Conversion To DITA – It Is Not An OverNight Event

As a content author, you’ve heard a lot about the benefits of migrating your unstructured content (Word, unstructured FrameMaker, etc.) to a structured format such as DITA. A key benefit to migration is that you can efficiently reorganize and reformat your content for multiple audiences, formats, and devices.

But migration is not just a matter of purchasing a software application and then clicking a “Convert” button. In my experience, migration is not a trivial effort. We know structured content is the end result, but we need to think about the entire transition process that migration requires.

Some organizations try to migrate all the content at once, which usually means stopping work on existing projects until completed. Some companies try to put together a massive pilot project and expect it to work perfectly in a short period of time. Neither of these approaches is realistic, which can doom migration to failure or even scare an organization from trying at all.

Successful migration projects are often broken into manageable chunks. This lowers the level of risk to the organization since they can learn from their mistakes and encounter delays without catastrophic results. However, this means that legacy unstructured content and new structured content will need to exist at the same time, which brings up an obvious problem: how do you mix both unstructured and structured content?

The solution is to use FrameMaker’s structured authoring mode. I don’t know of any other publishing tool that can mix and structured and unstructured documents. With FrameMaker, you can break down your migration project into bite-sized pieces. This is very reassuring to your managers, who like to minimize risk. To enable FrameMaker to understand both unstructured and structured documents: select File > Preferences > General, then change the “Product Interface” to be “Structured FrameMaker”.

I recommend the following migration workflow from unstructured to structured content:


migration workflow
Migration Workflow


You can tell FrameMaker to analyze the unstructured documents and automatically create a conversion table, which will be used to convert the unstructured content to structured content:

 Initial Conversion Table

This is how to generate the initial conversion table:

Initial Conversion Table

 Depending on how well the unstructured documents follow the style rules, FrameMaker can automatically generate up to 90 – 100% of the conversion. You can if necessary create XSLT to finish the conversion that FrameMaker cannot do automatically. If no one on your team wants to learn how to create XSLT, you can always have a conversion specialist consultant write them for you.

Once FrameMaker uses the conversion table to generate the structured documents, you can save as XML. FrameMaker can apply XSLT on save, if necessary to further modify your structured output. You can tell FrameMaker to publish the content to the desired type of output. As part of FrameMaker’s book feature, you can have any mixture of unstructured, structured and new structured XML documents, letting you publish your content at any point in the migration. This is a key ability that FrameMaker has to make your migration process manageable.

Once you have successfully completed the pilot migration, you can convert the next phase of documents. Although you will probably find additional issues that require that you modify the XSLT and the conversion table, your pilot project will probably have taken into account a large portion of the migration issues. Also, the pilot will form a strong foundation for your next migration phases.

By following these steps, FrameMaker will give your migration a much higher probability of success.

For more details on converting unstructured to structured content, watch the seven-part e-seminar “Unstructured to XML Workflow Series”:

Tom Aldous
Product Evangelist – Technical Communication Suite, FrameMaker and RoboHelp