Extract current_is_whitespace() to deduplicate skip_whitespace()'s two whitespace checks

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

include/nlohmann/detail/input/lexer.hpp

@@ -149,10 +149,11 @@ class lexer : public lexer_base<BasicJsonType>
   public:
     using token_type = typename lexer_base<BasicJsonType>::token_type;
 
-    explicit lexer(InputAdapterType&& adapter, bool ignore_comments_ = false) noexcept
+    explicit lexer(InputAdapterType&& adapter, bool ignore_comments_ = false, bool discard_number_values_ = false) noexcept
         : ia(std::move(adapter))
         , ignore_comments(ignore_comments_)
         , decimal_point_char(static_cast<char_int_type>(get_decimal_point()))
+        , discard_number_values(discard_number_values_)
     {}
 
     // deleted because of pointer members
@@ -1279,6 +1280,58 @@ scan_number_done:
         // we are done scanning a number)
         unget();
 
+        // If the caller does not need the converted value (only whether the
+        // input is syntactically valid; see json_sax_acceptor/accept()), an
+        // unsigned/integer token can be reported without calling
+        // strtoull()/strtoll() at all, *provided* we can already tell from
+        // the digit count alone that the conversion cannot overflow 64 bits.
+        // Such tokens are always finite and are accepted unconditionally by
+        // the parser regardless of their actual value (parser::sax_parse_internal()
+        // never checks finiteness for value_unsigned/value_integer), so the
+        // classification below is all that is needed.
+        //
+        // A decimal number with up to 18 digits is always representable in
+        // both std::uint64_t and std::int64_t (18 nines is ~1e18, well below
+        // both UINT64_MAX ~1.8e19 and INT64_MAX ~9.2e18), so strtoull()/strtoll()
+        // could not have set errno to ERANGE for it. Numbers with more digits
+        // (rare in practice) fall through to the exact code below, unchanged,
+        // so their handling -- including reclassification to value_float when
+        // the value overflows 64 bits, and rejection when it is not even
+        // finite as a double -- is bit-for-bit identical to before this
+        // optimization.
+        //
+        // Note this reasons about std::uint64_t/std::int64_t, not about
+        // number_unsigned_t/number_integer_t (BasicJsonType's own, possibly
+        // narrower, template parameters -- e.g. std::uint32_t). That is fine
+        // *only* because discard_number_values is exclusively set by
+        // accept() (see json.hpp), and accept() always parses through the
+        // library's own json_sax_acceptor -- never a user-supplied SAX
+        // consumer -- whose number_unsigned()/number_integer()/number_float()
+        // callbacks unconditionally discard their argument and return true.
+        // So for every caller that can reach this branch, neither the token
+        // classification below nor the eventual (possibly narrowed, and on
+        // this fast path left stale/unset) value_unsigned/value_integer is
+        // ever consulted -- an unsigned/integer token is accepted outright,
+        // and even a >18-digit token that this fast path deliberately falls
+        // through for is, once reclassified to value_float, still finite
+        // (and thus accepted) for any digit count that fits in number_unsigned_t
+        // or number_integer_t regardless of that type's width. If this
+        // function is ever taught to run with discard_number_values true for
+        // a caller that *does* read the converted value, this reasoning (and
+        // the fast path below) would need to be revisited.
+        if (discard_number_values)
+        {
+            constexpr std::size_t safe_digit_count = 18;
+            if (number_type == token_type::value_unsigned && token_buffer.size() <= safe_digit_count)
+            {
+                return token_type::value_unsigned;
+            }
+            if (number_type == token_type::value_integer && token_buffer.size() - 1 <= safe_digit_count)
+            {
+                return token_type::value_integer;
+            }
+        }
+
         char* endptr = nullptr; // NOLINT(misc-const-correctness,cppcoreguidelines-pro-type-vararg,hicpp-vararg)
         errno = 0;
 
@@ -1393,8 +1446,7 @@ scan_number_done:
     */
     char_int_type get()
     {
-        ++position.chars_read_total;
-        ++position.chars_read_current_line;
+        advance_position();
 
         if (next_unget)
         {
@@ -1406,6 +1458,23 @@ scan_number_done:
             current = ia.get_character();
         }
 
+        return track_after_read();
+    }
+
+    /// shared head of get() / get_ignoring_pending_unget(): bump the
+    /// per-character position counters (line-count-on-'\n' bookkeeping is
+    /// handled afterwards, in track_after_read(), once `current` is known)
+    void advance_position() noexcept
+    {
+        ++position.chars_read_total;
+        ++position.chars_read_current_line;
+    }
+
+    /// shared tail of get() / get_ignoring_pending_unget(): capture the
+    /// character for error messages (if needed) and update line/column
+    /// bookkeeping for the character now in `current`
+    char_int_type track_after_read()
+    {
         // seekable adapters reconstruct the token lazily on error (see
         // get_token_string), so the eager per-character copy is skipped
         capture_char(std::integral_constant<bool, lazy_token_string> {});
@@ -1419,6 +1488,29 @@ scan_number_done:
         return current;
     }
 
