasicJsonType>
     may contain \0 bytes), and token_buffer.size() is the number of bytes in the
     string.
 
+    @param[in] keep  whether the string is copied into token_buffer; if false
+                     (accept(), see scan()), it is only checked, with the same
+                     errors, and token_buffer stays empty
+
     @return token_type::value_string if string could be successfully scanned,
             token_type::parse_error otherwise
 
     @note In case of errors, variable error_message contains a textual
           description.
     */
-    token_type scan_string()
+    template<bool Keep>
+    token_type scan_string(std::integral_constant<bool, Keep> keep)
     {
         // reset token_buffer (ignore opening quote)
         reset();
@@ -520,7 +531,7 @@ class lexer : public lexer_base<BasicJsonType>
         {
             // bulk-consume ordinary characters from contiguous input, then
             // get the next character, the one that ends the run
-            switch (get_after_string_run(std::integral_constant<bool, bulk_scan>()))
+            switch (get_after_string_run(std::integral_constant<bool, bulk_scan>(), keep))
             {
                 // end of file while parsing the string
                 case char_traits<char_type>::eof():
@@ -542,35 +553,35 @@ class lexer : public lexer_base<BasicJsonType>
                     {
                         // quotation mark
                         case '\"':
-                            add('\"');
+                            add_string_char(keep, '\"');
                             break;
                         // reverse solidus
                         case '\\':
-                            add('\\');
+                            add_string_char(keep, '\\');
                             break;
                         // solidus
                         case '/':
-                            add('/');
+                            add_string_char(keep, '/');
                             break;
                         // backspace
                         case 'b':
-                            add('\b');
+                            add_string_char(keep, '\b');
                             break;
                         // form feed
                         case 'f':
-                            add('\f');
+                            add_string_char(keep, '\f');
                             break;
                         // line feed
                         case 'n':
-                            add('\n');
+                            add_string_char(keep, '\n');
                             break;
                         // carriage return
                         case 'r':
-                            add('\r');
+                            add_string_char(keep, '\r');
                             break;
                         // tab
                         case 't':
-                            add('\t');
+                            add_string_char(keep, '\t');
                             break;
 
                         // unicode escapes
@@ -638,10 +649,7 @@ class lexer : public lexer_base<BasicJsonType>
                             JSON_ASSERT(0x00 <= codepoint && codepoint <= 0x10FFFF);
 
                             // translate codepoint into bytes
-                            encode_utf8(static_cast<std::uint32_t>(codepoint), [this](std::uint32_t byte)
+                            add_codepoint(keep, codepoint);
-                            {
-                                add(static_cast<char_int_type>(byte));
-                            });
 
                             break;
                         }
@@ -824,7 +832,7 @@ class lexer : public lexer_base<BasicJsonType>
                 case 0x7E:
                 case 0x7F:
                 {
-                    add(current);
+                    add_string_char(keep, current);
                     break;
                 }
 
@@ -860,7 +868,7 @@ class lexer : public lexer_base<BasicJsonType>
                 case 0xDE:
                 case 0xDF:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!next_byte_in_range({0x80, 0xBF})))
+                    if (JSON_HEDLEY_UNLIKELY(!next_byte_in_range(keep, {0x80, 0xBF})))
                     {
                         return token_type::parse_error;
                     }
@@ -870,7 +878,7 @@ class lexer : public lexer_base<BasicJsonType>
                 // U+0800..U+0FFF: bytes E0 A0..BF 80..BF
                 case 0xE0:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0xA0, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0xA0, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -894,7 +902,7 @@ class lexer : public lexer_base<BasicJsonType>
                 case 0xEE:
                 case 0xEF:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x80, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -904,7 +912,7 @@ class lexer : public lexer_base<BasicJsonType>
                 // U+D000..U+D7FF: bytes ED 80..9F 80..BF
                 case 0xED:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0x9F, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x80, 0x9F, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -914,7 +922,7 @@ class lexer : public lexer_base<BasicJsonType>
                 // U+10000..U+3FFFF F0 90..BF 80..BF 80..BF
                 case 0xF0:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x90, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x90, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -926,7 +934,7 @@ class lexer : public lexer_base<BasicJsonType>
                 case 0xF2:
                 case 0xF3:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x80, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -936,7 +944,7 @@ class lexer : public lexer_base<BasicJsonType>
                 // U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
                 case 0xF4:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0x8F, 0x80, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x80, 0x8F, 0x80, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -1427,7 +1435,7 @@ scan_number_done:
     token_type convert_number(token_type number_type, std::size_t mantissa_end)
     {
         // accept() only needs to know whether the input is valid, so it sets
-        // discard_number_values (see json.hpp), and an integer token whose
+        // discard_values (see json.hpp), and an integer token whose
         // digit count shows that it fits is reported without calling
         // convert_integer(). A number with up to 18 digits always fits into
         // both std::uint64_t and std::int64_t (18 nines is about 1e18, below
@@ -1442,7 +1450,7 @@ scan_number_done:
         // json_sax_acceptor, whose number callbacks discard their argument and
         // return true, and the parser rejects neither integers nor finite
         // floats. value_unsigned/value_integer are left unset here, so a caller
-        // that reads the converted value must not set discard_number_values.
+        // that reads the converted value must not set discard_values.
         //
         // On contiguous input, scan_number_bulk_contiguous() converts integer
         // tokens itself and does not pass them to this function, unless
@@ -1450,7 +1458,7 @@ scan_number_done:
         // reached for input without bulk access (e.g. streams), with
         // JSON_DIAGNOSTIC_POSITIONS, or when scan_number_bulk_contiguous()
         // falls back to scan_number().
-        if (discard_number_values)
+        if (discard_values)
         {
             constexpr std::size_t safe_digit_count = 18;
             if (number_type == token_type::value_unsigned && token_buffer.size() <= safe_digit_count)
@@ -1701,7 +1709,7 @@ scan_number_done:
         // w < 10^19, so a value below 10^max_exponent10 cannot overflow, and a
         // value that underflows is a finite number as well. Only a token that
         // may overflow has to be converted to find out.
-        if (JSON_HEDLEY_UNLIKELY(discard_number_values && s.exponent + 19 <= std::numeric_limits<number_float_t>::max_exponent10))
+        if (JSON_HEDLEY_UNLIKELY(discard_values && s.exponent + 19 <= std::numeric_limits<number_float_t>::max_exponent10))
         {
             value_float = 0;
             return true;
@@ -2033,6 +2041,34 @@ scan_number_done:
         token_buffer.push_back(static_cast<typename string_t::value_type>(c));
     }
 
