Skip to main content.
home | support | download

Back to List Archive

Re: Does not count files

From: Bill Moseley <moseley(at)>
Date: Mon Mar 04 2002 - 19:34:23 GMT
At 11:24 AM 03/04/02 -0800, Paul Thomas wrote:
>Turns out it is 'IgnoreWords SwishDefault' as well as 'IgnoreWords File'
>that is gobbling up words I want indexed. Apparently when either of
>those two options are enabled, alot of words not in SwishDefault or
>File are ignored. By commenting out both those fields, things work better.

Are you not using IgnoreLimit?

I'd not bother with stopwords, in general.  Especially since you might want
to include those words in phrase searches.

>> Also looks like you are running an older version of swish.  2.1-dev is
>> still being developed (feature wise), but I'd recommend it over what you
>> are running.
>I'd be interested in new features for sure. 

Take a look at the CHANGES file for a summary of most changes.

>> BTW -- what kind of archive are you indexing?  If hypermail, I've got a
>> little perl program that automatically indexes them with swish.
>Sorry. I'm using Mhonarc. What does your script do?

Let's see if I can remember.  It uses a config file to define a collection
of archives.  Then it can do three things:

1) if you define where the mbox files are archived it can create a new
hypermail archive from the mbox files.

2) it can be run via cron to keep all the archives indexed by swish

3) it can be run via procmail and the script will decide which archive the
incoming mail message belongs to and adds it to that archive, and then
updates the swish index.

It also adds a search box to the top of all the hypermail generated pages,
so you don't need a separate "Search" page.

Bill Moseley
Received on Mon Mar 4 19:36:06 2002