Re: [WebMacro-user] CrankyEvaluationExceptionHandler

[Eric B. R. <eb...@tc...>]

On Monday, November 25, 2002, at 01:14  AM, Keats Kirsch wrote:

> +1 for the change.
>
> It has always bothered me that an exception in an #if is evaluated as=20=

> false.
>
> It seems like you could still choose to not throw in the DfltEEH if =
you
> wanted to preserve backwards compatability, but I'd vote against it.

Ditto, because changing DEEH would change it for normal variable=20
references also.

eric

>
> Keats
>
> Eric B. Ridge wrote:
>  > On Friday, November 22, 2002, at 07:22  AM, Endre St=F8lsvik wrote:
>  >
>  >> Shouldn't the cranky-dude kick on _everything_??
>  >
>  >
>  > I've spent some time looking at this, and I'm changing it, but the
>  > "fix" has an interesting side effect when using the
>  > DefaultEvaluationExceptionHandler.
>  >
>  > Currently, #if (b/c of Expression.java) treats all expressions that
>  > throw as "false".  When using CrankyEEH, this isn't good because =
you
>  > miss the exceptions (as Endre reported).  The solution is to make
>  > Expression.java (and IfDirective) honor whatever EEH says to do.
>  >
>  > However, with Endre's examples, WM is throwing
>  > PropertyException.NoSuchMethodException.  This is an exception that
>  > DefaultEEH also wants to throw.  So, with this fix, CrankyEEH and
>  > DefaultEEH both will throw when when you #if ($foo.noSuchMethod()),
>  > whereas previously, because the exception was being treated as=20
> "false",
>  > neither would throw.
>  >
>  > I think this is a GoodThing, but it's important to note because=20
> people
>  > that rely on this behavior with #if/DefaultEEH  (default=20
> configuration
>  > of WM) will start getting exceptions where previously they did not.=20=

>  If
>  > nothing else, at least now NoSuchMethodException is being handled
>  > consistently.
>  >
>  > If anyone thinks this is bad, please yell.
>  >
>  > eric
>
>
>