Re: [Boost-docs] HTML to PDF translations?

Subject: Re: [Boost-docs] HTML to PDF translations?
From: Daniel James (daniel_james_at_[hidden])
Date: 2009-02-11 22:23:39

2009/2/11 John Maddock <john_at_[hidden]>:
>
> I think this would be hard: I toyed with the idea of a quick and dirty HTML
> to DocBook converter so we could do just this, but one of the main issues is
> structure - or rather the lack of it in html documentation - there's nothing
> to tell you in what order to place the hmtl "pages".  Also things like
> manually written tables of contents don't work half as well in PDF format as
> a good PDF chapter/section outline.

It's not that hard to write a script to scrape the structure if the
markup's good enough. Joaquin's documentation looks like it would be
quite easy to deal with. This could then be used to create a single
html file, which could be converted into a pdf using something like
prince - which can work out the contents based on the headers. With a
bit of work the results can be reasonable.

> I had a quick experiment with converting the reST docs into PDF's, and these
> seem to suffer from the same issues - one reST document is one HTML page,
> with hyperlinks between them hardcoded as links to the generated html, and
> no way to tell what order to place them in inside a single PDF :-(
>
> So... sadly I don't think a fully automated solution is possible,

That's a pity. I had assumed from the pdf of the mpl reference
documentation that it might be. It seems to be using a python script
on top of docutils. One possibility is to use sphinx
(http://sphinx.pocoo.org/), although I haven't tried it.

Daniel