+    /// add a character of a string token; accept() checks the string but does
+    /// not keep its value (@a keep is false)
+    void add_string_char(std::true_type /*keep*/, char_int_type c)
+    {
+        token_buffer.push_back(static_cast<typename string_t::value_type>(c));
+    }
+
+    static void add_string_char(std::false_type /*keep*/, char_int_type /*c*/) noexcept {}
+
+    /// add the UTF-8 bytes of a code point from a \\u escape to a string token
+    void add_codepoint(std::true_type /*keep*/, int codepoint)
+    {
+        encode_utf8(static_cast<std::uint32_t>(codepoint), [this](std::uint32_t byte)
+        {
+            add(static_cast<char_int_type>(byte));
+        });
+    }
+
+    static void add_codepoint(std::false_type /*keep*/, int /*codepoint*/) noexcept {}
+
+    /// append a run of ordinary string characters (see get_after_string_run())
+    void append_string_run(std::true_type /*keep*/, const unsigned char* data, std::size_t n)
+    {
+        token_buffer.append(reinterpret_cast<const typename string_t::value_type*>(data), n);
+    }
+
+    static void append_string_run(std::false_type /*keep*/, const unsigned char* /*data*/, std::size_t /*n*/) noexcept {}
+
   public:
     /////////////////////
     // value getters
@@ -2427,7 +2463,11 @@ scan_number_done:
 
             // string
             case '\"':
-                return scan_string();
+                // accept() (discard_values) checks strings without copying them
+                // into token_buffer. The choice is made once per string: a
+                // check inside the string loop made number-heavy documents up
+                // to 3 % slower (mesh.json, Apple M1)
+                return discard_values ? scan_string(std::false_type {}) : scan_string(std::true_type {});
 
             // number
             case '-':
@@ -2512,11 +2552,13 @@ scan_number_done:
     std::size_t decimal_point_position = std::string::npos;
 
     /// whether the caller only needs the token types and never looks at the
-    /// converted numeric values; set only by accept(), which parses through
+    /// values of number and string tokens; set only by accept(), which parses
-    /// json_sax_acceptor. When set, convert_number() skips converting integer
+    /// through json_sax_acceptor. When set, convert_number() skips converting
-    /// tokens whose digit count guarantees that they fit into 64 bits (see
+    /// integer tokens whose digit count guarantees that they fit into 64 bits
-    /// there)
+    /// (see there), floats are converted only when they may overflow (see
-    const bool discard_number_values = false;
+    /// convert_accumulated_float()), and strings are checked without being
+    /// copied into token_buffer (see scan())
+    const bool discard_values = false;
 };
 
 }  // namespace detail

include/nlohmann/detail/input/parser.hpp

@@ -89,9 +89,9 @@ class parser
                     const bool allow_exceptions_ = true,
                     const bool ignore_comments = false,
                     const bool ignore_trailing_commas_ = false,
-                    const bool discard_number_values_ = false)
+                    const bool discard_values_ = false)
         : callback(std::move(cb))
-        , m_lexer(std::move(adapter), ignore_comments, discard_number_values_)
+        , m_lexer(std::move(adapter), ignore_comments, discard_values_)
         , allow_exceptions(allow_exceptions_)
         , ignore_trailing_commas(ignore_trailing_commas_)
     {

include/nlohmann/detail/output/binary_writer.hpp

@@ -2088,8 +2088,7 @@ class binary_writer
         /// arrays: the first index with one more digit; only maintained by
         /// @ref calc_bson_sizes
         std::uint64_t index_boundary = 10;
-        /// @ref calc_bson_sizes: where its size goes in the table; @ref
+        /// @ref calc_bson_sizes only: where its size goes in the table
-        /// write_bson_document_single_pass: the position of its size field
         std::size_t size_slot;
         /// @ref calc_bson_sizes only: the size of its entries seen so far
         std::size_t entries_size = 0;
@@ -2340,197 +2339,6 @@ class binary_writer
         }
     }
 
-    template<typename SinkType>
-    using sink_written_t = decltype(std::declval<const SinkType&>().written());
-
-    /*!
-    @brief whether BSON documents can be written in a single pass: the sink
-           can overwrite what it has written (output_vector_sink), and every
-           integer fits the 64 bits BSON has, so that writing an array of
-           numbers needs no range check
-    */
-    using bson_single_pass_sink = std::integral_constant < bool,
-          is_detected<sink_written_t, OutputSinkType>::value
-          && std::numeric_limits<typename BasicJsonType::number_integer_t>::is_specialized
-          && std::numeric_limits<typename BasicJsonType::number_integer_t>::digits <= std::numeric_limits<std::int64_t>::digits
-          && std::numeric_limits<typename BasicJsonType::number_unsigned_t>::is_specialized
-          && std::numeric_limits<typename BasicJsonType::number_unsigned_t>::digits <= std::numeric_limits<std::uint64_t>::digits >;
-
-    /// write a placeholder for the size of an object or array whose entries
-    /// follow; @return its position, for @ref patch_bson_size
-    std::size_t write_bson_size_placeholder()
-    {
-        const std::size_t pos = oa.written();
-        write_number<std::int32_t>(0, true);
-        return pos;
-    }
-
-    /// end the object or array whose size is at @a pos and write its size
-    /// there: everything from that field to the terminating null byte
-    /// @return false (and the size is not written) if it does not fit
-    bool close_bson_container(const std::size_t pos)
-    {
-        oa.write_character(to_char_type(0x00));
-        const std::size_t size = oa.written() - pos;
-        if (JSON_HEDLEY_UNLIKELY(!value_in_range_of<std::int32_t>(size)))
-        {
-            return false;
-        }
-        std::array<CharType, sizeof(std::int32_t)> bytes{};
-        store_little_endian<std::int32_t>(bytes.data(), static_cast<std::int32_t>(size));
-        oa.overwrite(pos, bytes.data(), bytes.size());
-        return true;
-    }
-
-    /*!
-    @brief check the value @a j of a BSON entry as @ref calc_bson_value_size
-           does, for @ref write_bson_document_single_pass
-    @return at least the size of the value
-    A number, boolean, or null needs no check there (its integers have at most
-    64 bits) and has at most 8 characters; a discarded value is rejected with
-    the same exception when it is written.
-    */
-    std::size_t check_bson_value(const BasicJsonType& j)
-    {
-        return (j.is_string() || j.is_binary()) ? calc_bson_value_size(j) : 8;
-    }
-
-    /// whether an entry of @a entry_size characters keeps the document that
-    /// begins at @a start, closed by @a open_containers null bytes, within
-    /// the 32-bit size of BSON
-    bool bson_entry_fits(const std::size_t start, const std::size_t entry_size, const std::size_t open_containers) const noexcept
-    {
-        const std::size_t limit = static_cast<std::size_t>((std::numeric_limits<std::int32_t>::max)());
-        const std::size_t written = oa.written() - start;
-        return written <= limit && entry_size <= limit - written && open_containers <= limit - written - entry_size;
-    }
-
-    /*!
-    @brief Serializes the JSON object @a document as a BSON document in a
-           single pass, writing the size of each object and array once it is
-           complete
-
-    Checks every key and value as @ref calc_bson_sizes does, in the same
-    order, so the same exception is thrown; the output written up to then is
-    discarded with the vector it was written to.
-
-    @return false as soon as the document or a string, binary value, object,
-            or array in it would be larger than the 32-bit BSON sizes allow;
-            nothing large has been written then, and the caller writes the
-            document in two passes, which reports that error
-    */
-    template < typename SinkType = OutputSinkType,
-               typename std::enable_if < is_detected<sink_written_t, SinkType>::value, int >::type = 0 >
-    bool write_bson_document_single_pass(const BasicJsonType& document)
-    {
-        const std::size_t start = oa.written();
-        // the object or array whose entries are being written (with the
-        // position of its size), and the ones it is in
-        bson_frame current(&document, write_bson_size_placeholder());
-        std::vector<bson_frame> parents;
-
-        while (true)
-        {
-            const BasicJsonType* nested = nullptr;
-            std::size_t nested_size_pos = 0;
-            if (current.value->is_object())
-            {
-                const auto& object = *current.value->m_data.m_value.object;
-                while (nested == nullptr && current.member != object.cend())
-                {
-                    const auto& el = *current.member;
-                    ++current.member;
-                    const std::size_t header_size = calc_bson_entry_header_size(el.first, el.second);
-                    const std::size_t value_size = el.second.is_structured() ? 0 : check_bson_value(el.second);
-                    if (JSON_HEDLEY_UNLIKELY(!bson_entry_fits(start, header_size + value_size, parents.size() + 1)))
-                    {
-                        return false;
-                    }
-                    if (el.second.is_structured())
-                    {
-                        write_bson_entry_header(el.first, el.second.is_object() ? 0x03 : 0x04);
-                        nested_size_pos = write_bson_size_placeholder();
-                        nested = &el.second;
-                    }
-                    else
-                    {
-                        write_bson_value(el.first, el.second);
-                    }
-                }
-            }
-            else
-            {
-                const auto& array = *current.value->m_data.m_value.array;
-                while (nested == nullptr && current.index < array.size())
-                {
-                    if (is_bson_fixed_size(array[current.index]))
-                    {
-                        write_bson_array_fixed_entries(current);
-                        if (current.index == array.size())
-                        {
-                            break;
-                        }
-                    }
-
-                    // a string, binary value, object or array
-                    const BasicJsonType& el = array[current.index];
-                    const string_t index_name(current.index_name.data(), current.index_name_size);
-                    current.next_index();
-                    const std::size_t value_size = el.is_structured() ? 0 : check_bson_value(el);
-                    if (JSON_HEDLEY_UNLIKELY(!bson_entry_fits(start, 2 + index_name.size() + value_size, parents.size() + 1)))
-                    {
-                        return false;
-                    }
-                    if (el.is_structured())
-                    {
-                        write_bson_entry_header(index_name, el.is_object() ? 0x03 : 0x04);
-                        nested_size_pos = write_bson_size_placeholder();
-                        nested = &el;
-                    }
-                    else
-                    {
-                        write_bson_value(index_name, el);
-                    }
-                }
-            }
-
-            if (nested != nullptr)
-            {
-                if (nested->is_array())
-                {
-                    // an array of booleans, numbers and null is written without a frame on the stack
-                    bson_frame leaf(nested, nested_size_pos);
-                    write_bson_array_fixed_entries(leaf);
-                    if (leaf.index == nested->m_data.m_value.array->size())
-                    {
-                        if (JSON_HEDLEY_UNLIKELY(!close_bson_container(nested_size_pos)))
-                        {
-                            return false;
-                        }
-                        continue;
-                    }
-                    parents.push_back(std::move(current));
-                    current = std::move(leaf);
-                    continue;
-                }
-                parents.push_back(std::move(current));
-                current = bson_frame(nested, nested_size_pos);
-                continue;
-            }
-
-            if (JSON_HEDLEY_UNLIKELY(!close_bson_container(current.size_slot)))
-            {
-                return false;
-            }
-            if (parents.empty())
-            {
-                return true;
-            }
-            current = std::move(parents.back());
-            parents.pop_back();
-        }
-    }
-
     /*!
     @brief Serializes the JSON object @a document as a BSON document
 
@@ -2542,39 +2350,6 @@ class binary_writer
     @pre       document.type() == value_t::object
     */
     void write_bson_document(const BasicJsonType& document)
-    {
-        write_bson_document(document, bson_single_pass_sink());
-    }
-
-    /// the sink can overwrite: one pass, unless the error handler replaces or
-    /// drops ill-formed UTF-8, which changes the sizes the checks compute
-    void write_bson_document(const BasicJsonType& document, std::true_type /*single_pass*/)
-    {
-        const std::size_t start = oa.written();
-        if ((error_handler == error_handler_t::keep || error_handler == error_handler_t::strict)
-                && write_bson_document_single_pass(document))
-        {
-            return;
-        }
-        // a length does not fit: start over, so that it is reported as the
-        // two passes report it, before anything is written
-        oa.rewind(start);
-        write_bson_document_two_pass(document);
-    }
-
-    void write_bson_document(const BasicJsonType& document, std::false_type /*single_pass*/)
-    {
-        write_bson_document_two_pass(document);
-    }
-
-    /*!
-    @brief Serializes the JSON object @a document as a BSON document: the
-           sizes of all objects and arrays first, then the entries
-
-    @param[in] document  the JSON object to serialize
-    @pre       document.type() == value_t::object
-    */
-    void write_bson_document_two_pass(const BasicJsonType& document)
     {
         std::vector<std::size_t> nested_sizes;
         const std::size_t document_size = calc_bson_sizes(document, nested_sizes);

include/nlohmann/detail/output/output_adapters.hpp

@@ -227,29 +227,6 @@ class output_vector_sink
     void flush() noexcept
     {}
 
-    /// the number of characters written so far, the position of the next one
-    std::size_t written() const noexcept
-    {
-        return position();
-    }
-
-    /// @brief replace @a length characters written before, from position @a pos
-    ///
-    /// Lets a writer put a length in front of data it has written since, such
-    /// as the size of a BSON document, without computing it in advance.
-    void overwrite(const std::size_t pos, const CharType* s, const std::size_t length) noexcept
-    {
-        JSON_ASSERT(pos + length <= position());
-        std::memcpy(v->data() + pos, s, length * sizeof(CharType));
-    }
-
-    /// drop the characters written from position @a pos on
-    void rewind(const std::size_t pos) noexcept
-    {
-        JSON_ASSERT(pos <= position());
-        cursor = v->data() + pos;
-    }
-
   private:
     /// number of characters of *v written so far
     std::size_t position() const noexcept

