WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: PDFs and NVDA


From: Duff Johnson
Date: Aug 21, 2011 3:45PM


First, Bevi is correct in her comments.

I'm going to make a couple of similarly pedantic-seeming points :-)

> -If most people use the tagged reading order, what are the uses for the
> other reading orders in Acrobat/Reader?

It's not that "most people" use the tagged reading order... it's that tags are the ONLY place for logical structure, which is vital to the reliable reading and navigation of a PDF using AT. That said, many (most) applications OTHER than Adobe's are unable (at this time) to create or process tags in a PDF. Microsoft Office supports tagged PDF generation from Word files, and so does Open Office.

> -If I don't use proper headers in a Word doc (which will be made into the
> PDF), the tagging can't tell what is a header, correct?

Key point: The tags COME FROM the headings (not "headers", per Bevi). Software that supports creation of accessible PDFs from the source file will convert the structures present in the source file into corresponding PDF tags. So, exactly as you surmise... use good structures in your source file, and you have the right to expect a matching structure in your PDF.

> It just guesses? To
> me with my testing, it seems like it guesses and I think that is what you
> said below, but just wanted to confirm.

PDFs generated by software that does not support tagged PDF for output may be handled for accessibility purposes in two basic ways.

1) The "raw print stream" (PDF name is "reading order" - see earlier threads). In practice, this is what most software reads instead of the tags. The results range from "so-so" on extremely simple documents to "chaotic" in other cases.

2) Some interpretation of the raw print stream using other information (page-position, font size, font choice, etc, etc).

It should be noted that both of the above approaches are totally illegitimate with respect to accessibility. There is only one option that can comply with WCAG 2.0 - support for tagged PDF in all its particulars. Yes... some users get good use from software that (primarily) uses reading order (cell phones, for example), but that's simply an example of something being better than nothing.

> -The touchup reading order in Acrobat can be used to tweak what the tagging
> does?

That tool tweaks BOTH the raw print stream order (content order) AND the tag order. While this is not an unreasonable thing to do in the right circumstances, execution is everything, and I do not recommend the use of this tool in its current manifestation for this purpose.

> I haven't used that one a lot before...I usually try to make well
> structured documents in Word and they usually come out okay, but maybe I
> should be using that touchup tool..

The first part... focus on getting the source file as good as possible so that tweaking in Acrobat or with other tools is minimized... is exactly correct.

Otherwise... all the fixup following file creation can generally be done in the tags panel. Some of the Touch Up Reading Order Tool elements can be handy, but it's not a great tool, sorry to say.

Duff Johnson

p +1.617.283.4226
t http://www.twitter.com/duffjohnson
w http://www.duff-johnson.com