Number Processing

Number reader

The library has a built-in high-performance number reader,
it will read numbers according to these rules by default:

• Positive integers are read as uint64_t. If an overflow occurs, it is converted to double.
• Negative integers are read as int64_t. If an overflow occurs, it is converted to double.
• Floating-point numbers are read as double with correct rounding.
• If a double number overflows (reaches infinity), an error is reported.
• If a number does not conform to the JSON standard, an error is reported.

There are 3 flags that can be used to adjust the number parsing strategy:

• YYJSON_READ_ALLOW_INF_AND_NAN: read nan/inf number or literal as double (non-standard).
• YYJSON_READ_NUMBER_AS_RAW: read all numbers as raw strings without parsing.
• YYJSON_READ_BIGNUM_AS_RAW: read big numbers (overflow or infinity) as raw strings without parsing.

See the Reader flag section for more details.

Number writer

The library has a built-in high-performance number writer,
it will write numbers according to these rules by default:

• Positive integers are written without a sign.
• Negative integers are written with a negative sign.
• Floating-point numbers are written using the ECMAScript format, with the following modifications:
  • If the number is Infinity or NaN, an error is reported.
  • The negative sign of -0.0 is preserved to maintain input information.
  • The positive sign in the exponent part is removed.
• The floating-point number writer will generate the shortest correctly rounded decimal representation.

There are several flags that can be used to adjust the number writing strategy:

• YYJSON_WRITE_ALLOW_INF_AND_NAN writes inf/nan numbers as Infinity and NaN literals without error (non-standard).
• YYJSON_WRITE_INF_AND_NAN_AS_NULL writes inf/nan numbers as null literal.
• YYJSON_WRITE_FP_TO_FLOAT writes real numbers as float instead of double.
• YYJSON_WRITE_FP_TO_FIXED(prec) writes real numbers using fixed-point notation.

See the Writer flag section for more details.

There are also some helper functions to control the output format of individual values:

• yyjson_set_fp_to_float(yyjson_val *val, bool flt) and yyjson_mut_set_fp_to_float(yyjson_mut_val *val, bool flt) write this real number with float or double precision.
• yyjson_set_fp_to_fixed(yyjson_val *val, int prec) and yyjson_mut_set_fp_to_fixed(yyjson_mut_val *val, int prec) write this real number using fixed-point notation, the prec should be in the range of 1 to 15.

Number conversion function

There are also two utility functions that provide direct access to the library’s internal number conversion logic.
They are intended for standalone use and typically do not allocate memory.

// parse a number from string
const char *yyjson_read_number(const char *dat,
                               yyjson_val *val,
                               yyjson_read_flag flg,
                               const yyjson_alc *alc,
                               yyjson_read_err *err);
// write a number to string
char *yyjson_write_number(const yyjson_val *val, char *buf);