groff
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: groff_man subsections in HTML: indentation and TOC


From: G. Branden Robinson
Subject: Re: groff_man subsections in HTML: indentation and TOC
Date: Sun, 18 Apr 2021 01:23:02 +1000
User-agent: NeoMutt/20180716

Hi, Hans!

At 2021-04-17T12:09:00+0200, Hans Unzner wrote:
> Thanks very much, Branden!
> This change also creates <h3></h3> HTML-tags already. So there is no
> need to change the HTML output driver at this point.

Right--not in this respect, at least.

> Does this change go into the official release or do I have to patch it
> always on my system?

It's on groff Git HEAD, which means it is slated for the next release
(candidate).  If the change meets with strong objections, or if I find
something horribly wrong with it that my testing yesterday did not
reveal, it can be reverted before groff 1.23.0.

> Another thing that shows up with that extension is the indentation of
> the subsections in the TOC.

A lack of indentation, I think you mean.

> First see this example:
> https://hansu.github.io/linuxcnc-doc/man-new-ss/man1/iov2.1.html
> If I want to indent the subsections in the TOC -- is there a
> similiarly easy way to modify the HTML output driver or does it have
> to be compiled from the sources?

Not as far as I know.  I checked the device-independent output, and it
does _not_ include all of the headings at the beginning of the document.
(That's good design--such redundancy is not needed.)  Therefore, the TOC
must be constructed by the HTML output driver itself (more precisely,
the post-grohtml command) from the "NH" devtags.

I understand little of post-grohtml.  I have a vague suspicion that
groff's HTML output, particularly for man pages, would be a lot more
popular if we made it easy to point the generated HTML at a
user-specified stylesheet[1].  The mechanism for this could be a
postprocessor option, as already widely used.

That's said, it's unwise to prejudice oneself much with design notions
before the problem is fully understood.  If the have time to opine, I
expect that Werner Lemberg and John Gardner would be helpful in sorting
out these matters.

Regards,
Branden

[1] Even more important, though, is probably real table support;
<https://savannah.gnu.org/bugs/?60052>.

Attachment: signature.asc
Description: PGP signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]