FastSort Manual

Using Supported File Types
FastSort Manual429834-003
C-3
Relative Files
Relative Files
You can use relative files as input or output files. For a relative output file, specify
either of the following:
R in the TYPE parameter of the TO command
1 in the out-file-type parameter of the SORTMERGESTART procedure
If the type of your first input file is relative, the default output file type is relative.
Entry-Sequenced Files
You can use entry-sequenced files as input and output files. For an entry-sequenced
output file you do not need to specify the type because entry-sequenced is the default
type (unless the first input file is not an entry-sequenced file). However, you can specify
either of the following:
E in the TYPE parameter of the TO command
2 in the out-file-type parameter of the SORTMERGESTART procedure
If the type of your first input file is entry-sequenced, the default output file type is entry-
sequenced.
Key-Sequenced Files
You can use key-sequenced files as input files and output files. No special
requirements apply to using key-sequenced input files.
If you use a key-sequenced output file, you can specify only one key field for sorting.
That field must be the same as the primary key field for the file. When using
commands, you must name the field in an ASCENDING command and specify
UNSIGNED as the data type. When using the SORTMERGESTART procedure, you
must specify ascending and BINARY UNSIGNED for the field in the key-block
parameter array.
To cause FastSort to create a key-sequenced output file or use an existing one, you
must specify the type. You can also specify the percentage of slack space for
accommodating future insertions of records in a new or existing key-sequenced file.
To specify the output file type, use either of the following:
K in the TYPE parameter of the TO command
3 in the out-file-type parameter of the SORTMERGESTART procedure
To specify the data and index slack, use any of the following:
The SLACK parameter of the RUN command, if you want the same percentage of
slack space in the data blocks and the index blocks