PDFUA-246: Order of content within a marked content sequence incorrectly set

[zakkinsey]

Jira issue originally created by user @BPeboeck:

This example demonstrates a failure condition in which the order of content within a marked content sequence does not match the author’s intent.

Often, when text is rotated, the content order within the marked content sequence is incorrect, like in this example. Even though English is written top to bottom and left to right, {color:#202124}the text in this example should be ordered strictly left to right.{color}

[zakkinsey]

Comment created by @PaulRayius:

The PDF/UA identifier should be turned off on fail cases, right? 

[zakkinsey]

Comment created by @PaulRayius:

Metadata:  For PDF/UA, I did not select 01-006, even though we typically do, because the use case in question in this example is not about the tag types used but only the order of the content.  While that doesn't match 100% with 09-001, that's the closest we have.  For WCAG, similarly, I did not use 1.3.1 because I didn't view this as incorrect "information" or "structure" but I did call out 1.3.2, "Meaningful Sequence" because the order of the content affects its meaning. 

[zakkinsey]

Comment created by @PaulRayius:

Testing against Acrobat's Accessibility Checker, I got a failure for the Tab Order.  But that should only apply to pages with Annotations and there aren't any annotations in this file. 

[zakkinsey]

Comment created by @unknown-jira-user-8:

The PDF/UA identifier should be turned off.

[zakkinsey]

Comment created by @unknown-jira-user-2:

Acrobat's Accessibility Checker still reports Tab Order failure.

[zakkinsey]

Comment created by @unknown-jira-user-8:

I've changed the Tabs key in the page object to S(structure order). It seems that Acrobat's Accessibility Checker requires a Tabs key even for pages without annotations.