include/nlohmann/json.hpp

@@ -184,11 +184,11 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         const bool allow_exceptions = true,
         const bool ignore_comments = false,
         const bool ignore_trailing_commas = false,
-        const bool discard_number_values = false
+        const bool discard_values = false
     )
     {
         return ::nlohmann::detail::parser<basic_json, InputAdapterType>(std::move(adapter),
-               std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas, discard_number_values);
+               std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas, discard_values);
     }
 
   private:

single_include/nlohmann/json.hpp

@@ -13124,10 +13124,10 @@ class lexer : public lexer_base<BasicJsonType>
   public:
     using token_type = typename lexer_base<BasicJsonType>::token_type;
 
-    explicit lexer(InputAdapterType&& adapter, bool ignore_comments_ = false, bool discard_number_values_ = false) noexcept
+    explicit lexer(InputAdapterType&& adapter, bool ignore_comments_ = false, bool discard_values_ = false) noexcept
         : ia(std::move(adapter))
         , ignore_comments(ignore_comments_)
-        , discard_number_values(discard_number_values_)
+        , discard_values(discard_values_)
     {
         // the token string of an error before the first token is everything
         // read so far
@@ -13244,6 +13244,8 @@ class lexer : public lexer_base<BasicJsonType>
     checks if it is inside the range. If a violation was detected, set up an
     error message and return false. Otherwise, return true.
 
+    @param[in] keep    whether the bytes are added to token_buffer (false for
+                       accept(), see scan())
     @param[in] ranges  list of integers; interpreted as list of pairs of
                        inclusive lower and upper bound, respectively
 
@@ -13252,17 +13254,18 @@ class lexer : public lexer_base<BasicJsonType>
 
     @return true if and only if no range violation was detected
     */
-    bool next_byte_in_range(std::initializer_list<char_int_type> ranges)
+    template<bool Keep>
+    bool next_byte_in_range(std::integral_constant<bool, Keep> keep, std::initializer_list<char_int_type> ranges)
     {
         JSON_ASSERT(ranges.size() == 2 || ranges.size() == 4 || ranges.size() == 6);
-        add(current);
+        add_string_char(keep, current);
 
         for (auto range = ranges.begin(); range != ranges.end(); ++range)
         {
             get();
             if (JSON_HEDLEY_LIKELY(*range <= current && current <= *(++range))) // NOLINT(bugprone-inc-dec-in-conditions)
             {
-                add(current);
+                add_string_char(keep, current);
             }
             else
             {
@@ -13275,14 +13278,16 @@ class lexer : public lexer_base<BasicJsonType>
     }
 
     /// contiguous input: bulk-append the run of ordinary characters and complete
-    /// well-formed UTF-8 sequences starting at the current read position, then
+    /// well-formed UTF-8 sequences starting at the current read position (only
+    /// skip it if @a keep is false), then
     /// read the first byte that needs individual handling (the closing quote, an
     /// escape, a control character, or an ill-formed UTF-8 byte) like get().
     /// That byte is taken from the buffer that was just scanned. Through get(),
     /// Apple Clang merged this read with the other get() calls of the scan
     /// function, which load the read position again right after it was stored
     /// (3% of json::accept() on jeopardy.json on Apple M1).
-    char_int_type get_after_string_run(std::true_type /*bulk*/)
+    template<bool Keep>
+    char_int_type get_after_string_run(std::true_type /*bulk*/, std::integral_constant<bool, Keep> keep)
     {
         // a pending unget must be consumed through the normal path first
         if (next_unget)
@@ -13300,7 +13305,7 @@ class lexer : public lexer_base<BasicJsonType>
         const std::size_t pos = lexer_string_bulk_run(data, remaining);
         if (pos != 0)
         {
-            token_buffer.append(reinterpret_cast<const typename string_t::value_type*>(data), pos);
+            append_string_run(keep, data, pos);
             bulk_consume(reinterpret_cast<const char*>(data), pos);
         }
         if (pos == remaining)
@@ -13344,7 +13349,8 @@ class lexer : public lexer_base<BasicJsonType>
     }
 
     /// streaming input: no bulk fast path
-    char_int_type get_after_string_run(std::false_type /*bulk*/)
+    template<bool Keep>
+    char_int_type get_after_string_run(std::false_type /*bulk*/, std::integral_constant<bool, Keep> /*keep*/)
     {
         return get();
     }
@@ -13358,13 +13364,18 @@ class lexer : public lexer_base<BasicJsonType>
     may contain \0 bytes), and token_buffer.size() is the number of bytes in the
     string.
 
+    @param[in] keep  whether the string is copied into token_buffer; if false
+                     (accept(), see scan()), it is only checked, with the same
+                     errors, and token_buffer stays empty
+
     @return token_type::value_string if string could be successfully scanned,
             token_type::parse_error otherwise
 
     @note In case of errors, variable error_message contains a textual
           description.
     */
-    token_type scan_string()
+    template<bool Keep>
+    token_type scan_string(std::integral_constant<bool, Keep> keep)
     {
         // reset token_buffer (ignore opening quote)
         reset();
@@ -13376,7 +13387,7 @@ class lexer : public lexer_base<BasicJsonType>
         {
             // bulk-consume ordinary characters from contiguous input, then
             // get the next character, the one that ends the run
-            switch (get_after_string_run(std::integral_constant<bool, bulk_scan>()))
+            switch (get_after_string_run(std::integral_constant<bool, bulk_scan>(), keep))
             {
                 // end of file while parsing the string
                 case char_traits<char_type>::eof():
@@ -13398,35 +13409,35 @@ class lexer : public lexer_base<BasicJsonType>
                     {
                         // quotation mark
                         case '\"':
-                            add('\"');
+                            add_string_char(keep, '\"');
                             break;
                         // reverse solidus
                         case '\\':
-                            add('\\');
+                            add_string_char(keep, '\\');
                             break;
                         // solidus
                         case '/':
-                            add('/');
+                            add_string_char(keep, '/');
                             break;
                         // backspace
                         case 'b':
-                            add('\b');
+                            add_string_char(keep, '\b');
                             break;
                         // form feed
                         case 'f':
-                            add('\f');
+                            add_string_char(keep, '\f');
                             break;
                         // line feed
                         case 'n':
-                            add('\n');
+                            add_string_char(keep, '\n');
                             break;
                         // carriage return
                         case 'r':
-                            add('\r');
+                            add_string_char(keep, '\r');
                             break;
                         // tab
                         case 't':
-                            add('\t');
+                            add_string_char(keep, '\t');
                             break;
 
