WAD @EVAL

[vefatica]

This is expected.

v:\> echo %@eval[(0 & 1) + 0]
0

This is not expected.

v:\> set n=0

v:\> echo %@eval[(%n & 1) + %n]
TCC: Divide by zero "(0 & 1) + %n"

 

[samintz]

I don't think you need the %'s inside of EVAL.

set n=0
echo %@eval[(n & 1) + n]
0
set n=1
echo %@eval[(n & 1) + n]
2

 

[vefatica]

I don't think you need the %'s inside of EVAL.

set n=0
echo %@eval[(n & 1) + n]
0
set n=1
echo %@eval[(n & 1) + n]
2

I know. But I always use them anyway. I've never seen them hurt (before).

 

[rconn]

Scott is right - don't use % unless you are referencing an internal variables or function. In you example, part of the line is being expanded by the parser and part by @eval.

 

[vefatica]

Scott is right - don't use % unless you are referencing an internal variables or function. In you example, part of the line is being expanded by the parser and part by @eval.

I have always used "%" and probably will continue. This problem is as old as the hills. The help could call some attention to it. How do you fix this one?

... a bunch of ways ...

 

[rconn]

There's two choices here:

1. I can write a custom parser solely for handling @eval variable expansion. (Not going to happen.)

2. I can remove the @eval function from TCC.

 

[vefatica]

Or you could put something prominently in @EVAL's help. That's the first place I went after getting unexpected behavior, and all l could find was

& same as AND

 

[rconn]

I have always used "%" and probably will continue. This problem is as old as the hills. The help could call some attention to it. How do you fix this one?

Double quote the @EVAL expression:

echo %@eval["(%n&1)+1"]