flex-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Maxim Solodovnik <solomax...@gmail.com>
Subject Re: do-while-false
Date Thu, 19 Nov 2015 08:18:50 GMT
I personally prefer

if(conditiona && conditionb && conditionc &&conditiond)
{
//do something
}

it is shorter and more clear to me
not sure about performance, but I guess "if(conditiona && conditionb &&
conditionc &&conditiond)" should produce better native code

On Thu, Nov 19, 2015 at 2:16 PM, Harbs <harbs.lists@gmail.com> wrote:

> There’s a coding pattern that I like to use which I picked up from the
> InDesign SDK. When there’s some code which needs a lot of conditions to be
> executed, it’s hard to write the conditions in a way that’s easily human
> readable.
>
> You can either do:
> if(conditiona && conditionb && conditionc &&conditiond(
> {
> //do something
> }
>
> or:
> if(conditiona){
>   if(conditionb){
>     if(conditionc){
>       if(conditiond){
>         // do something
>       }
>     }
>   }
> }
>
> Both of these are kind of hard on the eyes and make fixes error-prone.
>
> The do-while-false solution is much more ledgible than both of these and
> it goes like this:
>
> do{
> if(!conditiona){break};
> if(!conditionb){break};
> if(!conditionc){break};
> if(!conditiond){break};
> //do something
> }while(false);
>
> The reason it works is that do-while-false executes exactly once, and
> break leaves the “loop”.
>
> The pattern reverses the logic and instead of checking when you *should*
> execute the code, it checks when you should bail out. The reason I like
> this pattern is because it makes for much flatter code and each condition
> stands on its own. That makes it easy to read and fix conditions at a later
> point.
>
> How do folks feel about trying to use this pattern?
>
> What prompted this post is commit b29975c which attempts to make a mess of
> conditions for invalidateSkinState() a bit clearer.
>
> Thoughts?
> Harbs




-- 
WBR
Maxim aka solomax

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message