[Htmlparser-user] Re: Htmlparser-user digest, Vol 1 #195 - 3 msgs

[<wf...@ma...>]

From: "Somik Raha" <so...@ya...>
To: <htm...@li...>
Subject: Re: [Htmlparser-user] Anyone around using htmlparser together=20
with=20
>Lotus Domino?
>Date: Sat, 15 Feb 2003 20:23:09 -0800

>Thats interesting - can you tell us how you are using the parser with=20
Lotus
>Domino, and what your doubt is ?

Thank you for your reply, Somik.
Since Domino R6 things have changed a little, however it will take some=20
time until this release becomes widely accepted. So what I'm investigatin=
g=20
is related with R5 that supports Java 1.1.8 natively. There are several=20
things I'm investigating:

1) Referrer Spamming:
This is becoming increasingly popular since referrers can be tweaked so=20
easily. The blogging scene often presents a list of recent referrers w/o=20
any validation. This can trick webmasters and visitors into clicking=20
spammed ones. I'm looking for a way to filter for valid references only.
Using Domino one can retrieve a HTML page including a list of hyperlinks=20
however a) performance is not impressive and b) this requires a web=20
interface database (perweb.nsf) is set up on the server. I'd prefer to us=
e=20
the HTMLParser class instead. This looks like a simple one.

2) HTML translation/validation/repair
Domino's proprietary rich text format dates back to the 80s when HTML=20
wasn't a standard. Domino's rich-text capabilites are impressive,=20
including nested interactive sections, features like hotspots,=20
script-enabled buttons, tabbed forms and alike. Due to compatibility=20
reasons Domino was web-enabled mainly not by downsizing this format to=20
HTML's native capabilites but by adding a richtext-to-html task and addin=
g=20
a special URL syntax. Although displayed properly by browsers the=20
generated HTML is not clean, e.g. list tags are not closed, stuff like=20
this. I'm investigating if HTMLParser could be used to do some automatic=20
repair - content will be edited in Domino's RTF for convenience and the=20
resulting HTML is parsed, corrected and seperately stored for web=20
delivery. I assume to parse HTML forgivingly the parser needs to perform=20
some stack correction and I hope this can easily be used for HTML repair=20
as well?

--=20

Mit freundlichen Gr=FC=DFen / Kind regards

Wolfgang Flamme
wf...@ma...

Am Jungst=FCck 32
55130 Mainz-Laubenheim
Tel.: +49 (6131) 8 74 02
Mobil: +49 163 25 43 166