Manual and batch-conversion processes

An important decision in data-conversion planning is to decide whether to use a manual or batch process for converting characters.

Manual approach

If the number of required changes is limited, a manual approach might be practical. The output report generated by the codepageutil analyze_tables subcommand lists the location of every character in the database that is not supported by the preferred data code page.

Depending on the tables where the unsupported characters appear, you can change the characters by using the Designer, the Maintenance Tool, HCL Compass client applications, or editing tools from your database vendor.

A manual approach provides a great deal of control over the changes and ensures that each one is reviewed so that the converted information makes sense to users. A disadvantage to this approach is that it is labor-intensive, prone to user error, and not practical if a large number of changes must be made.

Batch method

If the number of changes is large, a batch approach might be more effective.

Your database vendor might already have provided this capability (for example, the Oracle Character Set Scanner utility). Consult your vendor database documentation. Otherwise, you can write a batch-conversion program.

Running the codepageutil analyze_tables subcommand with the -xml option generates an XML file that shows the location of every unsupported character. You can use this file can as input to your batch-conversion program.

After using a batch-conversion process to convert characters, it is a good practice to review a few text fields such as the Description, Notes, and Headline to verify that the conversions are meaningful.