Convert bare URLs to active links in a HTML to PDF conversion

[jdlail]

I am trying to convert some HTML files into PDFs.

Unfortunately, the HTML files have bare URLs in the body. For most URLs this is not an issue, they get automagically converted. But URLs that contain a colon (or multiple colons) are either ignored or the active URL is only up to the Colon.

Example URL:

https://www.familysearch.org/ark:/61903/1:1:6ZSL-PHGC

I'm not a programmer and I was wondering what would be the best approach to rewriting any bare URLs to properly formed HTML? Could a Lua filter do this?

In the AST output, all of the content is in a single RawBlock except for the data-meta tag which is passed from the command line.

[jgm]

Colons are legal in the path part of a URL. You could try replacing the : with %3A, either manually in the source or using a Lua filter.

But I don't really understand your question. Are you using pandoc to convert these HTML files to PDF? With what command exactly? Your last comment really confuses me. I think we need more context.

[jdlail]

Thanks, John. I'm sorry if I wasn't clear.

Here is a simplified version of my command and my source and output files are attached.

pandoc new-test.html --pdf-engine=weasyprint -s -o new-test.pdf

I renamed the HTML file as .txt to upload it.

When I double click on the one link in the PDF, it does not include the entire URL. The link does not have a soft or hard return in the HTML file.

Is this a pandoc issue or weasyprint issue or something else?

new-test.pdf
new-test.txt

[jgm]

As you can see if you use -f html -t native, pandoc doesn't parse anything here as a link -- as one would expect, since there are no <a> elements in the HTML. Perhaps WeasyPrint has some code that automatically linkifies URLs; if so, it is that code that is at fault. This has nothing to do with pandoc.