CONVERSION OVERVIEW#

The Conversion Application is made up of several forms, reports, functions and procedures. All of these programs have been created with the same development tools as the software application. As a result, the Conversion application has a similar look and feel as the software. All the functions to import, manipulate and convert the data from comma-delimited flat files can be performed from within this application.

One of the other effects that this has on converting data into the software application is that the conversion programs cannot read the data directly from flat files stored on the server. All records that are to be loaded into the software data structure must first be loaded into interface tables within the Oracle database itself, before being moved to the tables owned by P2K. Once the information has been loaded in these interface tables it can be edited or reported with the supplied screens and reports. Both the screens and report can be accessed from within the menu structure designed into the Conversion application.

Loading the software conversion data into the interface tables gives us the ability to verify that the loaded data is correct and in the format that the conversion programs require, prior to altering the Wiki tables. This also allows us to perform any additional translation of the data that could not be performed prior to loading the information into the tables. Several third party applications are capable of reading, manipulating and storing the altered data back into Oracle tables. The use of SQL scripts to perform the same translation of the data is also available, and can be done with the utilities provided by Oracle.

The Wiki screens allow the users to key in all of the client data into the application in a structured format. The records need to be entered into the screen in a specific order and the data builds onto one another as it goes. If the quantity of records to be entered into some of the tables is small, keying them in will actually be faster that loading them in. However, there are a number of tables that need to have the records loaded into them in bulk.

The purpose of these conversion utilities is to provide this bulk load mechanism for the Wiki application. The records will still have to be loaded in the table in the same order that they would have to be keyed in. This is actually dictated by the data integrity constraints defined in the database, as part of the application configuration.

During the initial installation of the application environment, three separate database where created. One of these databases will be called CONV or PROD, depending on the specific installation of Oracle and the Wiki application.

This database is the one used for all the conversion processes. However, this database does not have any data in it. This database is usually created as a shell only. The following is the sequence of events that is followed to perform the conversion:

1. Install Oracle components needed for conversion on the client PC.
2. Move a copy of the PILOT data into the database used for conversion
3. Drop the test employees generate in the PILOT database from conversion.
4. Extract data from the legacy system into pre-defined, comma delimited data files (CSV files). Note; this step is only discussed for High Line products.
5. Load these files into interface tables within the target database
6. Make changes to the load process to format the data as it is loaded to match Wiki needs
7. Drop the contents of the INF table and repeat step 5, above.
8. Apply any scripts needed to fix the data prior to converting into the Wiki tables.
9. Run the stored procedures to convert, or move, the records from the interface tables into the P2K tables.
10. Verify the data loaded into the target database through screens and reports of the application and custom reports (I.E. Discoverer). If needed, drop the contents of the specific Wiki table and reload the data (step 5).
11. As changes continue to be made in the PILOT that will affect the conversion process we will need to drop the contents of the target database and start over again at step 2, above.
12. Once all the table conversion has been tested, the whole process is repeated in the production database and parallel payroll processing is started.

Installing the Oracle components and configuring the client for running the Conversion application is done once, for each computer or user that will participate in the conversion project. Typically, this is only one user on one computer. The remaining steps will be repeated several times until the process is working for all the tables that will be converted.