Skip to main content.
home | support | download

Back to List Archive

error indexing pdf files

From: Jody Cleveland <Cleveland(at)not-real.mail.winnefox.org>
Date: Mon Apr 14 2003 - 19:33:03 GMT
Hi Bill,

You're going to hate me for this, but I'm very lost.

Here's how I've been running swish for indexing:
$ ./spider.pl | /var/www/cgi-bin/search/src/swish-e -S prog -i stdin -f
/var/www/cgi-bin/search/oshkosh.index

Which is fine, but I get NULL for descriptions of returned search items. So,
I'd like to do the StoreDescriptions thing. So, I added it to config file,
and run this:
$ ./spider.pl | /var/www/cgi-bin/search/src/swish-e -S prog -i stdin -f
/var/www/cgi-bin/search/oshkosh.index -c
/var/www/cgi-bin/search/conf/oshkosh.config

Well, I don't get any errors, but it seems to ignore what's in that config
file. Am I doing this the right way?

Jody

> -----Original Message-----
> From: Bill Moseley [mailto:moseley@hank.org]
> Sent: Monday, April 14, 2003 1:57 PM
> To: Jody Cleveland
> Cc: Multiple recipients of list
> Subject: Re: [SWISH-E] error indexing pdf files
> 
> 
> On Mon, 14 Apr 2003, Jody Cleveland wrote:
> 
> > Is there a way to put things like Index File and 
> StoreDescription in the
> > SwishSpiderConfig.pl file?
> 
> Swish doesn't know about SwishSpiderConfig.pl -- all it knows is it's
> running a progam and that program is returning documents.  
> spider.pl just
> happens to look for SwishSpiderConfig.pl by default.
> 
> You can turn things around and call swish from a program.  So 
> instead of:
> 
>   swish-e -c config -S prog -i /path/to/program
> 
> where swish reads input from /path/to/program, you can do:
> 
>   /path/to/program | swish-e -c config -S prog -i stdin
> 
> that "stdin" is a hack to make swish read from standard 
> input.  So you can
> extend that, and from within a program you can say, for example:
> 
> open CONF, "swish.conf" or die $!;
> print CONF <<EOF;
> 
> DefaultContents HTML*
> StoreDescripiton HTML* <body>
> PropertyNames foo
> EOF
> 
> close CONF;
> 
> open SWISH, "swish-e -c swish.conf -S prog -i stdin|" or die $!;
> while ( $doc = fetch_next_doc() ) {
>     print SWISH $doc;
> }
> close SWISH or die "failed to close";
> 
> unlink swish.conf;
> 
> 
> 
> 
> -- 
> Bill Moseley moseley@hank.org
> 
Received on Mon Apr 14 19:36:56 2003