Try to heuristically break up non-Tagged-PDF content into separate objects
Submitted by Joanmarie Diggs @joanmarie
Link to original bug (#701816)
Description
Once we have support in Evince for Tagged PDFs (bug 701749), we can exposed tagged elements to assistive technologies (bug 701814).
In the case of non-Tagged PDFs, it would be nice if we could find some heuristic to break up the giant text blob that is the accessible document view into separate objects (i.e. paragraphs). This may not be reasonably doable, but we should at least investigate the possibility. If in the end it is not reasonably doable, we can close this out as a WONTFIX.
Version: git master