[Python-Dev] unexpected reload() behavior

[Skip Montanaro]

In working on the doc change for <http://python.org/sf/919099> I came across
this statement:

    The names in the module namespace are updated to point to any new or
    changed objects. Names of unchanged objects, or of objects no longer
    present in the new module, remain pointing at the old objects.

(The bit about names of unchanged objects still pointing to the old objects
is incorrect.  I'll fix that.)

Not believing that old objects remained after the reload() I wrote a short
test:

    a = 5
    b = 7
    c = (1,2,3)

imported it, modified it to

    a = 9
    c = (1,2,3)

then reloaded it.  I was surprised to find that reloadtst.b did indeed still
exist:

    >>> import reloadtst
    >>> dir(reloadtst)
    >>> dir(reloadtst)
    ['__builtins__', '__doc__', '__file__', '__name__', 'a', 'b', 'c']
    >>> # edit reloadtst.py
    ...
    >>> reload(reloadtst)
    <module 'reloadtst' from 'reloadtst.py'>
    >>> dir(reloadtst)
    ['__builtins__', '__doc__', '__file__', '__name__', 'a', 'b', 'c']

It seems counterintuitive to me that reloadtst.b should still be defined.
Is that behavior intention or accidental?

Skip