Rectangle vs. Graphic Frame

[Roland Dreger]

Hi all,

does anyone have any idea how to differentiate between these two rectangles using ExtendScript.

 

About the creation process: An XML tag was assigned to the lower rectangle, which was then removed. This way, the appearance in the layer palette can be recreated.

They do not differ in IDML. Thus, the lower element becomes a “normal” rectangle when the IDML file is opened.

Thanks for any hints
Roland

[James Gifford—NitroPress]

I am not sure what your goal or needs are, but ID frames are... malleable entities that shift with the designer's whim. Drag out a "Rectangle"; drag out a rectangular "Frame"; either can be used as a graphics element or tagged as a graphics or text frame by subsequent actions.

Not a Script Guy, but I'd wager that script-level actions are no more differentiated. That is, I'd bet you can create a (graphics) Rectangle and follow up with assigning and formatting text in it, or placing a graphic, and vicey-versey-alleey-roundy.

[Roland Dreger]

Thank you for your answer.  That's right, you can convert it back to its original state. But what I want to find out is how these two states of the rectangle can be distinguished at script level.

Both are rectangle frames, both have (it seems) the same script properties and are the same in IDML. Only one rectangle was tagged once (XML) and the tag removed again.

Background, why I want to find out: one rectangle is tagged when exporting a PDF, the other is not.

[Robert at ID-Tasker]

@Roland Dreger

Maybe because of XML?

What is your end goal? 

[Roland Dreger]

In principle, I just want to identify these specific spline items. I get an unknown document, loop through the items and analyze them. These items that are not “normal” should be listed.

The reason is that they are treated differently when exported to a tagged PDF.

InDesign:

 

PDF

 

 

[Robert at ID-Tasker]

Define "not normal". 

As I've suspected - XML creates a "structure" in your document - tree to your object.

Still - what is your end goal? 

Tag objects that hasn't been "touched" by XML? 

[Roland Dreger]

Define "not normal". 

By @Robert at ID-Tasker

Normal is perhaps the wrong term. But look in the layer palette, there the two rectangles differ in their automatically assigned names. The tag behavior is different when exporting. If you create an IDML and open a document from it, “Rectangle” appears again in the layer palette. So they are obviously different, but not in the script properties. And I think that's strange.

[Robert at ID-Tasker]

@Roland Dreger

OK, but still - what is your end goal? Why do you need to find the "wrong" ones?

What do you have to do with them? 

[Robert at ID-Tasker]

@Roland Dreger 

What options do you have set in the PDF export dialog?

[Robert at ID-Tasker]

@Roland Dreger 

What options do you have set in the PDF export dialog?

Because in print and interactive - I got this from you file:

[Roland Dreger]

No special settings. Only Interactive PDF and “Create PDF with tags”. You get the same result, don't you?

[Robert at ID-Tasker]

No special settings. Only Interactive PDF and “Create PDF with tags”. You get the same result, don't you?

By @Roland Dreger

Looks that way:

I think it's because of XML. Some kind of leftovers after tagging?

[Robert at ID-Tasker]

@Roland Dreger 

But tagging is pulled from ObjectStyle / XML info - objects don't have a "accessiblity tag" property...

https://www.indesignjs.de/extendscriptAPI/indesign-latest/#TagType.html

[James Gifford—NitroPress]

In principle, I just want to identify these specific spline items. I get an unknown document, loop through the items and analyze them. These items that are not “normal” should be listed.

The reason is that they are treated differently when exported to a tagged PDF.

So, I'm still coming away with the impression that this is more intellectual curiosity/tech puzzle than anything relevant to a real-world need. Yes, I understand the mania obsession regulatory-pressured focus on accessibility, and there are lots of grimy cracks and broken corners in PDF export with respect to that... but, simple question:

Does this difference actually result in any variation of accessiblity in an otherwise conforming PDF? Or, is it ever likely to?

Not saying it's not worth pursuing and possibly "fixing" even if it's benign, but there is a difference between need and... curiosity. Or so I think.

[Roland Dreger]

This (I'll call it) bug is certainly a rare case and apparently only occurs with documents that previously contained an XML structure.

But there is also a practical relevance. If this is not dealt with, a check (in this case PAC) results in an error: "Alternative text missing for 'Figure' structure element"

 

But of course, curiosity also plays a role here.

[m1b]

Hi @Roland Dreger, I couldn't find a single thing that differentiated these two frames, based on doing diff between the two dumps in the ExtendScript debugger. Maybe the exported XML is different?

I don't know your use case, but I wondered if the following code might be safe to use:

var doc = app.activeDocument,
    item1 = doc.rectangles[0],
    item2 = doc.rectangles[1];

resetGraphicRectangle(item1);
resetGraphicRectangle(item2);

function resetGraphicRectangle(item) {

    if (
        ContentType.GRAPHIC_TYPE !== item.contentType
        || 0 === item.graphics.length
    )
        return;

    item.contentType = ContentType.UNASSIGNED;
    item.contentType = ContentType.GRAPHIC_TYPE;

};

Edit: fixed typo.

- Mark

[Eugene Tyson]

Good idea @m1b 

Just wonder if the core issue is addressed - rather than resetting.

Maybe some Metadata or Hidden Attributes could be added to the frame. While the obvious properties are identical, this would mean a hidden attributes or metadata (perhaps a label, taggedPDFData, or custom object metadata) that might be accessible by ExtendScript.

I'd try with testing export Tags,  since Roland mentioned differences in PDF tagging, maybe a script could compare export tag settings (perhaps tagged PDF Data or other related properties).

Could a check for Custom Label where untagging elements could leave residual label data or other flags? Perhaps there's a way for a script to loop through all objects and differentiate label properties.

[Roland Dreger]

Hi Mark, hi Eugene,

thank you for your investigation. My use case: Before the PDF export, it should be checked whether a tag (tagged PDF) is exported for an object. I would prefer not to change anything in the document during this check, otherwise it could potentially irritate the users.

But for an action to correct these issues, this is a good idea. Possibly also check whether an XML element is still associated with the object.

Roland

[m1b]

Alas, as you can guess, I'm out of my depth here. I hope someone more knowledgeable about this area of scripting can help. Eugene, I'm not sure I follow your idea, but if you have time, perhaps you could pursue it with Roland's sample file? It sounds intriguing.

[Eugene Tyson]

I can give it a go - I'm not the best scripter but I don't really know the end goal here or what the purpose of it is. 

If there was more info to give that would be great. 

[Roland Dreger]

Here are some screenshots But don't do too much work. It's a special case. And I'd just be interested to know how you can identify these special items.

[Robert at ID-Tasker]

Here are some screenshots But don't do too much work. It's a special case. And I'd just be interested to know how you can identify these special items.

By @Roland Dreger

Can you share some INDD document(s)? 

I could run them through my ID-Tasker tool. 

[m1b]

@Robert at ID-Tasker, there is a document attached to Roland's first post.

[m1b]

Yes @Eugene Tyson, "don't do too much work" is good advice. I believe Roland has more scripting experience than either of us and probably has probably experimented copiously before posting the question here. If we don't know a fix by our prior experience, looking for a solution now will probably double-up work that Roland has already done. (But I don't want to put you off if you want the learning experience, which is usually time well spent.)