Question about alt-text in Access-Aide

[DNSB]

Ran 0.9.7 against my 4 ePubs and same results. A quick hex search showed at most one APP1 (0xFFE1) in any of the files so likely why I did not run into the issue.

[oston]

I just came across what seems to be a new addition to InDesign that provides exactly the same functionality to this new addition to Access-Aide. It is in Object Styles under Export Options.
Two new options have been added to the Alt-Text-Source in addition to the three XMP options that were already there: They are:
From XMP: Alt Text (Accessibility) and
From XMP: Extended Description (Accessibility).
I do not know when these options were added.

[KevinH]

Thanks for letting me know. I never use InDesign and so was unaware of what they do or don't do. I tried it early on but the quality of the xhtml code and css code it generated was just too poor for my taste.

[oston]

Yes, Kevin poor css and html. I have to use InDesign because ALL the files I am sent to convert to epub are InDesign files. But now I turn off having indesign generate css files and use my own.
Sigil is my main tool for my post export work.

[oston]

Just ran into a problem while using Access-Aide to fill the alt-text for an ebook that contained 145 images. The plugin worked wonderfully well after using exiftool to populate the AltTextAccessibility field in the image metadata.
But then the author sent a new file of corrected alt-text descriptions, without highlighting the corrections. So we had to repeat the process.
BUT the Access Aide plugin only works if the existing alt="....." is empty.
Is is possible to indicate to the plugin, that it's OK to overwrite any existing alt-text entries?
I did use a regex work around to remove the existing alt-text, but not being a frequent regex user, there were a couple of issues to solve.

[KevinH]

My first thought was that using regex, it would be quite straightforward to find and remove all alt attributes and their contents from img tags. And once you have that regex you can save it as a Saved search so it is easy to find and reuse. Access Aide uses the same approach for contents of title tags inside head tags (ie if you want Access Aide to fill them they must be empty).

What issues did you run into?

[oston]

Originally Posted by KevinH

My first thought was that using regex, it would be quite straightforward to find and remove all alt attributes and their contents from img tags. And once you have that regex you can save it as a Saved search so it is easy to find and reuse. Access Aide uses the same approach for contents of title tags inside head tags (ie if you want Access Aide to fill them they must be empty).

What issues did you run into?

Thanks, Kevin.
I just used this for the search: alt="(.+)" , and while it found all the alt-text , there were other strings that matched the search. So I had to step through the search. Sometimes it's quicker to do that than trying to find the correct regex search.

[Doitsu]

Originally Posted by oston

I just used this for the search: alt="(.+)" , and while it found all the alt-text , there were other strings that matched the search.

alt="(.+)" might match other strings, if the Minimal Match Sigil regex option is deactivated.

The following regex should only match the alt attribute, even if the Minimal Match option is deactivated.

alt="[^"]+"

[oston]

Many thanks for that regex, Doitsu. I've saved it for future use.