when run with cps-header, should pycps catch exceptions not explicitly raised?

[cam]

The following expression causes an exception without our code explicitly raising it:

try:

  y = (lambda x : x) / 1 # EXCEPTION: can't divide a lambda!

except:

  print("caught exception")

Should our lir/cps try handle this kind of an exception in the cps phase, or only raised exceptions?

[Petey A]

When I run it in python3, it catches the exception. This makes sense to me, too. I'd say you should catch _any_ exception raised while in a try block.

- Petey

--
You received this message because you are subscribed to the Google Groups "Utah Compilers, Spring 2013" group.
To unsubscribe from this group and stop receiving emails from it, send an email to utah-compilers-spri...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[cam]

I agree that we should handle these kinds of exceptions to match python3, but how relevant is it that the lir and cps reference implementations choke on these non-raised exceptions?  Is this a {lir,cps}-header.rkt issue and not an emitted code issue?

Is the interpreted program responsible to catch exceptions that the interpreter throws?  Because the header tries to divide the lambda by 1, then chokes.  How can our emitted code handle the interpreter's exception?

Please correct me if I am thinking about this wrong.

[Petey A]

Have you run it through the reference desugarer? Does it do the same?

- Petey

[cam]

Yes, the reference desugarer also crashes.  

- Petey

To unsubscribe from this group and stop receiving emails from it, send an email to utah-compilers-spring-2013+unsub...@googlegroups.com.

[Petey A]

At this point, I'd recommend that you make a decision, document it in your README, and move forward. The reference differs from python3, so all bets are off - except that you can be pretty sure that you won't be graded on this disparity.

- Petey

To unsubscribe from this group and stop receiving emails from it, send an email to utah-compilers-spri...@googlegroups.com.

[Matt Might]

Only worry about *explicitly* raised exceptions.

-Matt

On Thu, Apr 11, 2013 at 1:01 PM, cam <cams...@gmail.com> wrote: