Necessity of ``pass''

Phil Hunt philh at vision25.demon.co.uk
Tue Aug 24 03:04:12 CEST 1999


In article <hhemgu41z7.fsf at forcix.roof.lan> forcer at mindless.com "forcer" writes:
> Hi there.
> 
> Lately, i've been thinking about Python. While it is an
> exceptionally clean language, it has one very obvious wart:
> The ``pass'' statement.
> 
> To quote the reference manual:
> 
>    pass is a null operation -- when it is executed, nothing
>    happens. It is useful as a placeholder when a statement is
>    required syntactically, but no code needs to be executed, for
>    example:
> 
>    def f(arg): pass    # a function that does nothing (yet)
> 
> So it happens that Python has places where semantically nothing
> should be done, while syntactically it requires a statement.
> 
> I consider this suboptimal.
> 
> As a solution, one could allow the following:
> instead of
> 
>         if foo:
>            pass
>         bar
> 
> one could write
> 
>         if foo:
>         bar
> 
> Syntactically this does not introduce any ambiguity. The only
> problem i see is that the "pass" version is more readable.
> 
> But...
> Most of the time, people write
> 
>         if foo:pass
>         bar
> 
> anyways so the "readability" does not suffer from leaving pass
> out.
> 
> Also, the requirement for pass makes it impractical to comment
> out parts of an if expression, like in
> 
>         if foo:
>            # baz
>            # qux
>         bar
> 
> The new syntax would even allow a smooth transition, or
> coexistence of both versions. Pass can be retained to keep the
> possibility for a explicit do-nothing statment, while it's just
> not mandatory anymore.
> 
> I'd like to see such a change in the Python language, so it'll be
> even cleaner than it was before.
> 
> I hope my opinions and thoughts are of value to you,
>        --forcer

<aol>I agree with all this</aol>

-- 
Phil Hunt....philh at vision25.demon.co.uk





More information about the Python-list mailing list