[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [cobalt-developers] Search Engine Recommendations Research



On Sat, Apr 01, 2000 at 01:30:17AM -0500, Daniel Ouellet wrote:
> 1) Index the content for a search on a web site. So, this will search and
> index html pages of a specific web site.

I prefer to use swish-e http://sunsite.berkeley.edu/SWISH-E/
Very nice and very configurable.  Works fairly well on our
Solaris stations, RedHat and Slackware intel boxes, as well
as on the RaQ and RaQ2.

It stores the indices in it's own format rather than a foreign
RDBMS, but it includes a small program you can use to query 
the db and plenty of perl scripts exist to parse form and the
output to return very nice search results.

It allows you to specify which directories will be kept in which
index file and supply multiple search configurations if desired
so as to exclude certain types of files or directory paths, very
useful for excluding all the _vti* directories and files created
by FrontPage server extensions.

> I would like the end result of the search engine to be based on MySQL as the
> core engine.

Any particular reason you want to use MySQL?  The database format does
not usually matter so much as building all the indices -- which you
have to do no matter what engine you use...Nor does the engine really
have anything to do with the way your output looks.

-- 
Robert G. Fisher		     NEOCOM Microspecialists Inc. 
System Administrator/Programmer      (540) 666-9533 x 116