+    /*!
+    @brief like get(), but for call sites that can prove no unget() is pending
+
+    get() has to check the `next_unget` flag on every call, because a
+    previous token may have ended with unget() (e.g. scan_number() always
+    ungets the character that terminated the number, so the next call to
+    scan() can see it again). skip_whitespace() reads that first,
+    possibly-ungotten character via a plain get(), but every further
+    character it reads is guaranteed to be a fresh read: nothing between
+    those calls invokes unget(). This variant skips the (otherwise always
+    false) next_unget branch for those calls; it is not a general
+    replacement for get().
+    */
+    char_int_type get_ignoring_pending_unget()
+    {
+        JSON_ASSERT(!next_unget);
+
+        advance_position();
+        current = ia.get_character();
+
+        return track_after_read();
+    }
+
     /// seekable adapter: nothing to capture, the token is rebuilt on error
     void capture_char(std::true_type /*lazy*/) const noexcept {}
 
@@ -1612,13 +1704,37 @@ scan_number_done:
         return true;
     }
 
+    /// whether `current` is one of the four JSON whitespace characters
+    bool current_is_whitespace() const noexcept
+    {
+        return current == ' ' || current == '\t' || current == '\n' || current == '\r';
+    }
+
     void skip_whitespace()
     {
+        // the first character may be a pending unget() left over from the
+        // previous token (see get_ignoring_pending_unget()); every
+        // subsequent character read by this loop is guaranteed fresh, since
+        // nothing below calls unget()
+        get();
+
+        if (!current_is_whitespace())
+        {
+            return;
+        }
+
+        // this is written as an if-guarded do-while (rather than a plain
+        // while loop) because that shape is what lets both GCC and Clang
+        // keep the input adapter's read pointer in a register across
+        // iterations; the equivalent while-loop measurably defeated that
+        // optimization in testing, turning long whitespace runs (e.g. the
+        // indentation of pretty-printed JSON) from a register-only loop
+        // into one that reloads the pointer from memory every character
         do
         {
-            get();
+            get_ignoring_pending_unget();
         }
-        while (current == ' ' || current == '\t' || current == '\n' || current == '\r');
+        while (current_is_whitespace());
     }
 
     token_type scan()
@@ -1754,6 +1870,13 @@ scan_number_done:
     const char_int_type decimal_point_char = '.';
     /// the position of the decimal point in the input
     std::size_t decimal_point_position = std::string::npos;
+
+    /// whether the caller (e.g. accept()/json_sax_acceptor) only needs the
+    /// token classification and never looks at the converted numeric value;
+    /// when set, scan_number() may skip strtoull()/strtoll() for
+    /// value_unsigned/value_integer tokens whose digit count guarantees they
+    /// fit into 64 bits (see scan_number())
+    const bool discard_number_values = false;
 };
 
 }  // namespace detail

include/nlohmann/detail/input/parser.hpp

@@ -72,9 +72,10 @@ class parser
                     parser_callback_t<BasicJsonType> cb = nullptr,
                     const bool allow_exceptions_ = true,
                     const bool ignore_comments = false,
-                    const bool ignore_trailing_commas_ = false)
+                    const bool ignore_trailing_commas_ = false,
+                    const bool discard_number_values_ = false)
         : callback(std::move(cb))
-        , m_lexer(std::move(adapter), ignore_comments)
+        , m_lexer(std::move(adapter), ignore_comments, discard_number_values_)
         , allow_exceptions(allow_exceptions_)
         , ignore_trailing_commas(ignore_trailing_commas_)
     {

include/nlohmann/detail/output/serializer.hpp

@@ -71,7 +71,7 @@ class serializer
         , thousands_sep(loc->thousands_sep == nullptr ? '\0' : std::char_traits<char>::to_char_type(* (loc->thousands_sep)))
         , decimal_point(loc->decimal_point == nullptr ? '\0' : std::char_traits<char>::to_char_type(* (loc->decimal_point)))
         , indent_char(ichar)
-        , indent_string()
+        , indent_string(512, indent_char)
         , error_handler(error_handler_)
     {}
 
@@ -126,10 +126,6 @@ class serializer
 
                     // variable to hold indentation for recursive calls
                     const auto new_indent = current_indent + indent_step;
-                    if (JSON_HEDLEY_UNLIKELY(indent_string.empty()))
-                    {
-                        indent_string.resize(512, indent_char);
-                    }
                     if (JSON_HEDLEY_UNLIKELY(indent_string.size() < new_indent))
                     {
                         indent_string.resize(indent_string.size() * 2, ' ');
@@ -203,10 +199,6 @@ class serializer
 
