FM10 and S1000D query

You have asked a complex question in a very few words. So let’s take a look at what you are really asking.

You have a document in ATA iSpec 2200 and you wish to convert that to S1000D. First off, S1000D is not ATA iSpec compliant in content, format, or presentation. So you have to decide if you are going to be S1000D compliant, or simply use S1000D elements to recreate what you already have. In which case, you will not be S1000D compliant; at least not in format/presentation. Additionally, because S1000D does not support things like MTOSS codes and pageblock numbering, you will either have to abandon those concepts or develop a methodology to determine/create them.

The next consideration is what you plan to do with the data once you have it in S1000D. Do you only require generating a PDF/paper that matches the ATA iSpec output or do you need to output a PDF/paper in S1000D default format? Or will you exchange the xml data itself and have no requirement for PDF/paper? Maybe all of the above? Serious questions to address before beginning. Are we talking one document or do you have a library of documents to consider?

The ATA group responsible for this type of data is in the process of defining the Component Maintenance Publication (CMP) requirements for S1000D. This is the requirements and guidance rules that tells data developers (like yourself) how to create the equivalent industry accepted S1000D data that is equivalent to an ATA iSpec 2200 CMM.

I have just touched on a few of the very high-level items to consider. A complete understanding of your requirements and end goals is really required. Simply converting a document to S1000D may not meet your needs.

Please feel free to reach out directly if you need additional information.

I'm afraid that the reply to your question is not short .

What you are asking is something that we are asked from time to time and unfortunately the answer is far from straightforward. iSpec2200 is, as you are aware, based on PageBlocks which contain tasks which may be authored in several different ways. I know that iSpec says that you should only use one Pageblock type in a CMM but we know, because we also supply an iSpec2200 compliant CMM package, that our client's clients have in one instance specified that you can have different Task Styles in the same page block.

No conversion can be 100% and there will always be reworking required - how much of course depends on how consistent the source data is. Also, where conversion is concerned, it is often necessary for the entire publication to be re-verified to ensure FAA/CAA approval. Parts of iSpec do not directly convert into S1000D.

If the document has been created using unstructured FrameMaker where there has been a use of Paragraph and Character styles then it is possible that the built in FrameMaker conversion table feature could be used. This is a very useful feature which has the potential of providing a good start in getting the data into S1000D. I note that you have the Mekon Eclipse S1000D plug-in and we certainly do have clients who are using conversion table feature to quickly get the source document into S1000D. Of course the quality of the conversion at this stage depends on how good the conversion table definitions are and how consistent the source documents are. But because there is the mismatch between S1000D and iSpec2200 there will be quite a lot of remedial work involved - although FrameMaker is quite forgiving in some respects in that it can automatically do some correction for you.

If the source document is in Structured FrameMaker the question is "Is the structure in accordance with the iSpec2200 DTD?"  It is possible, of course, to write a transformation between the two structures but the same problem still exists and a fair amount of rework will be required because of the mismatch. If the MTOSS codes are used it may be possible to extract these to help in the population of the S1000D Data Module Code.

Whichever route is taken a good start is to allocate sufficient time to do a document analysis of the source document and decide where the source data is to be located in the S1000D Data Modules. Also relevant it a decision on what Data Module types are going to be used as this obviously makes a difference to where information is going to be located in the target.

So the bottom line is that this conversion is not easy and whichever route is taken be prepared for remedial work and allocating someone to the job who is familiar with both specifications so that proper decisions can be made about the conversion.

Have a look over in the Structured FM forum?