[Groff] Translating RS/RE to DocBook [was: The case against the case against .EX/.EE & .DS/.DE]

[Michael(tm) Smith]

"Eric S. Raymond" <address@hidden>, 2007-01-03 07:28 -0500:

> Sorry, I am in fact abusing <blockquote> to translate .RS/.RE.  And
> yes, I was aware this was an abuse, but I don't know of any better way
> to translate it.  Can you offer one? 

Nope. I think the case of .RS/.RE exposes the basic problem of
there being some genuine limitations on what a tool like doclifter
can do -- that is, doing what Norm Walsh calls "dragging
unstructued markup uphill". I assume that when most man-page
authors are using .RS/.RE, they're doing it not just for
presentational effect but instead to set apart some block of
content that has a particular semantic meaning -- a program
listing, a example, or whatever.

The only way for doclifter to generate proper markup for .RS/.RE
instances like that would be be for it to figure out what semantic
meaning the author intended for that indented block to have. Is it
a program listing, an example, or what? It sound like you already
have doclifter doing some AI stuff to make that determination in
some cases. But I'm sure there are many cases where it can't
figure out what the author's intention was.

I don't think generating <blockquote> for those cases is the right
thing to do. I guess the only thing I can suggest is that you
generate a, say, remap="RS-RE" attribute/value on whatever block
element you output (<para> or whatever) for the block that the
RS/RE instance contains. And maybe emit a warning: "RS/RE found
but can't determine intened semantics" (or whatever).

  --Mike

-- 
Michael(tm) Smith
http://www.w3.org/People/Smith/