Suspected bug

The code at the bottom of this email displays either a bug in REBOL 2.3.0.4.2 or a bug in my head. The sample was quickly cut down from a larger program under development to try to demonstrate the key issue, so it's a bit untidy. The code was supposed to perform a sort of 2-d traversal of the alphabet. After a letter is picked for the row, pick a letter for the column and, in the original program, do something which involves comparing the row and column letters. In the sample output immediately below, it is obvious that one method of getting the row letter (using a shared local variable) fails miserably, getting stuck on the dummy value to which that variable was initialized, and never sees any subsequent changes to the variable. The second method of getting the row letter (passing it as an argument) works, demonstrating that the letter is actually changing. A : ~ B : ~ C : ~ D : ~ E : ~ F : ~ G : ~ A : A B : A C : A D : A E : A F : A G : A A : ~ B : ~ C : ~ D : ~ E : ~ F : ~ G : ~ A : B B : B C : B D : B E : B F : B G : B A : ~ B : ~ C : ~ D : ~ E : ~ F : ~ G : ~ A : C B : C C : C D : C E : C F : C G : C A : ~ B : ~ C : ~ D : ~ E : ~ F : ~ G : ~ A : D B : D C : D D : D E : D F : D G : D A : ~ B : ~ C : ~ D : ~ E : ~ F : ~ G : ~ A : E B : E C : E D : E E : E F : E G : E A : ~ B : ~ C : ~ D : ~ E : ~ F : ~ G : ~ A : F B : F C : F D : F E : F F : F G : F A : ~ B : ~ C : ~ D : ~ E : ~ F : ~ G : ~ A : G B : G C : G D : G E : G F : G G : G So... Why does 'doindx1 fail, while 'doindx2 succeeds? Shouldn't changes to the local variable 'initial in 'run be visible to both of them (albeit by different mechanisms)? -jn- REBOL [] bug: make object! [ run: function [ ][ initial alphabet letter doindx ][ initial: #"~" letter: #"~" doindx1: func [] [ foreach letter alphabet [ prin [letter ":" initial " "] ] print " " ] doindx2: func [initial] [ foreach letter alphabet [ prin [letter ":" initial " "] ] print " " ] alphabet: "ABCDEFG" foreach initial alphabet [ doindx1 doindx2 initial print " " ] ] ]

when you do something like this: foreach someword someseries [ print someword ] then someword is set to the values of someseries, but only within the code block of the foreach loop. I think there is a new context created for the foreach invocation, and in that context the value of someword is indeed repeatedly set to each element of someseries. Notice that you won't be able to see someword outside the foreach loop before or after it, although you may be able to see something if someword has a value in the enclosing, possibly global context. This is actually a good feature of rebol, I like it. In other languages like c, pascal, or basic, you can see the looping variables value after the function executes, and in large software you have to worry about clobbering some other variable. But here you know that whatever word you choose to use with foreach will only be redefined within the code block, and not anywhere else. Now, doindx1 is defined much earlier, and it has a code block which when created will have find initial to be a local variable of the run function, so that is what is bound there inside doindx1 when the function creation is evaluated. To do what you want, you would have to move doindx1 down into the foreach loop, or else pass it as a parameter, or else don't use foreach, or else use some other global variable. -Galt

Galt wrote:

> I think there is a new context created for the foreach invocation, and in

Hello [joel--neely--fedex--com]! On 11-Ago-00, you wrote: j> foreach initial alphabet [ j> doindx1 j> doindx2 initial j> print " " j> ] No bug here. 'initial is local to FOREACH, so it is different from RUN's local 'initial. Try writing: foreach init alphabet [ initial: init doindx1 doindx2 initial print " " ] Regards, Gabriele. -- Gabriele Santilli <[giesse--writeme--com]> - Amigan - REBOL programmer Amiga Group Italia sez. L'Aquila -- http://www.amyresource.it/AGI/