scan_number()

template<typename BasicJsonType , typename InputAdapterType >

token_type nlohmann::detail::lexer< BasicJsonType, InputAdapterType >::scan_number ( )

inlineprivate

scan a number literal

This function scans a string according to Sect. 6 of RFC 8259.

The function is realized with a deterministic finite state machine derived from the grammar described in RFC 8259. Starting in state "init", the input is read and used to determined the next state. Only state "done" accepts the number. State "error" is a trap state to model errors. In the table below, "anything" means any character but the ones listed before.

state	0	1-9	e E	+	-	.	anything
init	zero	any1	[error]	[error]	minus	[error]	[error]
minus	zero	any1	[error]	[error]	[error]	[error]	[error]
zero	done	done	exponent	done	done	decimal1	done
any1	any1	any1	exponent	done	done	decimal1	done
decimal1	decimal2	decimal2	[error]	[error]	[error]	[error]	[error]
decimal2	decimal2	decimal2	exponent	done	done	done	done
exponent	any2	any2	[error]	sign	sign	[error]	[error]
sign	any2	any2	[error]	[error]	[error]	[error]	[error]
any2	any2	any2	done	done	done	done	done

The state machine is realized with one label per state (prefixed with "scan_number_") and goto statements between them. The state machine contains cycles, but any cycle can be left when EOF is read. Therefore, the function is guaranteed to terminate.

During scanning, the read bytes are stored in token_buffer. This string is then converted to a signed integer, an unsigned integer, or a floating-point number.

Returns

token_type::value_unsigned, token_type::value_integer, or token_type::value_float if number could be successfully scanned, token_type::parse_error otherwise

Note

The scanner is independent of the current locale. Internally, the locale's decimal point is used instead of . to work with the locale-dependent converters.

Definition at line 7566 of file json.hpp.

    {
        // reset token_buffer to store the number's bytes
        reset();
        // the type of the parsed number; initially set to unsigned; will be
        // changed if minus sign, decimal point or exponent is read
        token_type number_type = token_type::value_unsigned;
 
        // state (init): we just found out we need to scan a number
        switch (current)
        {
            case '-':
            {
                add(current);
                goto scan_number_minus;
            }
 
            case '0':
            {
                add(current);
                goto scan_number_zero;
            }
 
            case '1':
            case '2':
            case '3':
            case '4':
            case '5':
            case '6':
            case '7':
            case '8':
            case '9':
            {
                add(current);
                goto scan_number_any1;
            }
 
            // all other characters are rejected outside scan_number()
            default:            // LCOV_EXCL_LINE
                JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
        }
 
scan_number_minus:
        // state: we just parsed a leading minus sign
        number_type = token_type::value_integer;
 
        switch (get())
        {
            case '0':
            {
                add(current);
                goto scan_number_zero;
            }
 
            case '1':
            case '2':
            case '3':
            case '4':
            case '5':
            case '6':
            case '7':
            case '8':
            case '9':
            {
                add(current);
                goto scan_number_any1;
            }
 
            default:
            {
                error_message = "invalid number; expected digit after '-'";
                return token_type::parse_error;
            }
        }
 
scan_number_zero:
 
        // state: we just parse a zero (maybe with a leading minus sign)
        switch (get())
        {
            case '.':
            {
                add(decimal_point_char);
                goto scan_number_decimal1;
            }
 
            case 'e':
            case 'E':
            {
                add(current);
                goto scan_number_exponent;
            }
 
            default:
                goto scan_number_done;
        }
 
scan_number_any1:
 