                         // unicode escapes
@@ -13494,10 +13505,7 @@ class lexer : public lexer_base<BasicJsonType>
                             JSON_ASSERT(0x00 <= codepoint && codepoint <= 0x10FFFF);
 
                             // translate codepoint into bytes
-                            encode_utf8(static_cast<std::uint32_t>(codepoint), [this](std::uint32_t byte)
+                            add_codepoint(keep, codepoint);
-                            {
-                                add(static_cast<char_int_type>(byte));
-                            });
 
                             break;
                         }
@@ -13680,7 +13688,7 @@ class lexer : public lexer_base<BasicJsonType>
                 case 0x7E:
                 case 0x7F:
                 {
-                    add(current);
+                    add_string_char(keep, current);
                     break;
                 }
 
@@ -13716,7 +13724,7 @@ class lexer : public lexer_base<BasicJsonType>
                 case 0xDE:
                 case 0xDF:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!next_byte_in_range({0x80, 0xBF})))
+                    if (JSON_HEDLEY_UNLIKELY(!next_byte_in_range(keep, {0x80, 0xBF})))
                     {
                         return token_type::parse_error;
                     }
@@ -13726,7 +13734,7 @@ class lexer : public lexer_base<BasicJsonType>
                 // U+0800..U+0FFF: bytes E0 A0..BF 80..BF
                 case 0xE0:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0xA0, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0xA0, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -13750,7 +13758,7 @@ class lexer : public lexer_base<BasicJsonType>
                 case 0xEE:
                 case 0xEF:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x80, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -13760,7 +13768,7 @@ class lexer : public lexer_base<BasicJsonType>
                 // U+D000..U+D7FF: bytes ED 80..9F 80..BF
                 case 0xED:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0x9F, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x80, 0x9F, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -13770,7 +13778,7 @@ class lexer : public lexer_base<BasicJsonType>
                 // U+10000..U+3FFFF F0 90..BF 80..BF 80..BF
                 case 0xF0:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x90, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x90, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -13782,7 +13790,7 @@ class lexer : public lexer_base<BasicJsonType>
                 case 0xF2:
                 case 0xF3:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x80, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -13792,7 +13800,7 @@ class lexer : public lexer_base<BasicJsonType>
                 // U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
                 case 0xF4:
                 {
-                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0x8F, 0x80, 0xBF, 0x80, 0xBF}))))
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range(keep, {0x80, 0x8F, 0x80, 0xBF, 0x80, 0xBF}))))
                     {
                         return token_type::parse_error;
                     }
@@ -14283,7 +14291,7 @@ scan_number_done:
     token_type convert_number(token_type number_type, std::size_t mantissa_end)
     {
         // accept() only needs to know whether the input is valid, so it sets
-        // discard_number_values (see json.hpp), and an integer token whose
+        // discard_values (see json.hpp), and an integer token whose
         // digit count shows that it fits is reported without calling
         // convert_integer(). A number with up to 18 digits always fits into
         // both std::uint64_t and std::int64_t (18 nines is about 1e18, below
@@ -14298,7 +14306,7 @@ scan_number_done:
         // json_sax_acceptor, whose number callbacks discard their argument and
         // return true, and the parser rejects neither integers nor finite
         // floats. value_unsigned/value_integer are left unset here, so a caller
-        // that reads the converted value must not set discard_number_values.
+        // that reads the converted value must not set discard_values.
         //
         // On contiguous input, scan_number_bulk_contiguous() converts integer
         // tokens itself and does not pass them to this function, unless
@@ -14306,7 +14314,7 @@ scan_number_done:
         // reached for input without bulk access (e.g. streams), with
         // JSON_DIAGNOSTIC_POSITIONS, or when scan_number_bulk_contiguous()
         // falls back to scan_number().
-        if (discard_number_values)
+        if (discard_values)
         {
             constexpr std::size_t safe_digit_count = 18;
             if (number_type == token_type::value_unsigned && token_buffer.size() <= safe_digit_count)
@@ -14557,7 +14565,7 @@ scan_number_done:
         // w < 10^19, so a value below 10^max_exponent10 cannot overflow, and a
         // value that underflows is a finite number as well. Only a token that
         // may overflow has to be converted to find out.
-        if (JSON_HEDLEY_UNLIKELY(discard_number_values && s.exponent + 19 <= std::numeric_limits<number_float_t>::max_exponent10))
+        if (JSON_HEDLEY_UNLIKELY(discard_values && s.exponent + 19 <= std::numeric_limits<number_float_t>::max_exponent10))
         {
             value_float = 0;
             return true;
@@ -14889,6 +14897,34 @@ scan_number_done:
         token_buffer.push_back(static_cast<typename string_t::value_type>(c));
     }
 
+    /// add a character of a string token; accept() checks the string but does
+    /// not keep its value (@a keep is false)
+    void add_string_char(std::true_type /*keep*/, char_int_type c)
+    {
+        token_buffer.push_back(static_cast<typename string_t::value_type>(c));
+    }
+
+    static void add_string_char(std::false_type /*keep*/, char_int_type /*c*/) noexcept {}
+
+    /// add the UTF-8 bytes of a code point from a \\u escape to a string token
+    void add_codepoint(std::true_type /*keep*/, int codepoint)
+    {
+        encode_utf8(static_cast<std::uint32_t>(codepoint), [this](std::uint32_t byte)
+        {
+            add(static_cast<char_int_type>(byte));
+        });
+    }
+
+    static void add_codepoint(std::false_type /*keep*/, int /*codepoint*/) noexcept {}
+
+    /// append a run of ordinary string characters (see get_after_string_run())
+    void append_string_run(std::true_type /*keep*/, const unsigned char* data, std::size_t n)
+    {
+        token_buffer.append(reinterpret_cast<const typename string_t::value_type*>(data), n);
+    }
+
+    static void append_string_run(std::false_type /*keep*/, const unsigned char* /*data*/, std::size_t /*n*/) noexcept {}
+
   public:
     /////////////////////
     // value getters
@@ -15283,7 +15319,11 @@ scan_number_done:
 
             // string
             case '\"':
-                return scan_string();
+                // accept() (discard_values) checks strings without copying them
+                // into token_buffer. The choice is made once per string: a
+                // check inside the string loop made number-heavy documents up
+                // to 3 % slower (mesh.json, Apple M1)
+                return discard_values ? scan_string(std::false_type {}) : scan_string(std::true_type {});
 
