> I've been working on a database wrapper class for a while > now, MySQL to > be specific. Until now, I've simply had a fetch function > that returned > an array of all the rows that the database returned, only > because I have > gotten so tired of always writing the same while loop to > iterate through > all the rows. > > However, I have discovered that the method I'm using, passing around > large multidimensional arrays of text by copy, is extremely memory > inefficient, so I am working on a new method. Tell me if this > is any better:
[...snip...] > Where fetch($query, &$array) is the header. In the second case, the > fetch function would therefore write the rows directly to the array > which was passed as a reference rather than returning a copy. > Am I right > in thinking that this is a better method? I hope your new method is the right one, because I do it the same way and for the same reason :) I haven't done any benchmarks, though. I use procedural code, but the idea is the same. The calling routine builds the query string and passes it into a function that executes the query, processes the results set, puts the results into a hash, and returns a reference to the hash. The calling routine then gets what it needs out of the hash. So, one simple scalar gets passed in both directions. I use that one function for all queries. Kirk -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
