From kfm-devel  Mon Jun 11 06:41:54 2001
From: Andreas Hochsteger <e9625392 () student ! tuwien ! ac ! at>
Date: Mon, 11 Jun 2001 06:41:54 +0000
To: kfm-devel
Subject: RE: A more flexible solution for internet keywords
X-MARC-Message: https://marc.info/?l=kfm-devel&m=99224189506966

Hi!

On Sun, 10 Jun 2001, Yves Arrouye wrote:

> Andreas,
> 
> Sounds cool. Please just don't call it Internet Keywords, it's confusing.
> This is related to the Web shortcuts (whatever their names are today) that
> Simon started before I added keywords in the mix.

Thanks for the explanation.
I self got confused about the names too ;-)

I just saw in the code, that there is often occuring the string "ikws".
What does that mean, and should I replace ikw_* with ikws_* and IKWMap
with IKWSMap?

> 
> Also, if I remember correctly, I used to use \1 for the query and \2 for the
> charset. Now there is nothing for the charset apparently. And the Internet
> Keywords machines (those who provide the keyword functionality) expect \2
> for it. If you have to change that, please synchronize the change with me.
> Thanks!

Yes, I changed some of these references in the file kuriikwsfiltereng.cpp
Here's a small translation table:
\1 -> \0
\2 -> \ikw_charset
\3 -> \ikw_responsecharset
\| -> \ikw_fallbackuri

If there are some more changes to do, then they should follow the same
scheme (\ikw_<name> for internal references). Perhaps it makes sense to
add an additional parameter to formatURL which passes a "global" IKWMap
with predefined references. So it's much easier to add references outside
of formatURL without being responsible for the substitution of the
reference.

Perhaps someone can suggest a better solution for the replacement of
\ikw_fallbackuri, which is not done through the IKWMap, because it is
located in another method (ikwsQuery).

One change is still missing:
The GUI checks whether \1 was entered into a newly created web shortcut
and gives a message if not.
This should be changed to a check if '\' is entered.
I'll change this, when I found out where this is done ;-)

Has somebody took a look at the code?
I'm afread, the encoding of the URLs is not done right.
Can somebody tell me, how it's done right?

> 
> YA
> 

Thanks,

	Andreas