Intel 9800758-02 manual Figure B-1. Internal Representation of Numbers and Strings

BASIC-80 Calling Non-BASIC-80 Subroutines

To return a single-precision floating-point, double-precision floating-point, or string result, you must use the appropriate data type character (see table 2-5 for a list of these characters) before the subroutine number in the USR function. For example, to tell BASIC-SO that a user-written subroutine returns a double-precision floating- point value:

240 A# = USR #15 (VARPTR(L#), A1)

In this example, the sharp sign (#) following USR tells BASIC-SO to reserve an S- byte space for the double-precision result of USR #15. A# is also defined as double- precision, but the parameters passed may be of any numeric type.

Your subroutine must interpret the first parameter passed to it as the storage address of A#, and it must also place the result of USR #15 there.

References to string parameters are handled in a similar manner. VARPTR of a string is the address of the string descriptor, not the string data. Thus if a user subroutine returns a string then the user should code USR$[n]. BASIC allocates a 255-character string and pass the address of the string descriptor to the subroutine. Your routine may change the string length in the string descriptor to return a shorter string, but may not change the string's address. Neither parameter strings nor parameter string descriptors should be changed.

Array variables are passed as parameters by referencing the first element of the array. BASIC-SO follows row-major order.

To code in 8080/S0S5 assembly language, you must know the Intel format for representing integer, single-precision floating-point, and double-precision floating point numbers. Figure E-l shows these representations.

~~D~ESS	ILS_f_I~~I________________1_5_B_IT_S_O_F_DA_T_A________________~			ADDRESS
				LOW
		INTEGER REPRESENTATION: 2 BYTES, 16 BITS, LOW ORDER BYTE FIRST
		8 EXPONENT BITS	23 MANTISSA BITS
		SINGLE-PRECISION FLOATING·POINTREPRESENTATION:
		4 BYTES, 32 BITS, LOW ORDER BYTE FIRST
		11 EXPONENT BITS	52 MANTISSA BITS
		DOUBLE-PRECISION FLOATING-POINT REPRESENTATION:
		8 BYTES, 84 BITS, LOW ORDER BYTE FIRST
		LENGTH	ADDRESS

STRING DESCRIPTOR REPRESENTATION: 8 BITS, STRING LENGTH: 16 BITS, STRING ADDRESS.

DATADATADATADATA

STRING DATA REPRESENTATION: 8 BIT BYTES.

Figure B-1. Internal Representation of Numbers and Strings

E-3