Omni Systems, Inc.

  

Mif2Go User's Guide, Version 55

  

Valid HTML 4.01!

 

Made with Mif2Go

15 Converting to DITA XML > 15.1 Generating DITA XML with Mif2Go > 15.1.3 Clarifying your purpose for creating DITA output


15.1.3 Clarifying your purpose for creating DITA output

Mif2Go supports two general purposes for creating DITA output from FrameMaker:

Migrate legacy content to DITA XML

Export current content to DITA as needed.

A third potential purpose might be to use DITA as an intermediate step in converting documents from unstructured to structured FrameMaker. You could use Mif2Go to produce DITA XML from your unstructured files, then bring the results back into structured FrameMaker. This should be a lot faster than developing FrameMaker conversion tables.

Migrate legacy content to DITA XML

When you migrate legacy content from FrameMaker to DITA XML, completeness is less important than it would be if you retain source in FrameMaker. After converting your document you edit in an XML environment, so (for example) you can add metadata after conversion. Even validity can be relaxed, if your existing document does not quite measure up. As long as the XML is well formed, you can use XSLT to make adjustments. You can even run XSLT from within Mif2Go, with a command that invokes, for example, Saxon. See §34.4 Executing operating-system commands.

If you are just beginning a proposed migration, you might want to stay with FrameMaker until you are satisfied with the results of converting all your FrameMaker documents to DITA. Keep editing in FrameMaker, and see how the DITA comes out. Make your FrameMaker documents conform to DITA architecture, if you can; you are sure to need specializations, which Mif2Go handles. This way you can continue to produce decent deliverables all the way through the process. At the same time, you can test alternative production workflows with the DITA versions you generate. And if in the end you do not think you can live with the strictures that DITA enforces, you can walk away without having a disaster to clean up.

The simplest route is to go from unstructured FrameMaker direct to DITA version 1.1, then import the DITA files into your next application. There would not be much point to bringing the files back into structured FrameMaker if you do not plan to keep using FrameMaker. In fact, it would be a bad idea, because structured FrameMaker (version 8) currently supports only DITA version 1.0, so you would lose index features, graphics scaling, and so forth. Unstructured FrameMaker does allow DITA version 1.1 features natively, and Mif2Go output to DITA version 1.1 retains those features.

For a list of features added in DITA version 1.1 that would be lost if you went back to DITA version 1.0, see:

http://wiki.oasis-open.org/dita/DITA_1.1_Impact_Assessments

Export current content to DITA as needed

To continue using FrameMaker as source and export content to DITA as needed, you must interpolate into the DITA output any data required by DITA but not needed in FrameMaker. You can use FrameMaker markers or dedicated conditional paragraph formats for file-specific data, and Mif2Go configuration settings for general data items such as book revision level. You do not need XSLT for this purpose. In fact, you should not need XSLT at all, unless your FrameMaker document does not follow the same sequence of items that DITA expects. This is not likely, because DITA pretty much codifies what writers do anyway as good practice.

If you continue using FrameMaker as source, consider two possible ways to proceed:

DITA as an accommodation to others. For this purpose you would want minimum disturbance to FrameMaker files; for example, you would keep multi-topic chapters, and chunk them out using Mif2Go.

DITA as an authoring model. For this purpose you would make your FrameMaker files single-topic. Chapters would take on the role of ditamaps, and would import the topic files as insets.

If you are converting a FrameMaker book, the book file becomes a DITA map either way.

If you create content in DITA XML, then use the DITA Open Toolkit to generate other outputs, you might have difficulty producing some Help formats. Also, PDF output can look ugly. With Mif2Go you can continue to write in FrameMaker, and get a matching DITA set any time you need one. And you can produce DITA output from unstructured as well as from structured FrameMaker.



15 Converting to DITA XML > 15.1 Generating DITA XML with Mif2Go > 15.1.3 Clarifying your purpose for creating DITA output