Easy question on error catching

[Andrew Dalke]

Erik Max Francis:
> What you wrote will work, but catching a ValueError (except ValueError:
> ...) is a better solution.  Especially in dynamically typed languages
> like Python, it's a good idea to catch the most restrictive error you
> think you need, and then expand the net if it's required.

To elaborate, suppose you press control-C inside of the try
block.  That the signal for Python to stop what it's doing.  Python
converts that into an exception.  If nothing catches it then
the top-level reports the normal exception and traceback.

If you have a bare "except:" then in addition to telling Python
to ignore the ValueError when the string doesn't contain an
integer, you're also telling Python to ignore control-C, out-of-
memory exceptions and a few others.

I've actually had something like this happen to me
 try:
  v = flaot(s)
 except:
  v = 0.0

See the misspelling?  It raised a NameError which was caught
by the bare except: so for every case I was setting v to 0.0.

                    Andrew
                    dalke at dalkescientific.com