Related articles |
---|
"dynamic" parser uk7i@rz.uni-karlsruhe.de (jan) (1998-07-28) |
Re: "dynamic" parser qjackson@wave.home.com (Quinn Tyler Jackson) (1998-08-02) |
Re: "dynamic" parser nnylfv@ny.ubs.com (Olivier Lefevre) (1998-08-10) |
From: | Olivier Lefevre <nnylfv@ny.ubs.com> |
Newsgroups: | comp.compilers |
Date: | 10 Aug 1998 23:21:49 -0400 |
Organization: | Union Bank of Switzerland, New York site |
References: | 98-07-217 |
Keywords: | parse, WWW |
Unless your users want to browse a really wide range of sites,
I'd just special-case each site, as there aren't that many of them
that traders (unless they are into exotic stuff like emerging market
exchanges) really use and every site tends to have a standard format
for all its tables (which often look like Excel CSV dumps, by the way).
I did it once for a few US sites and it was no big deal. You will
probably have to take that route anyway because in some cases the
HTML annotations provided are really minimal: a mere wrapper around
an ASCII table.
If you really want to roll out a general solution, you might want to
look at the SGML parsers and transformation tools listed under
http://www.sil.org/sgml/publicSW.html
-- O.L.
--
Return to the
comp.compilers page.
Search the
comp.compilers archives again.