        // state: we just parsed a number 0-9 (maybe with a leading minus sign)
        switch (get())
        {
            case '0':
            case '1':
            case '2':
            case '3':
            case '4':
            case '5':
            case '6':
            case '7':
            case '8':
            case '9':
            {
                add(current);
                goto scan_number_any1;
            }
 
            case '.':
            {
                add(decimal_point_char);
                goto scan_number_decimal1;
            }
 
            case 'e':
            case 'E':
            {
                add(current);
                goto scan_number_exponent;
            }
 
            default:
                goto scan_number_done;
        }
 
scan_number_decimal1:
        // state: we just parsed a decimal point
        number_type = token_type::value_float;
 
        switch (get())
        {
            case '0':
            case '1':
            case '2':
            case '3':
            case '4':
            case '5':
            case '6':
            case '7':
            case '8':
            case '9':
            {
                add(current);
                goto scan_number_decimal2;
            }
 
            default:
            {
                error_message = "invalid number; expected digit after '.'";
                return token_type::parse_error;
            }
        }
 
scan_number_decimal2:
 
        // we just parsed at least one number after a decimal point
        switch (get())
        {
            case '0':
            case '1':
            case '2':
            case '3':
            case '4':
            case '5':
            case '6':
            case '7':
            case '8':
            case '9':
            {
                add(current);
                goto scan_number_decimal2;
            }
 
            case 'e':
            case 'E':
            {
                add(current);
                goto scan_number_exponent;
            }
 
            default:
                goto scan_number_done;
        }
 
scan_number_exponent:
        // we just parsed an exponent
        number_type = token_type::value_float;
 
        switch (get())
        {
            case '+':
            case '-':
            {
                add(current);
                goto scan_number_sign;
            }
 
            case '0':
            case '1':
            case '2':
            case '3':
            case '4':
            case '5':
            case '6':
            case '7':
            case '8':
            case '9':
            {
                add(current);
                goto scan_number_any2;
            }
 
            default:
            {
                error_message =
                    "invalid number; expected '+', '-', or digit after exponent";
                return token_type::parse_error;
            }
        }
 
scan_number_sign:
 
        // we just parsed an exponent sign
        switch (get())
        {
            case '0':
            case '1':
            case '2':
            case '3':
            case '4':
            case '5':
            case '6':
            case '7':
            case '8':
            case '9':
            {
                add(current);
                goto scan_number_any2;
            }
 
            default:
            {
                error_message = "invalid number; expected digit after exponent sign";
                return token_type::parse_error;
            }
        }
 
scan_number_any2:
 
        // we just parsed a number after the exponent or exponent sign
        switch (get())
        {
            case '0':
            case '1':
            case '2':
            case '3':
            case '4':
            case '5':
            case '6':
            case '7':
            case '8':
            case '9':
            {
                add(current);
                goto scan_number_any2;
            }
 
            default:
                goto scan_number_done;
        }
 
scan_number_done:
        // unget the character after the number (we only read it to know that
        // we are done scanning a number)
        unget();
        char *endptr = nullptr; // NOLINT(cppcoreguidelines-pro-type-vararg,hicpp-vararg)
        errno = 0;
 
        // try to parse integers first and fall back to floats
        if (number_type == token_type::value_unsigned)
        {
            const auto x = std::strtoull(token_buffer.data(), &endptr, 10);
            // we checked the number format before
            JSON_ASSERT(endptr == token_buffer.data() + token_buffer.size());
 
            if (errno == 0)
            {
                value_unsigned = static_cast<number_unsigned_t>(x);
 
                if (value_unsigned == x)
                {
                    return token_type::value_unsigned;
                }
            }
        }
 
        else if (number_type == token_type::value_integer)
        {
            const auto x = std::strtoll(token_buffer.data(), &endptr, 10);
            // we checked the number format before
            JSON_ASSERT(endptr == token_buffer.data() + token_buffer.size());
 
            if (errno == 0)
            {
                value_integer = static_cast<number_integer_t>(x);
 
                if (value_integer == x)
                {
                    return token_type::value_integer;
                }
            }
        }
 
        // this code is reached if we parse a floating-point number or if an
        // integer conversion above failed
        strtof(value_float, token_buffer.data(), &endptr);
        // we checked the number format before
        JSON_ASSERT(endptr == token_buffer.data() + token_buffer.size());
        return token_type::value_float;
    }