Manual

June 2004 Isis® Sonar User's Manual, Volume 2 6
Table A-4. Data representation types for Q-MIPS headers and footers
Data Types Type Definitions and Possible Range of Values
SHORT signed, two’s complement integer two bytes INT format
[-32,768 to 32,767] or four bytes LONG [-2,147,483,648 to
2,147,483,647] format
USHORT unsigned integer two bytes WORD format [0 to 65,535] or four
bytes DWORD format [0 to 4,294,967,296]
FLOAT floating point number in IEEE single precision standard four-byte
format FLOAT format [10-38 to1038]
DOUBLE IEEE double precision standard eight-byte DOUBLE format
[10-308 to 10 308]
CHAR signed character one byte CHAR format [-128 to 127]; can be
interpreted by ASCII code or as an 8-bit signed integer
BYTE unsigned character one byte BYTE format [0 to 255]; can be
interpreted by ASCII code or as an 8-bit unsigned integer
The exact byte and word ordering for each of these types and the IEEE floating
point formats are described below. Any unused fields are filled with zeros, that is,
each bit within the field will be a zero instead of a one. The integers will return the
value
0, the floating point numbers will return the value 0.0, and the characters
will return the value
0 or the null character \0, depending on how they are
interpreted. In Table A-5 on page 9 and Table A-6 on page 16, which describe
the complete header and footer formats, respectively, one column indicates the
type for each field.
The Q-MIPS format data are stored according to the Intel processor scheme for
representation of numbers in memory. This is described below. If a number is
represented by N bytes from most significant (MSB, Byte N-1) to least significant
(LSB, Byte 0), and N x 8 bits from bit N x 8 -1 (MSB) to bit 0 (LSB), the stored
bytes are found in the order described below. The bits within each byte are
always in order from most significant (bit 7) to least significant (bit 0). All Intel
80x86 chips are Little Endian.
Appendix A: Q-MIPS File Format