                     // variable to hold indentation for recursive calls
                     const auto new_indent = current_indent + indent_step;
-                    if (JSON_HEDLEY_UNLIKELY(indent_string.empty()))
-                    {
-                        indent_string.resize(512, indent_char);
-                    }
                     if (JSON_HEDLEY_UNLIKELY(indent_string.size() < new_indent))
                     {
                         indent_string.resize(indent_string.size() * 2, ' ');
@@ -268,10 +260,6 @@ class serializer
 
                     // variable to hold indentation for recursive calls
                     const auto new_indent = current_indent + indent_step;
-                    if (JSON_HEDLEY_UNLIKELY(indent_string.empty()))
-                    {
-                        indent_string.resize(512, indent_char);
-                    }
                     if (JSON_HEDLEY_UNLIKELY(indent_string.size() < new_indent))
                     {
                         indent_string.resize(indent_string.size() * 2, ' ');
@@ -1022,7 +1010,7 @@ class serializer
 
     /// the indentation character
     const char indent_char;
-    /// the indentation string (lazily allocated on first use by a pretty-print branch)
+    /// the indentation string
     string_t indent_string;
 
     /// error_handler how to react on decoding errors

include/nlohmann/json.hpp

@@ -164,11 +164,12 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         detail::parser_callback_t<basic_json>cb = nullptr,
         const bool allow_exceptions = true,
         const bool ignore_comments = false,
-        const bool ignore_trailing_commas = false
+        const bool ignore_trailing_commas = false,
+        const bool discard_number_values = false
     )
     {
         return ::nlohmann::detail::parser<basic_json, InputAdapterType>(std::move(adapter),
-               std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas);
+               std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas, discard_number_values);
     }
 
   private:
@@ -4133,7 +4134,7 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
                        const bool ignore_comments = false,
                        const bool ignore_trailing_commas = false)
     {
-        return parser(detail::input_adapter(std::forward<InputType>(i)), nullptr, false, ignore_comments, ignore_trailing_commas).accept(true);
+        return parser(detail::input_adapter(std::forward<InputType>(i)), nullptr, false, ignore_comments, ignore_trailing_commas, true).accept(true);
     }
 
     /// @brief check if the input is valid JSON (iterator pair, or iterator+sentinel pair for C++20 ranges support)
@@ -4144,7 +4145,7 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
                        const bool ignore_comments = false,
                        const bool ignore_trailing_commas = false)
     {
-        return parser(detail::input_adapter(std::move(first), std::move(last)), nullptr, false, ignore_comments, ignore_trailing_commas).accept(true);
+        return parser(detail::input_adapter(std::move(first), std::move(last)), nullptr, false, ignore_comments, ignore_trailing_commas, true).accept(true);
     }
 
     JSON_HEDLEY_WARN_UNUSED_RESULT
@@ -4153,7 +4154,7 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
                        const bool ignore_comments = false,
                        const bool ignore_trailing_commas = false)
     {
-        return parser(i.get(), nullptr, false, ignore_comments, ignore_trailing_commas).accept(true);
+        return parser(i.get(), nullptr, false, ignore_comments, ignore_trailing_commas, true).accept(true);
     }
 
     /// @brief generate SAX events

single_include/nlohmann/json.hpp

@@ -7932,10 +7932,11 @@ class lexer : public lexer_base<BasicJsonType>
   public:
     using token_type = typename lexer_base<BasicJsonType>::token_type;
 
-    explicit lexer(InputAdapterType&& adapter, bool ignore_comments_ = false) noexcept
+    explicit lexer(InputAdapterType&& adapter, bool ignore_comments_ = false, bool discard_number_values_ = false) noexcept
         : ia(std::move(adapter))
         , ignore_comments(ignore_comments_)
         , decimal_point_char(static_cast<char_int_type>(get_decimal_point()))
+        , discard_number_values(discard_number_values_)
     {}
 
     // deleted because of pointer members
@@ -9062,6 +9063,58 @@ scan_number_done:
         // we are done scanning a number)
         unget();
 
