DITA - Multiple objects in anchored frame

Report · Mar 19, 2009

Hi,

I'm using the DITA features in FrameMaker 9. I'd like to insert multiple objects in an anchored frame in a DITA topic.

Right now, if I save an image element containing multiple objects, FrameMaker discards any extra objects in the frame. I've also experimented with adding extra graphics elements to the DTD using the attributes listed in the Structure Application Developer Guide. I can save DITA topics containing the custom elements, but they are not added to the XML file though I can see them in the FrameMaker interface. I have not yet experimented with a DITA specialization.

When exporting anchored frames in Structured FrameMaker to XML, FrameMaker bundles multiple objects in a frame into a single CGM file.

Is there a way to export anchored frames to CGM files before saving DITA topics? Is there a plug-in out there that will automatically save the objects in a frame? Can I use read/write rules to define how to treat these frames? How about creating a specialization on the objects element?

I'd like to use the tools in FrameMaker to modify elements in an anchored frame. I don't see why doing this should break the DITA model if these objects are bundled into a single file on export.

Thanks,
- Maura

Report · Mar 19, 2009

Hi Maura,

This behaviour is not directly linked to DITA.

Framemaker's anchored frame with multiple facets (or objects) has no counterpart in XML. THe default behaviour is tht when FM detects multiple facets in an anchored frame, it tries to combine them to a single graphic object in CGM format. The results are usually ugly..

I think it would be a major programming task to add multiple facet handler to export/import XML. It involves XML structure to support layered graphics with rich attribute set to support positioning/stacking etc. For example, MS supports this in OOXML, roundtrip works (though the XML looks horrible), but generating HTML with Word 2007 produces funny results sometimes.

I don't know of any FDK app that would separately produce the CGM's.

BR, Martti

Report · Mar 19, 2009

Hi Maura...

The short answer is .. you can't do that.

DITA knows nothing about "frames" and only allows a single object to be referenced by an image element, so to remain in compliance with the DITA specification the content of a frame is tightly controlled by FM. There's nothing that you'll be able to do via read/write rules or EDD tricks to get FM to write out the data from multiple objects as a CGM .. this is overridden by the internal DITA plugin within Frame. You're correct that in theory FM could merge multiple graphics into a CGM, but this really wouldn't be a good way to operate since the text or other objects would no longer be editable. You can do this now using Illustrator or other applications that let you store the text and objects on separate layers, then save to a single file. Reference this single file in the image element in DITA.

That said .. we are looking into supporting some subset of multiple graphic objects within a frame in DITA-FMx, and this feature may be available in an upcoming beta.

http://leximation.com/dita-fmx/

Cheers,

...scott

Scott Prentice
Leximation, Inc.
www.leximation.com

Report · Mar 20, 2009

Thanks for replying!

We add cross-references to our images in FrameMaker, so numbers placed on a screenshot refer to specific steps in a procedure. We can do the same sort of thing in Illustrator, but we would need to update the numbering in the source file whenever the order of steps in a procedure change.

We don't include screenshots in our html output, so we aren't concerned about compatibility issues.

I guess I was hoping to get the benefits of DITA without losing all of the features in FrameMaker. Oh well, I'm still onboard the DITA bandwagon for now.

- Maura

Report · Mar 20, 2009

Yeah .. unfortunately, the DITA specification doesn't support all of features in FM .. especially when it comes to image handling. We are trying to find ways to support the most popular FM features through DITA-FMx, but it may take some time. It sounds like the limited support for additional graphic objects (text, lines, circles, boxes) that we're planning for the 1.1 release will do what you're looking for.

Cheers,

...scott