Embedded switches?

Kay C Lan lan.kc.macmail at gmail.com
Mon Sep 28 21:15:27 EDT 2009


On Tue, Sep 22, 2009 at 4:03 AM, Mark Wieder <mwieder at ahsoftware.net> wrote:

>
> Actually there's no speed penalty for if/then vs switch constructs.
> They compile down to the same comparisons. If you're going to be
> applying the same tests repeatedly then you may find a negligible
> difference, but overall I wouldn't worry about this.
>
> Not precisely true. You may wish to look back at these two threads, lots of
benchmarking, sorry I haven't figured out how to create links to old threads
:-(

Subject: if statements vs case
Date: 27 feb 07
From: Hershel Fisch

Subject: switch case question
Date: 10/22/06
From: Mark Swindell

Basically, timings are similar BUT Dar Scott in the 2006 thread determined
those cases (sorry for the pun) where one will be faster than the other. I
believe the general conclusion was, if speed meant everything to you, you'd
need to benchmark both.

For Richmond, there is a simpler, cleaner option than complex CASE or IF
structures containing ever deepening embedded complex CASE or IF structures.
A single complex CASE structure (if it's complex I avoid IF structures)
containing many custom handlers. The custom handlers may then contain a
single complex CASE structure which may contain further custom handlers,
which may contain....., you get the picture.

General rule of thumb, if you spend more than a minute trying to match
IF/END IF, CASE/BREAK (or find yourself constantly using tRev's brilliant
Control Structure hilight feature) then it's probably time to Cut and Paste
some code into it's own custom handler.

HTH



More information about the use-livecode mailing list