earlier replies
Replying to @aaronpk @isaac32767
PHP for conditional rendering of content for the epub vs print editions, e.g. "on page X" vs hyperlink, and to get images to flow properly
1
Instead of jumping through hoops to generate the XML, consider authoring in XML in the first place.
1
I'm not going to maintain individual XML files for the ncx/opf files. Those need to be generated automatically from my content.
1
I'm not talking about those files. I'm talking about the content files you generate them *from* 1/2
1
There are a lot of off-the-shelf workflows for converting XML to any user doc format you could name. So if you *start* in XML ... 2/2
1
I'm starting in HTML which is what Prince uses to generate the print PDF. That part works great.
1
In other words, you're using a legacy workflow. Moving on from it may not be practical, but you should at least consider it.
1
What's legacy about that? It's an actively developed tool, and I was following their rec's: princexml.com
1
I suppose "legacy" can be a subjective word. Most of us consider FORTRAN a legacy language, but a lot of scientists still use it.
1
Or maybe I'm just mis-stating the argument, which might be really between advocates of structure and those who just want to hack out content
2
Replying to @isaac32767
There's plenty of structure in these HTML files, that's how the ToC is generated.

Nov 1, 2017 · 7:36 PM UTC

1
Replying to @aaronpk
Not the kind of structure that makes a document base maintainable and reusable. Having written huge amounts of HTML content, I know whereof