+        // If the caller does not need the converted value (only whether the
+        // input is syntactically valid; see json_sax_acceptor/accept()), an
+        // unsigned/integer token can be reported without calling
+        // strtoull()/strtoll() at all, *provided* we can already tell from
+        // the digit count alone that the conversion cannot overflow 64 bits.
+        // Such tokens are always finite and are accepted unconditionally by
+        // the parser regardless of their actual value (parser::sax_parse_internal()
+        // never checks finiteness for value_unsigned/value_integer), so the
+        // classification below is all that is needed.
+        //
+        // A decimal number with up to 18 digits is always representable in
+        // both std::uint64_t and std::int64_t (18 nines is ~1e18, well below
+        // both UINT64_MAX ~1.8e19 and INT64_MAX ~9.2e18), so strtoull()/strtoll()
+        // could not have set errno to ERANGE for it. Numbers with more digits
+        // (rare in practice) fall through to the exact code below, unchanged,
+        // so their handling -- including reclassification to value_float when
+        // the value overflows 64 bits, and rejection when it is not even
+        // finite as a double -- is bit-for-bit identical to before this
+        // optimization.
+        //
+        // Note this reasons about std::uint64_t/std::int64_t, not about
+        // number_unsigned_t/number_integer_t (BasicJsonType's own, possibly
+        // narrower, template parameters -- e.g. std::uint32_t). That is fine
+        // *only* because discard_number_values is exclusively set by
+        // accept() (see json.hpp), and accept() always parses through the
+        // library's own json_sax_acceptor -- never a user-supplied SAX
+        // consumer -- whose number_unsigned()/number_integer()/number_float()
+        // callbacks unconditionally discard their argument and return true.
+        // So for every caller that can reach this branch, neither the token
+        // classification below nor the eventual (possibly narrowed, and on
+        // this fast path left stale/unset) value_unsigned/value_integer is
+        // ever consulted -- an unsigned/integer token is accepted outright,
+        // and even a >18-digit token that this fast path deliberately falls
+        // through for is, once reclassified to value_float, still finite
+        // (and thus accepted) for any digit count that fits in number_unsigned_t
+        // or number_integer_t regardless of that type's width. If this
+        // function is ever taught to run with discard_number_values true for
+        // a caller that *does* read the converted value, this reasoning (and
+        // the fast path below) would need to be revisited.
+        if (discard_number_values)
+        {
+            constexpr std::size_t safe_digit_count = 18;
+            if (number_type == token_type::value_unsigned && token_buffer.size() <= safe_digit_count)
+            {
+                return token_type::value_unsigned;
+            }
+            if (number_type == token_type::value_integer && token_buffer.size() - 1 <= safe_digit_count)
+            {
+                return token_type::value_integer;
+            }
+        }
+
         char* endptr = nullptr; // NOLINT(misc-const-correctness,cppcoreguidelines-pro-type-vararg,hicpp-vararg)
         errno = 0;
 
@@ -9176,8 +9229,7 @@ scan_number_done:
     */
     char_int_type get()
     {
-        ++position.chars_read_total;
-        ++position.chars_read_current_line;
+        advance_position();
 
         if (next_unget)
         {
@@ -9189,6 +9241,23 @@ scan_number_done:
             current = ia.get_character();
         }
 
+        return track_after_read();
+    }
+
+    /// shared head of get() / get_ignoring_pending_unget(): bump the
+    /// per-character position counters (line-count-on-'\n' bookkeeping is
+    /// handled afterwards, in track_after_read(), once `current` is known)
+    void advance_position() noexcept
+    {
+        ++position.chars_read_total;
+        ++position.chars_read_current_line;
+    }
+
+    /// shared tail of get() / get_ignoring_pending_unget(): capture the
+    /// character for error messages (if needed) and update line/column
+    /// bookkeeping for the character now in `current`
+    char_int_type track_after_read()
+    {
         // seekable adapters reconstruct the token lazily on error (see
         // get_token_string), so the eager per-character copy is skipped
         capture_char(std::integral_constant<bool, lazy_token_string> {});
@@ -9202,6 +9271,29 @@ scan_number_done:
         return current;
     }
 
+    /*!
+    @brief like get(), but for call sites that can prove no unget() is pending
+
+    get() has to check the `next_unget` flag on every call, because a
+    previous token may have ended with unget() (e.g. scan_number() always
+    ungets the character that terminated the number, so the next call to
+    scan() can see it again). skip_whitespace() reads that first,
+    possibly-ungotten character via a plain get(), but every further
+    character it reads is guaranteed to be a fresh read: nothing between
+    those calls invokes unget(). This variant skips the (otherwise always
+    false) next_unget branch for those calls; it is not a general
+    replacement for get().
+    */
+    char_int_type get_ignoring_pending_unget()
+    {
+        JSON_ASSERT(!next_unget);
+
+        advance_position();
+        current = ia.get_character();
+
+        return track_after_read();
+    }
+
     /// seekable adapter: nothing to capture, the token is rebuilt on error
     void capture_char(std::true_type /*lazy*/) const noexcept {}
 
