Thanks, most of that I knew but the grouping in your first example is the one that allowed me to only use records matching any field but still limited by the one db field. So thanks again. I was not aware of the parenthesis for grouping objects like a mathematical equation where anything inside the paren's gets executed first and still conditional on the remainder of the equation. TG wrote: > Couple of little pointers. > > If you're doing the sub-select, then you don't need the "group like 'mac'" > because you've already limited your query in the subselect to a specific > groupname. > > The subselect is probably unnecessary since what you're doing is relatively > simple and uses the same table. It probably just adds overhead. But in > relation to that, you need to segregate your OR statements from your final > AND statement or it won't limit by group properly. > > WHERE ( `ordernum` LIKE "35132" > OR `price` LIKE "35132" > OR `partnum` LIKE "35132" > OR `vendor` LIKE "35132" > OR `purpose` LIKE "35132" > OR `tracking` LIKE "35132" > OR `contact` LIKE "35132" ) > AND `group` LIKE 'mac' > > Notice the ( and ). > > Next pointer.. LIKE is used when you're doing a non-exact search, but you > haven't used any wildcards to indicate a partial search. So what you're > essentially doing is the same as "price = 35132". If you're doing a > multi-field search and want to use LIKE, you'd do this on each: > > `price` LIKE "%35132%" > > The % is sorta like * in other systems. Any quantity of any characters can > match that space. Or you can do `price` LIKE "35132%" if you want to > search the beginning of the field (note the % at the end, but not the > beginning of the search string this time). > > One last thing... I don't know if it increases speed or not, but since you're > using so many fields this MAY speed up the query. Depends on how much > data you're searching through and if you're doing %search% or need to find > start/end strings. > > Anyway, here's the test.. what's faster, what you're trying to do: > > WHERE ( `ordernum` LIKE "35132" OR `price` LIKE "35132" OR `partnum` LIKE > "35132" > OR `vendor` LIKE "35132" OR `purpose` LIKE "35132" OR > `tracking` LIKE "35132" > OR `contact` LIKE "35132" ) > AND `group` LIKE 'mac' > > or something like this.... > > WHERE CONCAT(`ordernum`, `price`, `partnum`, `vendor`, `purpose`, `tracking`, > `contact`) LIKE "%35132%" > AND `group` LIKE '%mac%' > > > All the LIKE comparisons against a ton of data MAY be more taxing on the > server than doing a CONCAT of all the fields then doing a single LIKE. If > you're doing LIKE %search% where it can appear anywhere in any of the > fields, then CONCAT + LIKE would work just as good as LIKE OR LIKE OR ... > results-wise. I don't know if it's faster/less intensive or not though. > You'd have to do some tests. > > Also, comparisons like "=" should (if I recall) be faster than LIKE > comparisons. So if you really meant to use "=", do that instead. > > Let me know if any of that's unclear. I know I get some kooky ideas > sometimes. Also, anyone see anything I screwed up or have thoughts on this > matter? > > Good luck! > > -TG > > > > > > > ----- Original Message ----- > From: Jas <jason.gerfen@xxxxxxxxxxxx> > To: php-db@xxxxxxxxxxxxx > Date: Wed, 26 Sep 2007 12:08:53 -0600 > Subject: Re: mysql statement [SOLVED] > >> Got if figured out, needed a sub-select type of query: >> >> mysql> SELECT * >> -> FROM ( SELECT * FROM `orders` >> -> WHERE `group` = "groupname" ) >> -> AS orders UNION SELECT * FROM `orders` >> -> WHERE `ordernum` LIKE "35132" >> -> OR `price` LIKE "35132" >> -> OR `partnum` LIKE "35132" >> -> OR `vendor` LIKE "35132" >> -> OR `purpose` LIKE "35132" >> -> OR `tracking` LIKE "35132" >> -> OR `contact` LIKE "35132" >> -> AND `group` LIKE 'mac' >> -> ORDER BY `ordernum` >> -> LIMIT 0 , 30; -- PHP Database Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
