Mailing List Archive: How do you index ms office (.doc, .xls, .ppt) files with kinosearch

How do you index ms office (.doc, .xls, .ppt) files with kinosearch

Aug 25, 2008, 4:12 AM

Post #1 of 3 (7359 views)

hi
I've red through most of the documentation trying to understand what
filetypes KS supports. There is the interesting oscon presentation on
http://www.rectangular.com/downloads/KinoSearch_OSCON2006.pdf, where
you can find the statement on page 13:

What is KinoSearch not?
...
- Not a file parser
...

So if I get this right, kinosearch doesn't care about your .doc, .xls,
.ppt files. As much as I personally try to avoid this formats, I think
its realistic to assume that you have to index such files when
creating something like an intranet search.

My question is, what would you suggest for indexing office formats ?
How do you extract text without ole and and an office installation on
the server?

thanks in advance
ben

_______________________________________________
KinoSearch mailing list
KinoSearch@rectangular.com
http://www.rectangular.com/mailman/listinfo/kinosearch

Re: How do you index ms office (.doc, .xls, .ppt) files with kinosearch [ In reply to ]

henka at cityweb

Aug 25, 2008, 6:42 AM

Post #2 of 3 (7053 views)

Permalink

On Mon, August 25, 2008 1:12 pm, Ben Aurel wrote:
> My question is, what would you suggest for indexing office formats ?
> How do you extract text without ole and and an office installation on
> the server?

You use file conversion utilities such as pdftotext, xlhtml, wvHtml etc.
Most of these are far from perfect, sometimes crashing, etc.

Regards
Henry

_______________________________________________
KinoSearch mailing list
KinoSearch@rectangular.com
http://www.rectangular.com/mailman/listinfo/kinosearch

Re: How do you index ms office (.doc, .xls, .ppt) files with kinosearch [ In reply to ]

peter at peknet

Aug 25, 2008, 7:00 AM

Post #3 of 3 (7021 views)

Permalink

On 08/25/2008 08:42 AM, Henry wrote:
> On Mon, August 25, 2008 1:12 pm, Ben Aurel wrote:
>> My question is, what would you suggest for indexing office formats ?
>> How do you extract text without ole and and an office installation on
>> the server?
>
> You use file conversion utilities such as pdftotext, xlhtml, wvHtml etc.
> Most of these are far from perfect, sometimes crashing, etc.
>

Also, check out SWISH::Filter on CPAN, which uses many of those tools underneath but which
provides a common interface for converting them to parse-able text.
--
Peter Karman . peter@peknet.com . http://peknet.com/

_______________________________________________
KinoSearch mailing list
KinoSearch@rectangular.com
http://www.rectangular.com/mailman/listinfo/kinosearch