PPI won't descend into functions or methods

Doesn't stop perlcritic from doing the descending, or running B::Xref

and can't descend into XS

descending into XS is not insurmountable either

You can only get so far with static analysis when trying to find all modifications of a global variable at or below your current call site.

You can well get pretty far I tell you what

who says we have to find them all? or even descend? or even all globals?

I see no use case for relying on the behaviour we see here , so IMHO it seems like a very safe low level warning for every  $global ... func to be turned into  "$global" ... func, esp for $1

Doesn't stop perlcritic from doing the descending, or running B::Xref

How is P::C to descend into code that's not loaded until runtime?

descending into XS is not insurmountable either

How is P::C to descend into compiled code that's not loaded until runtime and for which the source code is not available?

IMHO it seems like a very safe low level warning for every $global ... func to be turned into "$global" ... func , esp for $1

How do you know you've caught every global? How do you know stringification is desirable?

How is P::C to descend into code that's not loaded until runtime?

It could, wait for it, load the code

do the scandeps dance

not do anything at all

 

How is P::C to descend into compiled code that's not loaded until runtime and for which the source code is not available?

Who said P::C is supposed to do it ?

How do you know you've caught every global? How do you know stringification is desirable?

Clearly stringification of $1 is undesireable -- I am humbled by your logic