• Paul D Anderson <paul.d.anderson.removethis comcast.andthis.net> Mar 28 2008
• BCS <BCS pathlink.com> Mar 28 2008

Paul D Anderson <paul.d.anderson.removethis comcast.andthis.net> writes:

BCS Wrote:
 
 As for part 1, the grammar is incomplete and wrong, but only in a few 
 tiny places. I'm working on fixing them and am going to submit a patch 
 to Walter in a few days. This is a result of something connected with 
 part 2; I have a sed script that extracts the grammar from the HTML and 
 converts it to EBNF (or a dialect there of) I'll release it if anyone is 
 interested.
 


Thanks. I look forward to it.

Paul

BCS <BCS pathlink.com> writes:

Paul D Anderson wrote:
 BCS Wrote:
 
As for part 1, the grammar is incomplete and wrong, but only in a few 
tiny places. I'm working on fixing them and am going to submit a patch 
to Walter in a few days. This is a result of something connected with 
part 2; I have a sed script that extracts the grammar from the HTML and 
converts it to EBNF (or a dialect there of) I'll release it if anyone is 
interested.


 
 Thanks. I look forward to it.
 
 Paul


It's posted:

http://www.dsource.org/projects/scrapple/browser/trunk/dparser

the relevant files are: "cut_gram2" and and "gram.sed".

"cut_gram2" is a bash script that runs the sed script and does a little 
filtering. Call it from the directory containing the .html docs or pass 
it the directory as the first arg.

The output will end up in "gram.inp" and isn't quite right (formatting 
errors and whatnot) I have another program that will convert it to 
something usable (and also factor out left recursions) but it's not 
ready for posting.

p.s. It is intended to be run on the doc as patched by me. With any 
luck, Walter will get my edits in post haste.