ftp://info.cern.ch/pub/incoming/libHTML-930121.tar.Z
There's a lot of work left to be done, but the architecture
I've come up with looks quite workable. For example, you can
do and of these with the same basic mechanism:
% sgml_test HTML MIF <some_html_file >mif_equivalent
% sgml_test plaintext HTML <any_old_file >html_equivalent
% sgml_test plaintext MIF <any_old_file >mif_equivalent
I gotta go. It's a rush job, but I'd like folks to get a look
at the HMDoc architecture.
Besides... HTML to MIF is handy!
(and plaintext to HTML might be handy for servers...)
Dan