             // number
             case '-':
@@ -15368,11 +15408,13 @@ scan_number_done:
     std::size_t decimal_point_position = std::string::npos;
 
     /// whether the caller only needs the token types and never looks at the
-    /// converted numeric values; set only by accept(), which parses through
+    /// values of number and string tokens; set only by accept(), which parses
-    /// json_sax_acceptor. When set, convert_number() skips converting integer
+    /// through json_sax_acceptor. When set, convert_number() skips converting
-    /// tokens whose digit count guarantees that they fit into 64 bits (see
+    /// integer tokens whose digit count guarantees that they fit into 64 bits
-    /// there)
+    /// (see there), floats are converted only when they may overflow (see
-    const bool discard_number_values = false;
+    /// convert_accumulated_float()), and strings are checked without being
+    /// copied into token_buffer (see scan())
+    const bool discard_values = false;
 };
 
 }  // namespace detail
@@ -22581,9 +22623,9 @@ class parser
                     const bool allow_exceptions_ = true,
                     const bool ignore_comments = false,
                     const bool ignore_trailing_commas_ = false,
-                    const bool discard_number_values_ = false)
+                    const bool discard_values_ = false)
         : callback(std::move(cb))
-        , m_lexer(std::move(adapter), ignore_comments, discard_number_values_)
+        , m_lexer(std::move(adapter), ignore_comments, discard_values_)
         , allow_exceptions(allow_exceptions_)
         , ignore_trailing_commas(ignore_trailing_commas_)
     {
@@ -26870,29 +26912,6 @@ class output_vector_sink
     void flush() noexcept
     {}
 
-    /// the number of characters written so far, the position of the next one
-    std::size_t written() const noexcept
-    {
-        return position();
-    }
-
-    /// @brief replace @a length characters written before, from position @a pos
-    ///
-    /// Lets a writer put a length in front of data it has written since, such
-    /// as the size of a BSON document, without computing it in advance.
-    void overwrite(const std::size_t pos, const CharType* s, const std::size_t length) noexcept
-    {
-        JSON_ASSERT(pos + length <= position());
-        std::memcpy(v->data() + pos, s, length * sizeof(CharType));
-    }
-
-    /// drop the characters written from position @a pos on
-    void rewind(const std::size_t pos) noexcept
-    {
-        JSON_ASSERT(pos <= position());
-        cursor = v->data() + pos;
-    }
-
   private:
     /// number of characters of *v written so far
     std::size_t position() const noexcept
@@ -29337,8 +29356,7 @@ class binary_writer
         /// arrays: the first index with one more digit; only maintained by
         /// @ref calc_bson_sizes
         std::uint64_t index_boundary = 10;
-        /// @ref calc_bson_sizes: where its size goes in the table; @ref
+        /// @ref calc_bson_sizes only: where its size goes in the table
-        /// write_bson_document_single_pass: the position of its size field
         std::size_t size_slot;
         /// @ref calc_bson_sizes only: the size of its entries seen so far
         std::size_t entries_size = 0;
@@ -29589,197 +29607,6 @@ class binary_writer
         }
     }
 
