[
https://issues.apache.org/jira/browse/TIKA-3026?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17020513#comment-17020513
]
Tim Allison commented on TIKA-3026:
-----------------------------------
I trust that this will be far more complex that I currently imagine, but I was
able to extract some structure from some of our current test files:
[https://twitter.com/_tallison/status/1219698912333746176?s=20]
This could be a game changer for the cases when PDFs are actually tagged
correctly.
> Consider extracting structure/tags where possible in PDFs with the
> PDFMarkedContentExtractor
> --------------------------------------------------------------------------------------------
>
> Key: TIKA-3026
> URL: https://issues.apache.org/jira/browse/TIKA-3026
> Project: Tika
> Issue Type: Task
> Reporter: Tim Allison
> Priority: Major
>
> Some PDFs contain tags that _may_ be useful in understanding the structure of
> the elements within a PDF, e.g. table markup, paragraph breaks, headers, etc.
>
>
>
> The quality of the tags depends entirely on the software and human generating
> the PDF. There are no guarantees. Nevertheless, it might be useful in some
> cases for users to be able to extract content with structure tags.
>
> Some references:
> [https://acrobatusers.com/tutorials/what-are-pdf-tags-and-why-should-i-care/]
> [https://www.adobe.com/accessibility/products/acrobat/pdf-repair-add-tags.html]
> [https://www.pdfa.org/resource/tagged-pdf-best-practice-guide-syntax/]
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
