Jython

This might be more of a python question, but here goes anyway.  If I 
import a module that has an error in it, the import correctly fails.  But 
if I then import the same module again, the import succeeds, and the only 
attributes on that module are those that were processed before the failure 
was hit.  Is this correct behavior?  It seems very dangerous.
e.g.

<a.py>
foo='foo'
errorhere
bar='bar'

<interactive>
>>> import a
...NameError...
>>> import a
>>> a.foo
'foo'
>>> a.bar
.. AttributeError...



Matt

Re: [Jython-users] partial imports From: Samuele Pedroni <pedroni@in...> - 2002-02-27 22:21 Hi. From: <Matt_Conway@...> > This might be more of a python question, but here goes anyway. If I > import a module that has an error in it, the import correctly fails. But > if I then import the same module again, the import succeeds, and the only > attributes on that module are those that were processed before the failure > was hit. Is this correct behavior? It seems very dangerous. > e.g. > > <a.py> > foo='foo' > errorhere > bar='bar' > > <interactive> > >>> import a > ...NameError... > >>> import a > >>> a.foo > 'foo' > >>> a.bar > .. AttributeError... > Is the BDFL (Guido van Rossum) intended behavior, the point is that, if an import fails somehow, you are on your own, if you know how to cleanup things you can do that explicitly OTOH the interpreter does not try to illude you to have properly cleaned up things, because there is no general notion of what that means ... the precise rationale is something also Guido continously forget about. regards, Samuele Pedroni.

Re: [Jython-users] partial imports From: Kevin Butler <kbutler@ca...> - 2002-02-27 22:40 Samuele Pedroni wrote: > From: <Matt_Conway@...> > > >>> import a > > ...NameError... > > >>> import a > > >>> a.foo > Is the BDFL (Guido van Rossum) intended behavior, > > the point is that, if an import fails somehow, you are on your own, > if you know how to cleanup things you can do that explicitly > ... > the precise rationale is something also Guido continously forget > about. I've found it quite handy helping to debug the problem that caused the error. import x <error> from x import * <look at values that led to the failure...> It is easy enough to re-execute the failed import with the reload() function: import x <error> import x reload( x ) # executes your fixed code, or triggers the failure again. Standard caveats about Python's non-recursive reload() apply. :-) kb

1 message has been excluded from this view by a project administrator.