|
|
sybperl-l Archive
Up Prev Next
From: LNowak at aol dot com
Subject: Re: Query re Sybase::CTlib's ct_sql returning alot of data
Date: Aug 16 2000 4:49PM
In a message dated 08/16/2000 12:21:50 PM Eastern Daylight Time,
tim.wilson@barclayscapital.com writes:
<< "
This one step call is very useful for writing simple, clear code, but has
one major drawback: the entire query result set is loaded into memory, which
makes it unusable for processing large result sets.
" >>
When working with data sources 2 ways come to mind:
Reading all the data into variable / array / hash array or streaming the data
in one line/record at a time into variable/array / hash array.
Lets say you have 100mb data source - you can dump all 100 megs in to @data
taking up 100+mb of ram memory or you can stream it in one record at a time
keeping amount memory less that one record size. Now if you know the result
sets going to be small loaded it all up into memory - if the result set is
unknown or known to be rather large don't risk running out memory and stream
it.
As for using Sybase::CTlib - I haven't used it very much - I am a big fan of
Sybase::DBlib.
Leigh
|
