B

The B (binary) format specification is used for fast and efficient internal numeric storage. Binary format is machine specific. There are two types: BH stores the most significant byte first, and BL stores the least significant byte first. B format defaults to BH or BL, whichever is the native format of the hardware. When portability is desired, BH or BL should be used. When speed is desired, B should be used. In general, systems that utilize Intel chips will use a native format of BL, and those that don't will use BH. The following chart indicates the format native to several hardware/operating system combinations:

COMPUTER	NATIVE FORMAT
AT&T with UNIX	BH
IBM PC, XT, AT, with DOS	BL
IBM AT with LINUX	BL
NCR Tower with UNIX	BH

The field length for a B-formatted number can be 1, 2, 3 or 4. The value range for each field length is as follows:

FIELD	VALUE
LENGTH	RANGE
1	+ / - 127
2	+ / - 32, 767
3	+ / - 8, 388,607
4	+ / - 2, 147, 483, 647

Technical Considerations

1.) As with ZD and PD, the decimal point is implied for B format specifications; it is not stored with the number. To retrieve the same number as was written, the same format specification must be used. Reading a field with a B 4 that was written with a B 4.2 results in a number 100 times larger than was written. The number is rounded to the specified number of decimal positions before writing.