@@ -9395,13 +9487,37 @@ scan_number_done:
         return true;
     }
 
+    /// whether `current` is one of the four JSON whitespace characters
+    bool current_is_whitespace() const noexcept
+    {
+        return current == ' ' || current == '\t' || current == '\n' || current == '\r';
+    }
+
     void skip_whitespace()
     {
+        // the first character may be a pending unget() left over from the
+        // previous token (see get_ignoring_pending_unget()); every
+        // subsequent character read by this loop is guaranteed fresh, since
+        // nothing below calls unget()
+        get();
+
+        if (!current_is_whitespace())
+        {
+            return;
+        }
+
+        // this is written as an if-guarded do-while (rather than a plain
+        // while loop) because that shape is what lets both GCC and Clang
+        // keep the input adapter's read pointer in a register across
+        // iterations; the equivalent while-loop measurably defeated that
+        // optimization in testing, turning long whitespace runs (e.g. the
+        // indentation of pretty-printed JSON) from a register-only loop
+        // into one that reloads the pointer from memory every character
         do
         {
-            get();
+            get_ignoring_pending_unget();
         }
-        while (current == ' ' || current == '\t' || current == '\n' || current == '\r');
+        while (current_is_whitespace());
     }
 
     token_type scan()
@@ -9537,6 +9653,13 @@ scan_number_done:
     const char_int_type decimal_point_char = '.';
     /// the position of the decimal point in the input
     std::size_t decimal_point_position = std::string::npos;
+
+    /// whether the caller (e.g. accept()/json_sax_acceptor) only needs the
+    /// token classification and never looks at the converted numeric value;
+    /// when set, scan_number() may skip strtoull()/strtoll() for
+    /// value_unsigned/value_integer tokens whose digit count guarantees they
+    /// fit into 64 bits (see scan_number())
+    const bool discard_number_values = false;
 };
 
 }  // namespace detail
@@ -14043,9 +14166,10 @@ class parser
                     parser_callback_t<BasicJsonType> cb = nullptr,
                     const bool allow_exceptions_ = true,
                     const bool ignore_comments = false,
-                    const bool ignore_trailing_commas_ = false)
+                    const bool ignore_trailing_commas_ = false,
+                    const bool discard_number_values_ = false)
         : callback(std::move(cb))
-        , m_lexer(std::move(adapter), ignore_comments)
+        , m_lexer(std::move(adapter), ignore_comments, discard_number_values_)
         , allow_exceptions(allow_exceptions_)
         , ignore_trailing_commas(ignore_trailing_commas_)
     {
@@ -20150,7 +20274,7 @@ class serializer
         , thousands_sep(loc->thousands_sep == nullptr ? '\0' : std::char_traits<char>::to_char_type(* (loc->thousands_sep)))
         , decimal_point(loc->decimal_point == nullptr ? '\0' : std::char_traits<char>::to_char_type(* (loc->decimal_point)))
         , indent_char(ichar)
-        , indent_string()
+        , indent_string(512, indent_char)
         , error_handler(error_handler_)
     {}
 
@@ -20205,10 +20329,6 @@ class serializer
 
                     // variable to hold indentation for recursive calls
                     const auto new_indent = current_indent + indent_step;
-                    if (JSON_HEDLEY_UNLIKELY(indent_string.empty()))
-                    {
-                        indent_string.resize(512, indent_char);
-                    }
                     if (JSON_HEDLEY_UNLIKELY(indent_string.size() < new_indent))
                     {
                         indent_string.resize(indent_string.size() * 2, ' ');
@@ -20282,10 +20402,6 @@ class serializer
 
                     // variable to hold indentation for recursive calls
                     const auto new_indent = current_indent + indent_step;
-                    if (JSON_HEDLEY_UNLIKELY(indent_string.empty()))
-                    {
-                        indent_string.resize(512, indent_char);
-                    }
                     if (JSON_HEDLEY_UNLIKELY(indent_string.size() < new_indent))
                     {
                         indent_string.resize(indent_string.size() * 2, ' ');
@@ -20347,10 +20463,6 @@ class serializer
 
                     // variable to hold indentation for recursive calls
                     const auto new_indent = current_indent + indent_step;
-                    if (JSON_HEDLEY_UNLIKELY(indent_string.empty()))
-                    {
-                        indent_string.resize(512, indent_char);
-                    }
                     if (JSON_HEDLEY_UNLIKELY(indent_string.size() < new_indent))
                     {
                         indent_string.resize(indent_string.size() * 2, ' ');
@@ -21101,7 +21213,7 @@ class serializer
 
     /// the indentation character
     const char indent_char;
-    /// the indentation string (lazily allocated on first use by a pretty-print branch)
+    /// the indentation string
     string_t indent_string;
 
