Rob,
OK, now I have my head back in the right place this is how I
understand the situation.
Makeindex makes special use of the characters:
! ^ |
Index entries using these should be escaped by "
(and also " should be escaped).
When I did texpdf I tried to work around the fact that this
was not happening, just covering the cases that actually
mattered to me.
It looks like I tried to escape single occurrences and
reduce multiple occurrences so that it would be possible to
use @@ instead of @. And it looks like I only did this for
the @. I got something that seemed to work OK for my
documents, but I knew it wasn't complete and that it wasn't
the right way to do it.
The right solution presumably is for the handling of the
index brackets to do the escaping, and then Makeindex can be
used.
The benefits are substantial, not just that you get the full
range of indexing facilities (which I do care about even
though I don't use them at present in ProofPower documents)
but also because it eliminates duplicate entries in the
index (compressing them into one entry). This can
substantially shorten an index and has a significant impact
on how long it takes to find an entry.
This is important enough that I will dig in and try to do it
myself, if you will assure me that you are willing to put
this into ProofPower. However, it is pretty trifling and you
could probably do it yourself faster than reassuring
yourself that what I've done is OK.
The other issue is what scripts there should be.
Should we just use docdvi even if we want PDF?
Should there be two names for the same script?
Roger Jones
_______________________________________________
Proofpower mailing list
Proofpower@lemma-one.com
http://lemma-one.com/mailman/listinfo/proofpower_lemma-one.com
