Use an array of objects instead. You would create a class that has
5 properties that represent the columns. Then New the objects and add
them to your singly-dimensioned array. Put this array in another
class (a 'wrapper' class) that provides a method that 'looks like' an
accessor for a doubly-dimensioned array. This approach is probably
faster, and definitely more extendable. Plus, many of the useful
array methods in RB only work with single-dimension arrays anyway.
Multi-dimension arrays, IMHO, are only useful when all the data
elements are the same type (such as for a bar graph, data plot, pie
chart, etc...) and even then, implementing some of the 'dimensions'
as objects in singly-dimensioned arrays is - for me anyway - preferable.
On Feb 15, 2007, at 2:01 PM, Octave Julien wrote:
Hi,
I'm using a method that analyses a text given as an argument and
returns a two-dimensional array. This array has 5 columns, one
filled with strings, two with integers, and three with booleans.
The returned array is defined as an array of type variant.
It works and returns an array with around 2000 rows from a text of
2 or 3 pages. But it is so slow ! It takes 6 or 7 seconds on my
PowerBook G4 867 mHz, with 384 Mo. With the same sample text it
used to work very fast with a function filling up 6 different one-
dimensional arrays defined before calling the method and given as
arguments.
So my question is : is it slower because I'm using a
multidimensional array, or because this array is defined with the
variant type ? Or both ?
Thanks in advance.
Octave
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives:
<http://support.realsoftware.com/listarchives/lists.html>
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives:
<http://support.realsoftware.com/listarchives/lists.html>