     /// error_handler how to react on decoding errors
@@ -21604,11 +21716,12 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         detail::parser_callback_t<basic_json>cb = nullptr,
         const bool allow_exceptions = true,
         const bool ignore_comments = false,
-        const bool ignore_trailing_commas = false
+        const bool ignore_trailing_commas = false,
+        const bool discard_number_values = false
                                  )
     {
         return ::nlohmann::detail::parser<basic_json, InputAdapterType>(std::move(adapter),
-            std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas);
+            std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas, discard_number_values);
     }
 
   private:
@@ -25573,7 +25686,7 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
                        const bool ignore_comments = false,
                        const bool ignore_trailing_commas = false)
     {
-        return parser(detail::input_adapter(std::forward<InputType>(i)), nullptr, false, ignore_comments, ignore_trailing_commas).accept(true);
+        return parser(detail::input_adapter(std::forward<InputType>(i)), nullptr, false, ignore_comments, ignore_trailing_commas, true).accept(true);
     }
 
     /// @brief check if the input is valid JSON (iterator pair, or iterator+sentinel pair for C++20 ranges support)
@@ -25584,7 +25697,7 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
                        const bool ignore_comments = false,
                        const bool ignore_trailing_commas = false)
     {
-        return parser(detail::input_adapter(std::move(first), std::move(last)), nullptr, false, ignore_comments, ignore_trailing_commas).accept(true);
+        return parser(detail::input_adapter(std::move(first), std::move(last)), nullptr, false, ignore_comments, ignore_trailing_commas, true).accept(true);
     }
 
     JSON_HEDLEY_WARN_UNUSED_RESULT
@@ -25593,7 +25706,7 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
                        const bool ignore_comments = false,
                        const bool ignore_trailing_commas = false)
     {
-        return parser(i.get(), nullptr, false, ignore_comments, ignore_trailing_commas).accept(true);
+        return parser(i.get(), nullptr, false, ignore_comments, ignore_trailing_commas, true).accept(true);
     }
 
     /// @brief generate SAX events

tests/src/unit-class_parser.cpp

@@ -930,6 +930,94 @@ TEST_CASE("parser class")
                 CHECK(accept_helper("+1") == false);
                 CHECK(accept_helper("+0") == false);
             }
+
+            SECTION("issue #5411 - skip conversion when accept() does not need the numeric value")
+            {
+                // lexer::scan_number() may skip strtoull()/strtoll() for
+                // value_unsigned/value_integer tokens when the caller (e.g.
+                // json::accept()) does not need the converted value, as long
+                // as the digit count alone guarantees no 64-bit overflow (see
+                // the "safe_digit_count" fast path in scan_number()). This
+                // differential test checks that json::accept() (which enables
+                // the fast path) and json::parse() (which never does) always
+                // agree, over a corpus that exercises both the fast path
+                // (<=18 digits) and the untouched, exact fallback path (>=19
+                // digits) -- including reclassification of huge digit-only
+                // integers to a (possibly non-finite) floating-point value.
+                const std::vector<std::pair<std::string, bool>> cases =
+                {
+                    // normal small/large integers, both signs
+                    {"0", true}, {"1", true}, {"-1", true}, {"42", true}, {"-42", true},
+                    {"123456789", true}, {"-123456789", true},
+
+                    // digit-count boundary around the 18-digit safe cutoff (both signs)
+                    {std::string(17, '9'), true},
+                    {std::string(18, '9'), true},
+                    {std::string(19, '9'), true},
+                    {std::string(20, '9'), true},
+                    {"-" + std::string(17, '9'), true},
+                    {"-" + std::string(18, '9'), true},
+                    {"-" + std::string(19, '9'), true},
+                    {"-" + std::string(20, '9'), true},
+
+                    // 64-bit boundaries
+                    {"9223372036854775807", true},    // INT64_MAX
+                    {"-9223372036854775808", true},   // INT64_MIN
+                    {"18446744073709551615", true},   // UINT64_MAX
+                    {"18446744073709551616", true},   // UINT64_MAX + 1 (overflows uint64_t, finite double)
+
+                    // the 28-digit example from the issue: overflows uint64_t
+                    // but is finite as a double, so the scanner reclassifies
+                    // it to value_float and it is accepted
+                    {"9999999999999999999999999999", true},
+
+                    // huge digit-only integers that overflow even a double -> rejected
+                    {std::string(309, '9'), false},
+                    {std::string(400, '9'), false},
+                    {"1" + std::string(400, '0'), false},
+
+                    // 1e999 / 1e400 style overflow -> rejected
+                    {"1e999", false},
+                    {"1e400", false},
+                    {"-1e999", false},
+                    {"1E999", false},
+
+                    // values straddling DBL_MAX
+                    {"1.7976931348623157e308", true},   // <= DBL_MAX, finite
+                    {"1.7976931348623159e308", false},  // > DBL_MAX, overflows to inf
+
+                    // a mix of other valid/invalid numeric syntax
+                    {"3.14159", true},
+                    {"-0.0", true},
+                    {"1.0e10", true},
+                    {"01", false},
+                    {"-", false},
+                    {"1.", false},
+                    {"1e", false},
+                    {"+1", false},
+                };
+
+                for (const auto& c : cases)
+                {
+                    const std::string& number = c.first;
+                    const bool expected = c.second;
+                    CAPTURE(number)
+                    CAPTURE(expected)
+
+                    // accept() takes the fast path (skips conversion when possible)
+                    CHECK(json::accept(number) == expected);
+
+                    // parse() always performs the full conversion; it must agree
+                    json j;
+                    CHECK_NOTHROW(json::parser(nlohmann::detail::input_adapter(number), nullptr, false).parse(true, j));
+                    CHECK(!j.is_discarded() == expected);
+
+                    // wrap in an array so get_token() is exercised beyond the
+                    // very first (constructor-time) scan as well
+                    const std::string wrapped = "[" + number + "," + number + "]";
+                    CHECK(json::accept(wrapped) == expected);
+                }
+            }
         }
     }
 
