Gossamer Forum
(Advanced Search)
for
Home : Products : Gossamer Links : Version 1.x :

perl nph-index.cgi [problem]

Jun 22, 2000, 9:30 AM
daywizard
Novice (29 posts)
Jun 22, 2000, 9:30 AM
Post #1 of 4
Views: 3042
Shortcut
Quote Reply
perl nph-index.cgi [problem]
# perl nph-index.cgi
HTTP/1.0 200 OK
Content-Type: text/html; charset=ISO-8859-1

<!DOCTYPE HTML
PUBLIC - "-//W3C//DTD HTML 5.01 Transitional//EN"
"http://www.w3.org/TR/html4/loose.dtd">
<HTML LANG="en-US"><HEAD><TITLE>Links Manager: Indexing Links</TITLE>
</HEAD><BODY><pre>Indexing 1681022 links. Started at Thu Jun 22 12:23:49 2000

Ubdexubg 286264 Categories. Started at Thu Jun 22 12:23:49 2000

Done (0 s)</pre>

</BODY></HTML>

# :(
#

Then it stops, doing nothing. Previous to this a successful 'perl Parse_RDF.pl' was accomplished.

Possible solutions?




Jun 22, 2000, 12:41 PM
pugdog
Veteran / Moderator (6956 posts)
Jun 22, 2000, 12:41 PM
Post #2 of 4
Views: 2963
Shortcut
Quote Reply
Re: perl nph-index.cgi [problem] In reply to
Did you give the links a "weight" in the Admin/Table/Update area?

nph-index will only index fields in the Links or Category tables that have a weight >= 1



http://www.postcards.com
FAQ: http://www.postcards.com/FAQ/LinkSQL/

Jun 22, 2000, 2:00 PM
daywizard
Novice (29 posts)
Jun 22, 2000, 2:00 PM
Post #3 of 4
Views: 2955
Shortcut
Quote Reply
Re: perl nph-index.cgi [problem] In reply to
What would be the optimal weighting for an index all for the full dmoz data in your opinion (or how you would do it), for the following:

ID
Title
URL
Add_Date
Mod_Date
CategoryID
Description
Contact_Name
Contact_Email
Hits
isNew
isChanged
isPopular
Rating
Votes
ReceiveMail
Status
Date_Checked

?

The goal is to produce a nice searchable database that makes sense to your average general user.

Thanks for the feedback.

Daywizard


Jun 22, 2000, 3:16 PM
pugdog
Veteran / Moderator (6956 posts)
Jun 22, 2000, 3:16 PM
Post #4 of 4
Views: 2971
Shortcut
Quote Reply
Re: perl nph-index.cgi [problem] In reply to
It really depends on what you want to do with it.

The three fields that make the most sense to index are URL, Title and Description. What sort of weight you want to give them is up to you.

I've added "keyword" field to all my databases, and weight that a 5, which pretty much overshadows the other fields which I rate either a 1 or 3.

If you trust the description, you can weight that higher than the Title, or if you think the URL means something, you can index that.

No one does it quite the same. I have 1 for the URL, 1 for the Title, and 1 for the description and 5 for the keywords. I also have a "short" description on some of the sites, which I give a 3.

If you give everything a '1' then all hits count equal, and that might be the best all around.

http://www.postcards.com
FAQ: http://www.postcards.com/FAQ/LinkSQL/