-    template<typename SinkType>
-    using sink_written_t = decltype(std::declval<const SinkType&>().written());
-
-    /*!
-    @brief whether BSON documents can be written in a single pass: the sink
-           can overwrite what it has written (output_vector_sink), and every
-           integer fits the 64 bits BSON has, so that writing an array of
-           numbers needs no range check
-    */
-    using bson_single_pass_sink = std::integral_constant < bool,
-          is_detected<sink_written_t, OutputSinkType>::value
-          && std::numeric_limits<typename BasicJsonType::number_integer_t>::is_specialized
-          && std::numeric_limits<typename BasicJsonType::number_integer_t>::digits <= std::numeric_limits<std::int64_t>::digits
-          && std::numeric_limits<typename BasicJsonType::number_unsigned_t>::is_specialized
-          && std::numeric_limits<typename BasicJsonType::number_unsigned_t>::digits <= std::numeric_limits<std::uint64_t>::digits >;
-
-    /// write a placeholder for the size of an object or array whose entries
-    /// follow; @return its position, for @ref patch_bson_size
-    std::size_t write_bson_size_placeholder()
-    {
-        const std::size_t pos = oa.written();
-        write_number<std::int32_t>(0, true);
-        return pos;
-    }
-
-    /// end the object or array whose size is at @a pos and write its size
-    /// there: everything from that field to the terminating null byte
-    /// @return false (and the size is not written) if it does not fit
-    bool close_bson_container(const std::size_t pos)
-    {
-        oa.write_character(to_char_type(0x00));
-        const std::size_t size = oa.written() - pos;
-        if (JSON_HEDLEY_UNLIKELY(!value_in_range_of<std::int32_t>(size)))
-        {
-            return false;
-        }
-        std::array<CharType, sizeof(std::int32_t)> bytes{};
-        store_little_endian<std::int32_t>(bytes.data(), static_cast<std::int32_t>(size));
-        oa.overwrite(pos, bytes.data(), bytes.size());
-        return true;
-    }
-
-    /*!
-    @brief check the value @a j of a BSON entry as @ref calc_bson_value_size
-           does, for @ref write_bson_document_single_pass
-    @return at least the size of the value
-    A number, boolean, or null needs no check there (its integers have at most
-    64 bits) and has at most 8 characters; a discarded value is rejected with
-    the same exception when it is written.
-    */
-    std::size_t check_bson_value(const BasicJsonType& j)
-    {
-        return (j.is_string() || j.is_binary()) ? calc_bson_value_size(j) : 8;
-    }
-
-    /// whether an entry of @a entry_size characters keeps the document that
-    /// begins at @a start, closed by @a open_containers null bytes, within
-    /// the 32-bit size of BSON
-    bool bson_entry_fits(const std::size_t start, const std::size_t entry_size, const std::size_t open_containers) const noexcept
-    {
-        const std::size_t limit = static_cast<std::size_t>((std::numeric_limits<std::int32_t>::max)());
-        const std::size_t written = oa.written() - start;
-        return written <= limit && entry_size <= limit - written && open_containers <= limit - written - entry_size;
-    }
-
-    /*!
-    @brief Serializes the JSON object @a document as a BSON document in a
-           single pass, writing the size of each object and array once it is
-           complete
-
-    Checks every key and value as @ref calc_bson_sizes does, in the same
-    order, so the same exception is thrown; the output written up to then is
-    discarded with the vector it was written to.
-
-    @return false as soon as the document or a string, binary value, object,
-            or array in it would be larger than the 32-bit BSON sizes allow;
-            nothing large has been written then, and the caller writes the
-            document in two passes, which reports that error
-    */
-    template < typename SinkType = OutputSinkType,
-               typename std::enable_if < is_detected<sink_written_t, SinkType>::value, int >::type = 0 >
-    bool write_bson_document_single_pass(const BasicJsonType& document)
-    {
-        const std::size_t start = oa.written();
-        // the object or array whose entries are being written (with the
-        // position of its size), and the ones it is in
-        bson_frame current(&document, write_bson_size_placeholder());
-        std::vector<bson_frame> parents;
-
-        while (true)
-        {
-            const BasicJsonType* nested = nullptr;
-            std::size_t nested_size_pos = 0;
-            if (current.value->is_object())
-            {
-                const auto& object = *current.value->m_data.m_value.object;
-                while (nested == nullptr && current.member != object.cend())
-                {
-                    const auto& el = *current.member;
-                    ++current.member;
-                    const std::size_t header_size = calc_bson_entry_header_size(el.first, el.second);
-                    const std::size_t value_size = el.second.is_structured() ? 0 : check_bson_value(el.second);
-                    if (JSON_HEDLEY_UNLIKELY(!bson_entry_fits(start, header_size + value_size, parents.size() + 1)))
-                    {
-                        return false;
-                    }
-                    if (el.second.is_structured())
-                    {
-                        write_bson_entry_header(el.first, el.second.is_object() ? 0x03 : 0x04);
-                        nested_size_pos = write_bson_size_placeholder();
-                        nested = &el.second;
-                    }
-                    else
-                    {
-                        write_bson_value(el.first, el.second);
-                    }
-                }
-            }
-            else
-            {
-                const auto& array = *current.value->m_data.m_value.array;
-                while (nested == nullptr && current.index < array.size())
-                {
-                    if (is_bson_fixed_size(array[current.index]))
-                    {
-                        write_bson_array_fixed_entries(current);
-                        if (current.index == array.size())
-                        {
-                            break;
-                        }
-                    }
-
-                    // a string, binary value, object or array
-                    const BasicJsonType& el = array[current.index];
-                    const string_t index_name(current.index_name.data(), current.index_name_size);
-                    current.next_index();
-                    const std::size_t value_size = el.is_structured() ? 0 : check_bson_value(el);
-                    if (JSON_HEDLEY_UNLIKELY(!bson_entry_fits(start, 2 + index_name.size() + value_size, parents.size() + 1)))
-                    {
-                        return false;
-                    }
-                    if (el.is_structured())
-                    {
-                        write_bson_entry_header(index_name, el.is_object() ? 0x03 : 0x04);
-                        nested_size_pos = write_bson_size_placeholder();
-                        nested = &el;
-                    }
-                    else
-                    {
-                        write_bson_value(index_name, el);
-                    }
-                }
-            }
-
-            if (nested != nullptr)
-            {
-                if (nested->is_array())
-                {
-                    // an array of booleans, numbers and null is written without a frame on the stack
-                    bson_frame leaf(nested, nested_size_pos);
-                    write_bson_array_fixed_entries(leaf);
-                    if (leaf.index == nested->m_data.m_value.array->size())
-                    {
-                        if (JSON_HEDLEY_UNLIKELY(!close_bson_container(nested_size_pos)))
-                        {
-                            return false;
-                        }
-                        continue;
-                    }
-                    parents.push_back(std::move(current));
-                    current = std::move(leaf);
-                    continue;
-                }
-                parents.push_back(std::move(current));
-                current = bson_frame(nested, nested_size_pos);
-                continue;
-            }
-
-            if (JSON_HEDLEY_UNLIKELY(!close_bson_container(current.size_slot)))
-            {
-                return false;
-            }
-            if (parents.empty())
-            {
-                return true;
-            }
-            current = std::move(parents.back());
-            parents.pop_back();
-        }
-    }
-
     /*!
     @brief Serializes the JSON object @a document as a BSON document
 
@@ -29791,39 +29618,6 @@ class binary_writer
     @pre       document.type() == value_t::object
     */
     void write_bson_document(const BasicJsonType& document)
-    {
-        write_bson_document(document, bson_single_pass_sink());
-    }
-
-    /// the sink can overwrite: one pass, unless the error handler replaces or
-    /// drops ill-formed UTF-8, which changes the sizes the checks compute
-    void write_bson_document(const BasicJsonType& document, std::true_type /*single_pass*/)
-    {
-        const std::size_t start = oa.written();
-        if ((error_handler == error_handler_t::keep || error_handler == error_handler_t::strict)
-                && write_bson_document_single_pass(document))
-        {
-            return;
-        }
-        // a length does not fit: start over, so that it is reported as the
-        // two passes report it, before anything is written
-        oa.rewind(start);
-        write_bson_document_two_pass(document);
-    }
-
-    void write_bson_document(const BasicJsonType& document, std::false_type /*single_pass*/)
-    {
-        write_bson_document_two_pass(document);
-    }
-
-    /*!
-    @brief Serializes the JSON object @a document as a BSON document: the
-           sizes of all objects and arrays first, then the entries
-
-    @param[in] document  the JSON object to serialize
-    @pre       document.type() == value_t::object
-    */
-    void write_bson_document_two_pass(const BasicJsonType& document)
     {
         std::vector<std::size_t> nested_sizes;
         const std::size_t document_size = calc_bson_sizes(document, nested_sizes);
@@ -37228,11 +37022,11 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         const bool allow_exceptions = true,
         const bool ignore_comments = false,
         const bool ignore_trailing_commas = false,
-        const bool discard_number_values = false
+        const bool discard_values = false
                                  )
     {
         return ::nlohmann::detail::parser<basic_json, InputAdapterType>(std::move(adapter),
-            std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas, discard_number_values);
+            std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas, discard_values);
     }
 
   private:

