The INCLUDE control statement requests that the binder use additional data sets as input. These can be any of the sequential data set types acceptable for primary input.

In addition, INCLUDE can refer to private libraries rather than sequential files. Concatenations must contain only libraries or sequential files (including library members), not both.

The INCLUDE statement specifies the ddname of a DD statement that describes the data set to be used as additional input. If the DD statement describes a library (partitioned data set, PDSE, or z/OS UNIX directory) the INCLUDE statement also contains the name of each member to be used. See INCLUDE statement for the syntax of the INCLUDE statement.

When an INCLUDE control statement is encountered, the binder processes the module or modules indicated. Figure 1 shows the processing of an INCLUDE statement. In the illustration, the primary input data set is a sequential data set named OBJMOD that contains an INCLUDE statement. After processing the included data set, the binder processes the next primary input item. The arrows indicate the flow of processing.

If an included data set also contains an INCLUDE statement, that INCLUDE is processed at the time it is encountered, effectively nesting includes. Any number of nested INCLUDE statements are possible with the binder. Figure 1 demonstrates the flow of processing for single INCLUDE statements. Note that the binder returns to the Include module after processing the included module whereas the linkage editor does not.

Figure 1. Processing of one INCLUDE control statement

Figure 2 demonstrates the flow of processing for nested INCLUDE statements.

Figure 2. Processing of nested INCLUDE control statements