char *
     ecvt(double value, int ndigit, int * restrict decpt,
         int * restrict sign);

     char *
     fcvt(double value, int ndigit, int * restrict decpt,
         int * restrict sign);

     char *
     gcvt(double value, int ndigit, char *buf);

DESCRIPTION

     These functions are provided for compatibility with legacy code.  New
     code should use the snprintf(3) function for improved safety and porta-
     bility.

     The ecvt(), fcvt() and gcvt() functions convert the double precision
     floating-point number value to a NUL-terminated ASCII string.

     The ecvt() function converts value to a NUL-terminated string of exactly
     ndigit digits and returns a pointer to that string.  The result is padded
     with zeroes from left to right as needed.  There are no leading zeroes
     unless value itself is 0.  The least significant digit is rounded in an
     implementation-dependent manner.  The position of the decimal point rela-
     tive to the beginning of the string is stored in decpt.  A negative value
     indicates that the decimal point is located to the left of the returned
     digits (this occurs when there is no whole number component to value).
     If value is zero, it is unspecified whether the integer pointed to by
     decpt will be 0 or 1.  The decimal point itself is not included in the
     returned string.  If the sign of the result is negative, the integer
     pointed to by sign is non-zero; otherwise, it is 0.

     If the converted value is out of range or is not representable, the con-
     tents of the returned string are unspecified.

     The fcvt() function is identical to ecvt() with the exception that ndigit
     specifies the number of digits after the decimal point (zero-padded as
     needed).

     The gcvt() function converts value to a NUL-terminated string similar to
     the %g printf(3) format specifier and stores the result in buf.  It pro-
     duces ndigit significant digits similar to the %f printf(3) format speci-
     fier where possible.  If ndigit does allow sufficient precision, the
     result is stored in exponential notation similar to the %e printf(3) for-
     mat specifier.  If value is less than zero, buf will be prefixed with a
     minus sign.  A decimal point is included in the returned string if value
     is not a whole number.  Unlike the ecvt() and fcvt() functions, buf is
     not zero-padded.

RETURN VALUES

     The ecvt(), fcvt() and gcvt() functions return a NUL-terminated string

STANDARDS

     The ecvt(), fcvt() and gcvt() functions conform to IEEE Std 1003.1-2001
     (``POSIX.1'').

BSD                            December 1, 2002                            BSD