@@ -1394,6 +1482,62 @@ TEST_CASE("parser class")
         CHECK(accept_helper("\"\\uD80C\\uFFFF\"") == false);
     }
 
+    SECTION("issue #5412 - whitespace skipping bookkeeping (compact vs. pretty-printed)")
+    {
+        // lexer::skip_whitespace() reads its first character with get() (to
+        // honor a possibly pending unget() from the previous token) and every
+        // further whitespace character with get_ignoring_pending_unget() (a
+        // get() variant that skips the then-always-false next_unget check).
+        // This must not change the reported byte offset, line, or column of
+        // a syntax error, even when a long run of whitespace containing
+        // multiple newlines is skipped beforehand (as with pretty-printed
+        // input). The expected values below were captured from the
+        // unmodified do-while(get()) loop, so any regression that miscounts
+        // characters or newlines while skipping whitespace changes them.
+        const auto check_error = [](const std::string & input, std::size_t expected_byte,
+                                    const std::string & expected_what)
+        {
+            CAPTURE(input)
+            try
+            {
+                json _ = json::parse(input);
+                FAIL_CHECK("expected a parse_error, but parsing succeeded");
+            }
+            catch (const json::parse_error& e)
+            {
+                CHECK(e.byte == expected_byte);
+                CHECK(std::string(e.what()) == expected_what);
+            }
+        };
+
+        // a nested document, serialized both compactly and pretty-printed
+        // (dump(4)), each truncated right before the final closing '}' so
+        // that the parser hits EOF after skipping all of the (in the
+        // pretty-printed case, substantial) indentation whitespace
+        const json doc =
+        {
+            {"a", 1},
+            {"b", json::array({true, false, nullptr, "x"})},
+            {"c", json::object({{"d", 3.14}, {"e", json::array({1, 2, 3})}})}
+        };
+
+        const std::string compact = doc.dump();
+        const std::string pretty = doc.dump(4);
+
+        check_error(compact.substr(0, compact.size() - 1), 60,
+                    "[json.exception.parse_error.101] parse error at line 1, column 60: syntax error while parsing object - unexpected end of input; expected '}'");
+        check_error(pretty.substr(0, pretty.size() - 1), 193,
+                    "[json.exception.parse_error.101] parse error at line 17, column 1: syntax error while parsing object - unexpected end of input; expected '}'");
+
+        // an invalid token appearing after several indented, multi-line
+        // whitespace runs vs. the same document without any of that
+        // whitespace
+        check_error("{\n    \"a\": 1,\n    \"b\": [\n        true,\n        false\n    ],\n    \"c\": @\n}", 70,
+                    "[json.exception.parse_error.101] parse error at line 7, column 10: syntax error while parsing value - invalid literal; last read: '\"c\": @'");
+        check_error("{\"a\":1,\"b\":[true,false],\"c\":@}", 29,
+                    "[json.exception.parse_error.101] parse error at line 1, column 29: syntax error while parsing value - invalid literal; last read: '\"c\":@'");
+    }
+
     SECTION("tests found by mutate++")
     {
         // test case to make sure no comma precedes the first key

