Editors have found it efficient to assimilate data in two main stages. The first is tabularization. In this stage, you produce a plain-text, tab-delimited file. We call it a tabular file. It provides the input into the second stage, serialization.

The following pages provide an overview of various aspect of the tabularization process:

In rare cases, you may decide to skip tabularization. This is most likely to be appropriate with a source that is already tabular and can be directly serialized, or with a source whose entries vary in structure.