I was able to track it down a little more. The problem is not about "subscenarios", but about scenarios in general, when you pass in a Slim variable that has been defined in a table before.

The result is here http://twitpic.com/a7m6l/full

My test wiki page looks like this.

I have a simple !-EchoAndEquality-! class.
{{{class EchoAndEquality
{
    public function valueEquals($a, $b)
    {
        return $a === $b;
    }

    public function echoValue($x)
    {
        return $x;
    }
} }}}
The class follows this specification and Slim variable assignment works.

!|script|echo and equality     |
|ensure |value     |b |equals|b|
|reject |value     |c |equals|d|
|$x=    |echo value|a          |
|ensure |value     |$x|equals|a|

Scenarios do not work as expected, when Slim variables are passed in as parameters and they are called in a table, after the variable was set the first time.

!|scenario|equality of|first value|and   |second value|
|ensure   |value      |@firstValue|equals|@secondValue|

This works, since $y was not defined in a table before. It would not work for $x passed into the scenario.

!|script                     |
|$y=        |echo value|b    |
|equality of|$y        |and|b|

This one fails in the scenario in a new table. The direct call without the scenario works.

!|script                            |
|$y=        |echo value|c           |
|equality of|$y        |and|c       |
|ensure     |value     |$y |equals|c|

Gregor







2009/7/10 Gregor Gramlich <gramlich@...>:
> Hello,
>
> I just switched to the new release and lots of my tests using PhpSlim
> started to fail. They make heavy use of scenarios invoking
> subscenarios.
> The bug seems to be due to a different behaviour of variables within
> subscenarios that are executed in a separate table after the variable
> had been set. So it probably was introduced with the table by table
> execution, but I am pretty sure, that it the fault is not on the Slim
> server side, but on the FitNesse side.
>
> The variable is REPLACED by the assigned value that was made in the
> table before. I put up a simple test.
>
> You can see the result at http://twitpic.com/9w9h2/full
>
> ---
>
> I have a simple !-SetterGetter-! class with a set(string x), get(),
> valueEquals(string x, string y) method.
>
> !|script|setter getter   |
> |set    |a               |
> |check  |get  |a         |
> |ensure |value|b|equals|b|
> |reject |value|c|equals|d|
>
> Scenarios that assign values to variables which are reused in
> subscenarios do not work anymore as expected.
>
> !|scenario  |set                          |value
>  |equals get                                                  |
> |set        |@value
>                                                              |
> |$x=        |get
>                                                              |
> |#          |The following line works and uses the correct value of x
>                                                              |
> |ensure     |value                        |$x
>  |equals                        |@value                       |
> |#          |But when we do it by a sub scenario, the $x value is not
> passed into it correctly. The last value assigned to $x is used|
> |equality of|$x                           |and
>  |@value                                                      |
>
> !|scenario|equality of|first value|and   |second value|
> |ensure   |value      |@firstValue|equals|@secondValue|
>
> The whole table here works.
>
> !|script            |
> |start|setter getter|
> |set  |a            |
> |$x=  |get          |
> |check|get    |$x   |
> |check|get    |a    |
> |set|b|equals get|
>
> This one fails in the subscenario.
>
> !|script         |
> |set|c|equals get|
>
> ---
>
> Gregor
>