On 12/8/2019 1:29 PM, R.Wieser wrote:
from the_file import ClassName
from somemod import name
has the same effect as
import somemod
name = somemod.name
del somemod
which is to import a particular 'name' linked to a particular python
object into the namespace where the import is executed.
sys.modules['somemod'] will become or continue to be the module
resulting from executing the 'somemod' code.
Note that in all cases when you import a module (either by import the_file
or from the_file importe whatever) you actually import ALL of it
The entire module is executed the first time it is imported.
So much for my assumption only the class itself would be loaded - and a
wrench into my idea to have a number of classes in a "library" file.
Not really. The resulting global objects do not normally take enough
space to worry about on normal modern desktops. It is normal for a
module to only use a subset, possibly just one, of the objects in an
imported module. For instance, itertools has 18 public classes but
almost no importer uses all of them.
if __name__ == '__main__':
# not run when imported
print("Hello world!")
Thanks for that. It means I do not have to block-quote the testcode every
time - which I'm certain I will forget now-and-again ...
Standard for in-file test is
def test(): <test code>
if __name__ == '__main__':
test()
Question: what is, in python, the convention in naming classes ?
The PEP 8 convention (except for basic builtin data structures like int,
str, list, tuple, set, and dict) is TitleCase. Follow or not as you wish.
Pre- or postfix it with "class" ?
No
--
Terry Jan Reedy
--
https://mail.python.org/mailman/listinfo/python-list
