Re: [PATCH] [mom]: Simplify `pdfhref` calls.

"G. Branden Robinson" <[email protected]> Fri, 16 Jan 2026 10:35:59 -0600

Newsgroups	gmane.comp.printing.groff.general
Message-ID	<20260116163559.j2mzyn7btf36nhyx@illithid>

Hi Deri,

At 2026-01-16T16:18:37+0000, Deri wrote:
> On Friday, 16 January 2026 04:11:42 GMT G. Branden Robinson wrote:
> > I noticed that the `-E` argument to a `.pdfhref L` call doesn't seem
> > to do anything.  I compared before-and-after, and while the PDF
> > checksums change, the file sizes do not.  I assume there is a shifty
> > data structure (like a hash table) or an embedded date stamp in the
> > PDF that changes.  I'd have more to report, but _pdfmom_ doesn't
> > support groff's `-Z` option.
> > 
> > Hyperlinks of various types, both internal (whether from tables of
> > contents or within body text) and external still work fine, and
> > appear to have correct link text.
> > 
> > Ok to push?
> 
> The -E flag is only documented to be used with .pdfhref M, so you are
> correct.

Ah, cool.  Peter, is it okay with you if I push?  It's important to me
to have sound specimens of fancy PDF feature usage in our release;
adventurous users are bound to go looking at "how we did that".

> Pdfmom does honour both -z (useful for examining forward references)
> and -Z.

I must have screwed up.  I'll try again more carefully next time.

> "test-pdfmom" is the same as test-groff but the last line is changed
> to call pdfmom rather than groff. I have a vague recollection of
> someone suggesting we use a variable GROFF_BIN in test-groff to select
> which executable we want to run in the test harness. (I may have
> dreamt this!).

It might be a dream you share(d) with Bjarni.  :)

Does the following ring a bell?

commit af13618a30cb04ce35eb122e70b716f217e7bf06
Author: G. Branden Robinson <[email protected]>
Date:   Thu Oct 23 14:16:01 2025 -0500

    doc/doc.am: Build PDF documents more carefully.

    * doc/doc.am (DOC_PDFMOM): Add environment variable `GROFF_COMMAND`,
      newly recognized by pdfmom(1), to make it run the build tree's groff
      via "test-groff", rather than (potentially) a "groff" in the build
      environment's $PATH.

      (PROCESSEDFILES_DEPS_PDF): Add "test-groff" script.

I think, though, that the foregoing might be one layer of wrappers
deeper than what you're talking about.  (Even now I can see Ingo
pinching the bridge of his nose.)

Regards,
Branden

signature.asc (application/pgp-signature, 833 B)