tests/src/unit-bson.cpp

@@ -2267,132 +2267,3 @@ TEST_CASE("BSON large strings and binaries (chunked reader)")
         }
     }
 }
-
-TEST_CASE("BSON written in one pass and in two passes")
-{
-    // to_bson() into a returned vector writes each size once the object or
-    // array is complete; through an output adapter, all sizes are computed
-    // first. Both must produce the same bytes and throw the same exceptions.
-    const auto two_pass = [](const json & j, const json::error_handler_t handler)
-    {
-        std::vector<std::uint8_t> result;
-        json::to_bson(j, result, handler);
-        return result;
-    };
-
-    const auto check_same = [&](const json & j, const json::error_handler_t handler)
-    {
-        CAPTURE(j.dump(-1, ' ', true, json::error_handler_t::replace))
-        CAPTURE(static_cast<int>(handler))
-#if !defined(JSON_NOEXCEPTION)
-        std::string one_pass_error;
-        std::string two_pass_error;
-        std::vector<std::uint8_t> one;
-        std::vector<std::uint8_t> two;
-        try
-        {
-            one = json::to_bson(j, handler);
-        }
-        catch (const json::exception& e)
-        {
-            one_pass_error = e.what();
-        }
-        try
-        {
-            two = two_pass(j, handler);
-        }
-        catch (const json::exception& e)
-        {
-            two_pass_error = e.what();
-        }
-        CHECK(one_pass_error == two_pass_error);
-        CHECK(one == two);
-#else
-        CHECK(json::to_bson(j, handler) == two_pass(j, handler));
-#endif
-    };
-
-    const json::error_handler_t handlers[] = {json::error_handler_t::keep, json::error_handler_t::strict,
-                                              json::error_handler_t::replace, json::error_handler_t::ignore
-                                             };
-
-    SECTION("corpus")
-    {
-        for (const auto& value : utils::round_trip_corpus::values())
-        {
-            const json document = value.is_object() ? value : json{{"value", value}};
-            for (const auto handler : handlers)
-            {
-                check_same(document, handler);
-            }
-        }
-    }
-
-    SECTION("nesting and arrays of every length class")
-    {
-        json document = json::object();
-        document["empty object"] = json::object();
-        document["empty array"] = json::array();
-        for (const std::size_t n :
-                {
-                    std::size_t(1), std::size_t(9), std::size_t(10), std::size_t(11), std::size_t(99), std::size_t(100), std::size_t(101), std::size_t(1000)
-                })
-        {
-            json numbers = json::array();
-            json mixed = json::array();
-            for (std::size_t i = 0; i < n; ++i)
-            {
-                numbers.push_back(i % 3 == 0 ? json(static_cast<double>(i) / 3) : json(static_cast<std::int64_t>(i) * 1000003));
-                mixed.push_back(i % 4 == 0 ? json("s" + std::to_string(i)) : i % 4 == 1 ? json::array({i, json::object({{"k", i}})}) : i % 4 == 2 ? json(nullptr) : json(i % 2 == 0));
-            }
-            document["numbers " + std::to_string(n)] = numbers;
-            document["mixed " + std::to_string(n)] = mixed;
-            document["nested " + std::to_string(n)] = json::object({{"a", numbers}, {"b", json::array({mixed, json::object()})}});
-        }
-        document["binary"] = json::binary({1, 2, 3}, 0x80);
-        document["binary old"] = json::binary({1, 2, 3}, 0x02);
-        for (const auto handler : handlers)
-        {
-            check_same(document, handler);
-        }
-    }
-
-    SECTION("errors at every position")
-    {
-        // an entry that cannot be written, at each place among valid ones
-        const std::string bad_utf8 = "\xC3\x28";
-        const json bad_entries[] =
-        {
-            json::object({{std::string("a\0b", 3), 1}}),
-            json::object({{"s", bad_utf8}}),
-            json::object({{bad_utf8, 1}}),
-            json::array({1, 2, bad_utf8}),
-            json::object({{"b", json::binary({1}, 0x100)}}),
-            json::object({{"d", json(json::value_t::discarded)}}),
-            json::array({json(json::value_t::discarded)}),
-        };
-        for (const auto& bad : bad_entries)
-        {
-            for (std::size_t pos = 0; pos < 4; ++pos)
-            {
-                json document = json::object();
-                json list = json::array();
-                for (std::size_t i = 0; i < 4; ++i)
-                {
-                    if (i == pos)
-                    {
-                        document["x" + std::to_string(i)] = bad;
-                        list.push_back(bad);
-                    }
-                    document["n" + std::to_string(i)] = json::array({i, "str", json::object({{"k", i}})});
-                    list.push_back(i);
-                }
-                document["list"] = list;
-                for (const auto handler : handlers)
-                {
-                    check_same(document, handler);
-                }
-            }
-        }
-    }
-}

tests/src/unit-class_parser.cpp

@@ -2982,6 +2982,50 @@ TEST_CASE("accept() agrees with parse() on floats around the overflow limit")
     }
 }
 
+TEST_CASE("accept() agrees with parse() on strings, which it checks without copying them")
+{
+    // accept() scans strings and object keys with a separate instantiation of
+    // the string scanner that does not copy them into the token buffer; it
+    // must accept and reject exactly what parse() does
+    const std::vector<std::string> pieces =
+    {
+        "", "abc", "\\\"\\\\\\/\\b\\f\\n\\r\\t", "\\u0041\\u00e4\\u20AC", "\\ud83d\\ude00", "\\uD83D\\uDE00",
+        "\\ud800", "\\udc00", "\\ud800\\u0041", "\\ud800x", "\\udbff\\udfff", "\\u12", "\\u12g4", "\\x", "\\",
+        "\xc3\xa4", "\xe2\x82\xac", "\xf0\x9f\x98\x80", "\xf4\x8f\xbf\xbf", "\xc0\xaf", "\xed\xa0\x80", "\xf4\x90\x80\x80",
+        "\xff", "\xe2\x82", "\xc3", "\x01", "\x1f", "\x7f", "\n", std::string(1, '\0')
+    };
+
+    for (std::size_t run = 0; run <= 70; ++run)
+    {
+        for (const std::string& piece : pieces)
+        {
+            // a run of ordinary characters puts the piece at every offset of the bulk scanners
+            const std::string str = std::string(run, 'a') + piece;
+            for (const std::string& text :
+                    {
+                        "\"" + str + "\"", "[\"" + str + "\"]", "{\"" + str + "\":1}", "{\"k\":\"" + str + "\",\"" + str + "\":[\"" + str + "\"]}",
+                        "[\"" + str
+                    })
+            {
+                CAPTURE(text)
+                for (int options = 0; options < 4; ++options)
+                {
+                    const bool ignore_comments = (options & 1) != 0;
+                    const bool ignore_trailing_commas = (options & 2) != 0;
+                    const bool parses = !json::parse(text, nullptr, false, ignore_comments, ignore_trailing_commas).is_discarded();
+                    CHECK(json::accept(text, ignore_comments, ignore_trailing_commas) == parses);
+
+                    std::istringstream stream(text);
+                    CHECK(json::accept(stream, ignore_comments, ignore_trailing_commas) == parses);
+
+                    const std::list<char> chars(text.begin(), text.end());
+                    CHECK(json::accept(chars.begin(), chars.end(), ignore_comments, ignore_trailing_commas) == parses);
+                }
+            }
+        }
+    }
+}
+
 namespace
 {
 // a DOM parser that reports parse errors without throwing, as in the sax_no_exception example