OK to pass big data in 'the result'?

[Sarah Reichelt]

On 3/16/07, Phil Davis <revdev at pdslabs.net> wrote:
> Is anyone aware of any drawbacks in using 'the result' as a container for
> passing, say, a bunch of records from one handler to another? Like this:
>
>
> on processRecords
>    getRecords
>    put the result into tRecordList
>    repeat for each line tLine in it
>      -- your code here
>    end repeat
> end processRecords
>
>
> on getRecords
>    put the uBigList of stack "XYZ" into tList
>    delete stack "XYZ" -- to remove it from memory
>    return tList
> end getRecords

I don't see any problems with this Phil. So long as the data is in
variables and not in fields, then access is extremely fast almost
regardless of size. As a matter of style, I would make getRecords a
function - as a general rule, if a handler has to return data, then
use a function.

So here is my version:

on processRecords
  put getRecords() into tRecordList
  -- other processing
end processRecords

function getRecords
  put the uBigList of stack "XYZ" into tList
  delete stack "XYZ" -- to remove it from memory
  return tList
end getRecords

Note the parentheses after the call to getRecord - with a function,
you have to have these and they would enclose any parameters if you
needed to send any.

Regards,
Sarah