Command Reference > Working with pmrep Files > Using the Persistent Input File

Using the Persistent Input File

When you run pmrep with some tasks, use a persistent input file to specify repository objects that you want to process. The persistent input file represents objects already in the repository. You can create a persistent input file manually or by using pmrep.
Use a persistent input file with the following pmrep commands:
The persistent input file uses the following format:
encoded ID, foldername, object_name, object_type, object_subtype, version_number, reusable|non-reusable

Creating a Persistent Input File with pmrep

You can create a persistent input file using the pmrep ExecuteQuery, Validate, or ListObjectDependencies commands. These commands create files that contain a list of objects with encoded IDs and a cyclic redundancy check (CRC) value. It also contains an encrypted repository GID. This ID identifies which repository the record comes from.
The pmrep commands that use a persistent input file get object information from the encoded IDs. The encoded IDs enable pmrep to process the input file quickly.
When you create a persistent input file with pmrep, it creates the file in the pmrep installation directory. You can specify a different path.
The following text shows a sample persistent input file:


You can use the ExecuteQuery command to create a persistent input file of objects to process in another pmrep command. For example, you want to export all logically deleted objects from the repository. You might create a query called find_deleted_objects. When you run the query with pmrep, as shown here, it finds all the deleted objects in the repository and outputs the results to a persistent input file:
ExecuteQuery -q find_deleted_objects -t private -u deletes_workfile
You can then use deletes_workfile as the persistent input file to ObjectExport:
ObjectExport -i deletes_workfile -u exported_del_file
ObjectExport exports all the referenced objects to an XML file called exported_del_file.

Creating a Persistent Input File Manually

If you want to run pmrep commands against a set of objects that you cannot identify through commands such as ExecuteQuery, you can manually create an input file.
Use the following rules and guidelines when you create a persistent input file:


The following example shows a manually created persistent input file:
In the first record, CustTgt is a target definition. Targets have no subtype, so you enter “none” for the object_subtype argument. A target cannot be reusable or non-reusable, so you drop the reusable argument. Note that the record has six arguments instead of seven.
In the second record, S_Orders is a session. Sessions have no subtype, so you leave the argument blank.
In the third record, you want the latest version of the transformation, so you enter “LATEST” for the version_number argument.