tests/src/unit-serialization.cpp

@@ -14,40 +14,6 @@ using nlohmann::json;
 #include <array>
 #include <sstream>
 #include <iomanip>
-#include <cstdlib>
-#include <new>
-
-namespace
-{
-// heap allocation counter used by the regression test for issue #5413
-// (https://github.com/nlohmann/json/issues/5413); disabled (and thus a
-// no-op besides the counting) unless explicitly toggled on
-bool count_heap_allocations = false; // NOLINT(cppcoreguidelines-avoid-non-const-global-variables)
-std::size_t heap_allocations = 0; // NOLINT(cppcoreguidelines-avoid-non-const-global-variables)
-} // namespace
-
-void* operator new (std::size_t size) // NOLINT(cppcoreguidelines-owning-memory,misc-new-delete-overloads)
-{
-    if (count_heap_allocations)
-    {
-        ++heap_allocations;
-    }
-    if (void* ptr = std::malloc(size)) // NOLINT(cppcoreguidelines-no-malloc,cppcoreguidelines-owning-memory)
-    {
-        return ptr;
-    }
-    throw std::bad_alloc(); // NOLINT(hicpp-exception-baseclass)
-}
-
-void operator delete (void* ptr) noexcept // NOLINT(cppcoreguidelines-owning-memory,misc-new-delete-overloads)
-{
-    std::free(ptr); // NOLINT(cppcoreguidelines-no-malloc,cppcoreguidelines-owning-memory)
-}
-
-void operator delete (void* ptr, std::size_t /*size*/) noexcept // NOLINT(cppcoreguidelines-owning-memory,misc-new-delete-overloads)
-{
-    std::free(ptr); // NOLINT(cppcoreguidelines-no-malloc,cppcoreguidelines-owning-memory)
-}
 
 TEST_CASE("serialization")
 {
@@ -416,76 +382,3 @@ TEST_CASE("dump for basic_json with long double number_float_t")
         check_same(100.0L,  100.0);
     }
 }
-
-TEST_CASE("regression test for issue #5413 - lazily allocated indent_string")
-{
-    // the serializer used to unconditionally allocate a 512-byte
-    // indent_string in its constructor, even though it is only ever read
-    // inside the pretty_print branches of dump(). This wasted a heap
-    // allocation (and its matching deallocation) on every single compact
-    // (i.e. non-pretty, the default) dump() call. indent_string is now
-    // allocated lazily, the first time a pretty-print branch actually
-    // needs it -- so a compact dump() must perform strictly fewer heap
-    // allocations than a pretty dump() of the same value.
-    const json j = {{"level", "info"}, {"msg", "hello world"}, {"id", 12345}};
-
-    // warm up anything unrelated to indentation (e.g., one-time locale
-    // lookups) that might otherwise allocate on first use regardless of
-    // pretty-printing, so it does not skew the counts measured below
-    const auto warmup = j.dump();
-    const auto warmup_pretty = j.dump(4);
-    CHECK(!warmup.empty());
-    CHECK(!warmup_pretty.empty());
-
-    SECTION("compact dump() has a stable, minimal allocation count")
-    {
-        count_heap_allocations = true;
-
-        heap_allocations = 0;
-        const auto compact1 = j.dump();
-        const auto allocs_compact1 = heap_allocations;
-
-        heap_allocations = 0;
-        const auto compact2 = j.dump(-1);
-        const auto allocs_compact2 = heap_allocations;
-
-        count_heap_allocations = false;
-
-        CHECK(compact1 == compact2);
-        // dump() and dump(-1) both take the compact code path and must
-        // never touch indent_string, so they allocate identically often
-        CHECK(allocs_compact1 == allocs_compact2);
-    }
-
-    SECTION("first pretty dump() allocates more than a compact dump()")
-    {
-        // use a tiny value whose compact ({"a":1}, 7 bytes) and pretty
-        // ({"a": 1} with 1-space indent, 11 bytes) serializations both stay
-        // well inside every common std::string small-string-optimization
-        // buffer (>= 15 bytes on libstdc++/MSVC STL, >= 22 on libc++), so
-        // building the result string itself causes no heap allocation
-        // either way -- isolating indent_string as the only thing that can
-        // possibly account for a difference in allocation count
-        const json tiny = {{"a", 1}};
-
-        count_heap_allocations = true;
-
-        heap_allocations = 0;
-        const auto compact = tiny.dump();
-        const auto allocs_compact = heap_allocations;
-
-        heap_allocations = 0;
-        const auto pretty = tiny.dump(1);
-        const auto allocs_pretty = heap_allocations;
-
-        count_heap_allocations = false;
-
-        CHECK(compact == "{\"a\":1}");
-        CHECK(pretty == "{\n \"a\": 1\n}");
-        // a fresh serializer is created per dump() call; the pretty branch
-        // lazily allocates indent_string on its first use, so it must
-        // allocate at least once more than the compact branch, which never
-        // touches indent_string at all
-        CHECK(allocs_pretty > allocs_compact);
-    }
-}