Suggestions other than grep -b3 :-)

[Robin Berjon]

On Apr 13, 2006, at 16:31, Jacqui Caren wrote:
> Simon Wistow wrote:
>> About the only thing I could suggest was either writing a Debugger  
>> or using PPI.
>
> a grep -b3 took 30 seconds to write and caught ~20 of the lill  
> buggers.
>
> I was hoping someone would mention a "use lint" or similar :-)

I would think that any regex that detects duplicate uses of the same  
variable name following my in the same file, and then weeding out the  
common ones ($i, $self) would find them all with minimal false  
positives. If there's a lot of variable name reuse, then checking  
that there isn't a "^sub " between the two should kill the remaining  
false positives.

Another option might be to use one of the B:: modules to walk all the  
pads and check that there isn't any variable shadowing. Might be a  
bit painful though.

-- 
Robin Berjon
    Senior Research Scientist
    Expway, http://expway.com/