Re: Re: New search related question

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

 



Inefficient, maybe.  Lazy, most likely yes.

I agree that htdig may be a better solution, however his current
solution requires upkeep if the static HTML is changed and requires that
the person populating the database pick all relevant words from the page
and if new ones are added to update the db.

For example, if you add the entry for the fakeFlowers.html and don't
think it's important to add "long lasting" to the db, even though it
appears on the page, then that search comes up empty.  Also, if the site
owner adds a new page or just updates the Flowers.html to include
"roses", then the db needs to be updated for that page or a new record
added for the new page, etc.

Unless, by FULLTEXT, you're implying that the full text of each page
should be in the db, then I would argue that there is negligible diff
between that and the grep.  Then the only major diff is the
maintainability, which the grep wins.

-Shawn

Robert Cummings wrote:
> On Mon, 2008-02-04 at 14:13 -0600, Shawn McKenzie wrote:
>> If there aren't many files and you don't intend to grow this site much
>> larger and intend to always have static HTML, any easy implementation
>> would be to read each file and search for the terms either in the
>> keywords tag or in the entire file.
>>
>> Optionally, if you're on a *nix host you could exec() a grep for the
>> terms which returns the matching lines in an array and display as needed.
> 
> Wow, that has got to be the most inefficient lazy method I've ever
> heard. I would never suggest such a route on a production server. His
> original plan is much more efficient and is generally along the lines
> how how search indexing works. As such for a simple site I'd do what he
> suggest using a FULLTEXT field in the database, or as Greg Donal
> suggested, use soemthing like htdig. A more involved solution would be
> something like Lucene. Either way, you don't want to be scanning the
> files on ever search request.
> 
> Cheers,
> Rob.

-- 
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php


[Index of Archives]     [PHP Home]     [Apache Users]     [PHP on Windows]     [Kernel Newbies]     [PHP Install]     [PHP Classes]     [Pear]     [Postgresql]     [Postgresql PHP]     [PHP on Windows]     [PHP Database Programming]     [PHP SOAP]

  Powered by Linux