The jargon file version 2.3.1 03 JAN 1991 follows in 11 parts

[David Curry]

In article <BZS.91Jan3192010 at world.std.com>, bzs at world.std.com (Barry
Shein) writes:
|>
|>Has anyone tried to put this under the webster server?
       
Just looking at the file, it would take quite a bit of work.  The problem is,
although there is a definite format to the jargon file, it's not nearly as
rigid as the one used by Webster.  Near as I can figure, the following are
true:

	1. Each entry begins flush left.  All lines following the first line
	   of the entry are indented by some amount.  There is a blank line
	   between entries.

	2. "word" entries have <> around them.  "symbol" entries don't.

	3. Most entries have pronunciations, some don't.

	4. Most entries have parts of speech, some don't, especially some
	   of the homonyms (see ACK, for example).

	5. "See also" is just another paragraph in the entries, not a
	   particular format.

Thus, the jargon file is too free-format for the existing webster server to
deal with.

You could do it in one of two ways:

	1. Hack the hell out of the webster server to understand the jargon
	   file format.  If nothing else, you'd have to do this for the
	   pronunciation part, since the code tries to do clever things to
	   print it out, and the jargon file uses different characters.

	2. Hack the hell out of the jargon file to impose the Webster format
	   on it.  This actually wouldn't be that bad an idea, since it would
	   certainly make the file more consistent.  On the other hand, the
	   format is pretty useless for just plain old persuing the file.

It'd probably be easier to just rewrite the webster server from scratch
to handle
the jargon file in its present format.

Dave Curry
(Author of the UNIX Webster server.)