7.2-t26-fail-a.pdf: NonStruct can have anything, including a TOC

[faceless2]

• 7.2-t26-fail-a.pdf

This one has a TOC inside a NonStruct. I think that's fine - ISO32K on NonStruct:

"A grouping element having no inherent structural significance; it serves solely for grouping purposes. This type of element differs from a division (structure type Div) in that it shall not be interpreted or exported to other document formats; however, its descendants shall be processed normally."

OK, so we process it's descendants normally but otherwise there are no restrictions. What about TOC?

"(Table of contents item) An individual member of a table of contents. This entry’s children may be any of the following structure types:"

It goes on to list restrictions on its children, but says nothing about its parent. So by the letter of the spec, TOCI in a NonStruct is valid.

[bdoubrov]

This is an interesting case. The same logic would then apply to <LI> outside of <L>, which looks a bit weird.

But we'll raise this question to TWG level to get a clarification for a larger group.

[bdoubrov]

I think in scope of PDF/UA-2 and ISO 32005, this all is clear:

• ISO 32000-2, Table 365 says: "A structure element with the type of NonStruct shall inherit the containment requirements and limitations of its parent element. Where the parent element is itself a structure element of type NonStruct, then the inheritance shall recurse to the first parent element whose type is not NonStruct."
• Per ISO 32005 TOCI cannot be a direct child of Document (as in this test file) if we skip NonStruct

Of course, this wasn't clearly specified in PDF/UA-1 and ISO 32000-1, but I think the intention was always the same.