RE: [greenstone-users] Browsing Classifiers Help - Special Case

From Nathan Einwechter
DateTue, 3 May 2005 09:24:00 -0400
Subject RE: [greenstone-users] Browsing Classifiers Help - Special Case
In-Reply-To (426DBB8B-2040009-cs-waikato-ac-nz)
Thanks for the suggestions. It has pointed me in the right direction.
I'm doing things a bit different, and should point that out when
elaborating and seeking clarification.

I now have what I need as far as one of the classifiers, which allows me
to use an IP and see which Title's contain that. Now what I need is to
see all the IP's under a title.

An example of the metadata I'm using is;

<FileSet>
<FileName>Test1.txt</FileName>
<Description>
<Metadata mode="override" name="dc.Title">Test1</Metadata>
<Metadata mode="accumulate"
name="fares.UnauthorizedChannel">192.168.2.255</Metadata>
<Metadata mode="accumulate"
name="fares.UnauthorizedChannel">204.22.185.38</Metadata>
<Metadata mode="accumulate"
name="fares.UnauthorizedChannel">204.22.177.202</Metadata>
<Metadata mode="accumulate"
name="fares.UnauthorizedChannel">204.22.177.78</Metadata>
<Metadata mode="accumulate"
name="fares.UnauthorizedChannel">204.22.180.6</Metadata>
<Metadata mode="accumulate"
name="fares.UnauthorizedChannel">204.22.176.6</Metadata>
</Description>
</FileSet>

Any ideas on how I could do this? Whenever I try various things, I just
get a listing of the Title's and then when I go to the second level, I
get the title and the source again.

Thanks, I really appreciate your input on this.

-- Nathan

-----Original Message-----
From: Katherine Don [mailto:kjdon&#64;cs.waikato.ac.nz]
Sent: Monday, April 25, 2005 11:55 PM
To: Nathan Einwechter
Cc: greenstone-users@list.scms.waikato.ac.nz
Subject: Re: [greenstone-users] Browsing Classifiers Help - Special Case

Hi Nathan,

Without knowing what your data looks like, here are a couple of
suggestions. I am assuming that in order to get your metadata into
Greenstone, you will either create metadata.xml files or write a plugin
to process your log files. This means that you have some leeway to
create metadata that is useful for greenstone.

Note that I haven't tried these suggestions, but hopefully they will
work enough to get you started.

1. Create an IP metadata that looks like
Title|Authorized Communication|address
Title|Unauthorized Communication|address.
Each document will have multiple values for IP: the Title part will be
the same within a document, but the addresses will be different.
Then use Hierarchy classifier:
classify Hierarchy -metadata IP
Hierarchy splits a metadata value on the | symbol to create a hierarchy.

2. Assuming you have metadata Title, UnauthIP, AuthIP
AZList -metadata UnauthIP -sort Title -buttonname "Unauthorized
Communication"

Then in the format statement use eg. [Title]<br>[sibling:UnauthIP]
And the same for authorised IPs.
This gives you one entry per document, with all the IPs displayed.

Alternatively you could try Michael's GenericList classifier.

(http://www.cs.waikato.ac.nz/~mdewsnip/greenstone/GenericList.pm.zip
Unzip this into your Greenstone perllib/classify directory, then delete
the C:Documents and Settings<Username>Application
DataGreenstoneGLIclassifiers.dat file (windows) or
<homedir>/.gli/classifiers.dat file (linux). )

classify GenericList -metadata Title/UnauthIP -buttonname "Unauthorized
Communication"

and in the format statement, display
{If}{[numleafdocs],[Title],[UnauthIP]}.

this gives you one entry per IP.

see
http://www.greenstone.org/cgi-bin/library?e=p-en-faq-utfZz-8&a=p&p=faqcu
stomize#customizeformat
for more formatting details.

Regards,
Katherine Don