Home Original page

import question

Steven D. Majewski sdm7g at Virginia.EDU
Tue Oct 16 11:36:12 EDT 2001 
On Tue, 16 Oct 2001, Vojin Jovanovic wrote:

> Suppose I have a file foo.py file that contains a function
> and now I do import foo.   Then, I have my function available
> in python.  But let's suppose that I decide to change the function in foo.py
> file
> Why is it that doing import foo again doesn't load the new definition  of
> the function
> which would be normal behavior in LISP for example?
> Because of not having such behavior in Python one has to exit the program
> and then
> do import foo in order to load the new definition!?

If you're going to compare Python statements to Lisp, Python's 'import' 
is closer to Lisp's 'require' than it is to 'load' : if it has already
been loaded it won't do it twice. There is very little overhead to 
subsequent 'imports' : all it has to do is make the namespace available.

You don't have to exit and restart the compiler (unless it's a module
written in C as there is no 'unload' capability -- mainly because that's
something that's not supported by the dynamic linking on all platforms.).

Use "reload module", which will cause the file to be reloaded and 
reexecuted.  If you did an import as "from foo import *" ,then you
need to do:

import foo	# get foo into namespace
reload foo	# reload it
from foo import *   # until you do this, the names in this namespace 
		    # still point to the names in the old foo module.



-- Steve Majewski
More information about the Python-list mailing list