Re: [Groff] Extending refer

[Pierre-Jean]

"Bernd Warken" wrote:

> 1) `refer.man' cites under `FILES':
> file.i                   Index files.

I did not investigate this much, but this should be the
conventional output file name of the 'mkey' program from
Heirloom/refer, which build an index of keys of a refer
file.

Another program, 'inv' make an inverted index from the mkey
output and produce a set of files: file.ia, file.ib,
file.ic, file.id.

Then, the 'hunt' program search and deliver items from the
inverted index.

For all these programms, the documentation is "Application
of inverted indexes in the Unix system" by M. E. Lesk. Based
on this document, I have made three manual pages, which
might be usefull:
http://utroff.org/man/mkey.html
http://utroff.org/man/inv.html
http://utroff.org/man/hunt.html


> 2) `refer.man' cites under `FILES':
> /usr/dict/papers/Ind     Default database.

I personnaly use refer a lot. And even if I copy all my
references in a central database, I find more usefull to
create a new database for each paper.

That said, it would be nice to share a refer database of the
various troff (and maybe unix) papers, which are sometimes
mentionned on man pages.

> 4) The programs in Heirloom and refer history
> - addbib

I find easier to add papers manually.

> - roffbib

This one can be replaced easilly by
refer -B file.ref | troff -ms

> - sortbib

Since refer is able to sort, I don't use this one a lot.

Cheers,

Pierre-Jean.