The conversion utility is provided as a REXX exec on the ISPF product tape.

The ISPDTLC exec can reside in a CLIST data set allocated to SYSPROC or in an EXEC data set allocated to SYSEXEC. For more information about the use of REXX execs on MVS™, refer to the z/OS TSO/E REXX User's Guide.

Additional Requirements:

• All data sets must be allocated before running the conversion utility. In addition, the conversion utility uses ISPF services to produce command table and key list output, which means that a partitioned data set must be allocated to ISPTABL. See the topic on allocating ISPF libraries in the z/OS ISPF User's Guide Vol I for more information.
• To allow the user to specify the source and destination data sets when using the conversion utility syntax, seven ddnames have been reserved in an allocation profile with associated data set names to be provided by the user.
  Note: ISPDTLC profiles from previous releases can be used without change. However, a warning message is issued if the DTLMIN or DTLNLS ddname records are encountered.
• A sample profile member ISPDTLP is included in the ISPSLIB skeleton library. You can modify the data set names for installation or user use. DTL format comments (<!--comment text--> or (<:--comment text-->) can be used in the profile data set or member. Do not modify the DDNAMEs in this table (column one). A sample user updated profile member follows:

  DDNAME

  Data Set

  DTLGML

  any.GML.input

  DTLPAN

  your.panel.output

  DTLMSG

  your.msg.output

  DTLLOG (*)

  your.log.output

  DTLLIST (*)

  your.list.output

  DTLSCR

  your.script.output

  DTLTAB

  your.table.output

  (*) The sequential data set name associated with the DTLLOG and DTLLIST ddnames should have the same characteristics and attributes as the LOG and LIST data sets for ISPF.

  DTLGML is the input file to the conversion utility. The last 6 files are for output and are usually the user's own data sets.

• For compatibility with previous ISPDTLC releases, the user can provide the allocation profile name on invocation:

  ISPDTLC source-filespec (disk PROFILE=User.profile

  The data set name following the PROFILE keyword must be a fully qualified data set name. When specifying the data set name, do not include quotes.

  The profile data-set-name can specify either a sequential or a partitioned data set. If the profile entry is part of a partitioned data set, then the member name must be included in the data-set-name specification. The profile data set and all data sets defined within the profile must be preallocated.

  The profile can contain multiple entries for each ddname. For output files, the first valid data set name in the profile is used. For the input GML file, each data set is checked in the order they are found in the profile for the member name specified. The first match by member name is used as the file to be converted.

  When the data set associated with either the DTLLOG or DTLLIST ddname in the profile is a PDS, the member name may be a single asterisk. When the asterisk notation is present, the conversion utility uses the same name for the log or list file as the source GML member name.