Building DLIB entries

If a COPY or COPYMOD statement is followed either by another COPY or COPYMOD statement or by an EXCLUDE statement, SMP/E assumes that the INDD library is totally copied to the OUTDD library. Because SMP/E does not look at either the DLIB or target library during JCLIN processing, it has no way of knowing what elements are contained in the INDD library. Therefore, a DLIB entry with the same name as the INDD ddname is created. This indicates that the entire library has been copied to the library specified by the OUTDD ddname. If the INDD ddname DLIB entry already exists, SMP/E adds the OUTDD ddname to the DLIB entry. SMP/E records these ddnames in alphabetical order. A DLIB entry can have at most two SYSLIB subentries. If the DLIB entry already has two SYSLIB subentries, SMP/E replaces the second SYSLIB ddname with the ddname specified on the OUTDD operand of the COPY or COPYMOD statement.

Later, whenever a SYSMOD is processed during APPLY or ACCEPT, SMP/E can check the DISTLIB ddname for each of the SYSMOD's elements to determine whether that element belonged to a totally copied distribution library. This is how DLIB entries are used during APPLY and ACCEPT processing:

For ++MOD statements, SMP/E can construct a MOD entry indicating that the module has been copied and that the load module name is the same as the module name. In addition, an LMOD entry can be built, indicating a copy with a SYSLIB equal to the SYSLIB value in the DLIB entry.
For ++MAC, ++MACUPD, ++SRC, ++SRCUPD, and data element message control statements, SMP/E can determine that the element is part of a totally copied library, and can then fill in the SYSLIB field in the appropriate element entry with the SYSLIB ddname saved in the DLIB entry. If there are two SYSLIB ddnames in the DLIB entry, SMP/E uses the second value.

Note: SMP/E treats copies with exclude as total copies. It does not retain information about which elements were excluded during the creation of the initial library.