qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Daniel P. Berrangé" <berrange@redhat.com>
To: Markus Armbruster <armbru@redhat.com>
Cc: John Snow <jsnow@redhat.com>, qemu-devel <qemu-devel@nongnu.org>,
	Peter Maydell <peter.maydell@linaro.org>,
	Michael Roth <michael.roth@amd.com>
Subject: Re: [PATCH 05/19] qapi/schema: make c_type() and json_type() abstract methods
Date: Wed, 22 Nov 2023 09:54:34 +0000	[thread overview]
Message-ID: <ZV3P2sDLVEanjK-9@redhat.com> (raw)
In-Reply-To: <87il5uw1zc.fsf@pond.sub.org>

On Wed, Nov 22, 2023 at 10:50:47AM +0100, Markus Armbruster wrote:
> Daniel P. Berrangé <berrange@redhat.com> writes:
> 
> > On Tue, Nov 21, 2023 at 11:28:17AM -0500, John Snow wrote:
> >> On Tue, Nov 21, 2023, 8:43 AM Daniel P. Berrangé <berrange@redhat.com>
> >> wrote:
> >> 
> >> > On Tue, Nov 21, 2023 at 02:36:54PM +0100, Markus Armbruster wrote:
> >> > > John Snow <jsnow@redhat.com> writes:
> >> > >
> >> > > > These methods should always return a str, it's only the default
> >> > abstract
> >> > > > implementation that doesn't. They can be marked "abstract" by raising
> >> > > > NotImplementedError(), which requires subclasses to override the method
> >> > > > with the proper return type.
> >> > > >
> >> > > > Signed-off-by: John Snow <jsnow@redhat.com>
> >> > > > ---
> >> > > >  scripts/qapi/schema.py | 8 ++++----
> >> > > >  1 file changed, 4 insertions(+), 4 deletions(-)
> >> > > >
> >> > > > diff --git a/scripts/qapi/schema.py b/scripts/qapi/schema.py
> >> > > > index c5fdd625452..4600a566005 100644
> >> > > > --- a/scripts/qapi/schema.py
> >> > > > +++ b/scripts/qapi/schema.py
> >> > > > @@ -233,8 +233,8 @@ def visit(self, visitor):
> >> > > >  class QAPISchemaType(QAPISchemaEntity):
> >> > > >      # Return the C type for common use.
> >> > > >      # For the types we commonly box, this is a pointer type.
> >> > > > -    def c_type(self):
> >> > > > -        pass
> >> > > > +    def c_type(self) -> str:
> >> > > > +        raise NotImplementedError()
> >> > > >
> >> > > >      # Return the C type to be used in a parameter list.
> >> > > >      def c_param_type(self):
> >> > > > @@ -244,8 +244,8 @@ def c_param_type(self):
> >> > > >      def c_unboxed_type(self):
> >> > > >          return self.c_type()
> >> > > >
> >> > > > -    def json_type(self):
> >> > > > -        pass
> >> > > > +    def json_type(self) -> str:
> >> > > > +        raise NotImplementedError()
> >> > > >
> >> > > >      def alternate_qtype(self):
> >> > > >          json2qtype = {
> >> > >
> >> > > I wish abstract methods could be done in a more concise way.
> >> >
> >> > The canonical way would be to use the 'abc' module:
> >> >
> >> >   from abc import ABCMeta, abstractmethod
> >> >
> >> >   class A(metaclass=ABCMeta):
> >> >       @abstractmethod
> >> >       def foo(self): pass
> >> >
> >> > Not sure if the @abstractmethod decorator is enough to keep the static
> >> > typing checker happy about a 'str' return type, when there is no body
> >> > impl
> >> 
> >> In practice, I'm under the belief that mypy and pylint both recognize a
> >> method raising NotImplementedError as marking an abstract method without
> >> needing to explicitly inherit from the ABC.
> >> 
> >> I suppose there's also
> >> https://docs.python.org/3/library/abc.html#abc.abstractmethod which I am
> >> guessing just does this same thing. I'll see what makes mypy happy. I'm
> >> assuming Markus would like to see something like this decorator to make it
> >> more obvious that it's an abstract method.
> >
> > The 'abc' module described is an official PEP standard
> >
> >   https://peps.python.org/pep-3119/
> 
> Compare:
> 
>     @abstractmethod
>     def c_type(self) -> str:
>         pass
> 
>     def c_type(self) -> str:
>         raise NotImplementedError()
> 
> I prefer the former, because it's more explicit.
> 
> Bonus: prevents accidental instantiation, and sub-classes don't need to
> know what's abstract in the super-class, they can blindly use super()
> calls. 

Being able to blindly call the parent impl via super() is more than
just a bonus, it is a super compelling reason to use this. It protects
your code against bugs from future re-factoring of the class hierarchy
whether adding or removing parents. Even if we don't expect to need it
for this particular class, I think this justifies having a policy of
using 'abc' everywhere we need abstract methods.

With regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|



  reply	other threads:[~2023-11-22  9:55 UTC|newest]

Thread overview: 76+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-11-16  1:43 [PATCH 00/19] qapi: statically type schema.py John Snow
2023-11-16  1:43 ` [PATCH 01/19] qapi/schema: fix QAPISchemaEntity.__repr__() John Snow
2023-11-16  7:01   ` Philippe Mathieu-Daudé
2023-11-16  1:43 ` [PATCH 02/19] qapi/schema: add pylint suppressions John Snow
2023-11-21 12:23   ` Markus Armbruster
2023-11-16  1:43 ` [PATCH 03/19] qapi/schema: name QAPISchemaInclude entities John Snow
2023-11-21 13:33   ` Markus Armbruster
2023-11-21 16:22     ` John Snow
2023-11-22  9:37       ` Markus Armbruster
2023-12-13  0:45         ` John Snow
2023-11-16  1:43 ` [PATCH 04/19] qapi/schema: declare type for QAPISchemaObjectTypeMember.type John Snow
2023-11-16  1:43 ` [PATCH 05/19] qapi/schema: make c_type() and json_type() abstract methods John Snow
2023-11-16  7:03   ` Philippe Mathieu-Daudé
2023-11-21 13:36   ` Markus Armbruster
2023-11-21 13:43     ` Daniel P. Berrangé
2023-11-21 16:28       ` John Snow
2023-11-21 16:34         ` Daniel P. Berrangé
2023-11-22  9:50           ` Markus Armbruster
2023-11-22  9:54             ` Daniel P. Berrangé [this message]
2023-11-16  1:43 ` [PATCH 06/19] qapi/schema: adjust type narrowing for mypy's benefit John Snow
2023-11-16  7:04   ` Philippe Mathieu-Daudé
2023-11-21 14:09   ` Markus Armbruster
2023-11-21 16:36     ` John Snow
2023-11-22 12:00       ` Markus Armbruster
2023-11-22 18:12         ` John Snow
2023-11-23 11:00           ` Markus Armbruster
2023-11-16  1:43 ` [PATCH 07/19] qapi/introspect: assert schema.lookup_type did not fail John Snow
2023-11-21 14:17   ` Markus Armbruster
2023-11-21 16:41     ` John Snow
2023-11-22  9:52       ` Markus Armbruster
2023-11-16  1:43 ` [PATCH 08/19] qapi/schema: add static typing and assertions to lookup_type() John Snow
2023-11-21 14:21   ` Markus Armbruster
2023-11-21 16:46     ` John Snow
2023-11-22 12:09       ` Markus Armbruster
2023-11-22 15:55         ` John Snow
2023-11-23 11:04           ` Markus Armbruster
2023-11-16  1:43 ` [PATCH 09/19] qapi/schema: assert info is present when necessary John Snow
2023-11-16  7:05   ` Philippe Mathieu-Daudé
2023-11-16  1:43 ` [PATCH 10/19] qapi/schema: make QAPISchemaArrayType.element_type non-Optional John Snow
2023-11-21 14:27   ` Markus Armbruster
2023-11-21 16:51     ` John Snow
2023-11-16  1:43 ` [PATCH 11/19] qapi/schema: fix QAPISchemaArrayType.check's call to resolve_type John Snow
2023-11-22 12:59   ` Markus Armbruster
2023-11-22 15:58     ` John Snow
2023-11-23 13:03       ` Markus Armbruster
2024-01-10 19:33         ` John Snow
2024-01-11  9:33           ` Markus Armbruster
2024-01-11 22:24             ` John Snow
2023-11-16  1:43 ` [PATCH 12/19] qapi/schema: split "checked" field into "checking" and "checked" John Snow
2023-11-22 14:02   ` Markus Armbruster
2024-01-10 20:21     ` John Snow
2024-01-11  9:24       ` Markus Armbruster
2023-11-16  1:43 ` [PATCH 13/19] qapi/schema: fix typing for QAPISchemaVariants.tag_member John Snow
2023-11-22 14:05   ` Markus Armbruster
2023-11-22 16:02     ` John Snow
2024-01-10  1:47       ` John Snow
2024-01-10  7:52         ` Markus Armbruster
2024-01-10  8:35           ` John Snow
2024-01-17  8:19             ` Markus Armbruster
2024-01-17 10:32               ` Markus Armbruster
2024-01-17 10:53                 ` Markus Armbruster
2024-02-01 20:54                   ` John Snow
2023-11-16  1:43 ` [PATCH 14/19] qapi/schema: assert QAPISchemaVariants are QAPISchemaObjectType John Snow
2023-11-23 13:51   ` Markus Armbruster
2024-01-10  0:42     ` John Snow
2023-11-16  1:43 ` [PATCH 15/19] qapi/parser: demote QAPIExpression to Dict[str, Any] John Snow
2023-11-23 14:12   ` Markus Armbruster
2024-01-10  0:14     ` John Snow
2024-01-10  7:58       ` Markus Armbruster
2023-11-16  1:43 ` [PATCH 16/19] qapi/schema: add type hints John Snow
2023-11-24 15:02   ` Markus Armbruster
2023-11-16  1:43 ` [PATCH 17/19] qapi/schema: turn on mypy strictness John Snow
2023-11-16  1:43 ` [PATCH 18/19] qapi/schema: remove unnecessary asserts John Snow
2023-11-28  9:22   ` Markus Armbruster
2023-11-16  1:43 ` [PATCH 19/19] qapi/schema: refactor entity lookup helpers John Snow
2023-11-28 12:06   ` Markus Armbruster

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ZV3P2sDLVEanjK-9@redhat.com \
    --to=berrange@redhat.com \
    --cc=armbru@redhat.com \
    --cc=jsnow@redhat.com \
    --cc=michael.roth@amd.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).