R devel - Aug 2017 - Problems with S4 methods dispatching on `...` (aka dotsMethods)

I attempted to fix it, and that example seems to work for me. It's
also a (passing) regression test in R. Are you sure you're using a new
enough R-devel?


On Tue, Apr 25, 2017 at 2:34 AM, Andrzej Ole? <andrzej.oles at gmail.com>
wrote:
> Hi Michael,
>
> thanks again for your patch! I've tested it and I'm happy to
confirm that
> `callNextMethod()` works with methods dispatching on `...`.
>
> However, the second issue I reported still seems to be unresolved. Consider
> the following toy example, where the `f()` calls differ in result depending
> on whether the dispatch happens on a formal argument or the `...` argument.
>
>
>     f = function(x, ..., a = b) {
>       b = "missing 'a'"
>       print(a)
>     }
>
>     f()
>     ## [1] missing 'a'
>
>     f(a = 1)
>     ## [1] 1
>
>     setGeneric("f", signature = "x")
>
>     # works as the non-generic version
>     f()
>     ## [1] missing 'a'
>
>     setGeneric("f", signature = "...")
>
>     # unexpectedly fails to find 'b'
>     f()
>     ## Error in print(a) : object 'b' not found
>
>
> Any chances of fixing this?
>
> Cheers,
> Andrzej
>
>
>
> On Fri, Apr 21, 2017 at 11:40 AM, Andrzej Ole? <andrzej.oles at
gmail.com>
> wrote:
>>
>> Great, thanks Michael for you quick response!
>>
>> I started off with a question on SO because I was not sure whether this
>> was an actual bug or I was just missing something obvious. I'm
looking
>> forward to the patch.
>>
>> Cheers,
>> Andrzej
>>
>>
>> On Thu, Apr 20, 2017 at 10:28 PM, Michael Lawrence
>> <lawrence.michael at gene.com> wrote:
>>>
>>> Thanks for pointing out these issues. I have a fix that I will
commit
>>> soon.
>>>
>>> Btw, I would never have seen the post on Stack Overflow. It's
best to
>>> report bugs on the bugzilla.
>>>
>>> Michael
>>>
>>> On Thu, Apr 20, 2017 at 8:30 AM, Andrzej Ole? <andrzej.oles at
gmail.com>
>>> wrote:
>>> > Hi all,
>>> >
>>> > I recently encountered some unexpected behavior with S4
generics
>>> > dispatching on `...`, which I described in
>>> >
>>> >
http://stackoverflow.com/questions/43499203/use-callnextmethod-with-dotsmethods
>>> >
>>> > TL;DR: `callNextMethod()` doesn't work in methods
dispatching on `...`,
>>> > and
>>> > arguments of such methods are resolved differently than the
arguments
>>> > of
>>> > methods dispatching on formal arguments.
>>> >
>>> > Could this indicate a potential problem with the
implementation of the
>>> > `...` dispatch?
>>> >
>>> > Cheers,
>>> > Andrzej
>>> >
>>> >         [[alternative HTML version deleted]]
>>> >
>>> > ______________________________________________
>>> > R-devel at r-project.org mailing list
>>> > https://stat.ethz.ch/mailman/listinfo/r-devel
>>
>>
>

You're right, I must have mixed up my R versions when running the example,
as the problem seems to be resolved in R-devel.

Sorry for the noise and thanks again for fixing this.

Andrzej

On Tue, Apr 25, 2017 at 3:55 PM, Michael Lawrence <lawrence.michael at
gene.com
> wrote:
> I attempted to fix it, and that example seems to work for me. It's
> also a (passing) regression test in R. Are you sure you're using a new
> enough R-devel?
>
>
> On Tue, Apr 25, 2017 at 2:34 AM, Andrzej Ole? <andrzej.oles at
gmail.com>
> wrote:
> > Hi Michael,
> >
> > thanks again for your patch! I've tested it and I'm happy to
confirm that
> > `callNextMethod()` works with methods dispatching on `...`.
> >
> > However, the second issue I reported still seems to be unresolved.
> Consider
> > the following toy example, where the `f()` calls differ in result
> depending
> > on whether the dispatch happens on a formal argument or the `...`
> argument.
> >
> >
> >     f = function(x, ..., a = b) {
> >       b = "missing 'a'"
> >       print(a)
> >     }
> >
> >     f()
> >     ## [1] missing 'a'
> >
> >     f(a = 1)
> >     ## [1] 1
> >
> >     setGeneric("f", signature = "x")
> >
> >     # works as the non-generic version
> >     f()
> >     ## [1] missing 'a'
> >
> >     setGeneric("f", signature = "...")
> >
> >     # unexpectedly fails to find 'b'
> >     f()
> >     ## Error in print(a) : object 'b' not found
> >
> >
> > Any chances of fixing this?
> >
> > Cheers,
> > Andrzej
> >
> >
> >
> > On Fri, Apr 21, 2017 at 11:40 AM, Andrzej Ole? <andrzej.oles at
gmail.com>
> > wrote:
> >>
> >> Great, thanks Michael for you quick response!
> >>
> >> I started off with a question on SO because I was not sure whether
this
> >> was an actual bug or I was just missing something obvious. I'm
looking
> >> forward to the patch.
> >>
> >> Cheers,
> >> Andrzej
> >>
> >>
> >> On Thu, Apr 20, 2017 at 10:28 PM, Michael Lawrence
> >> <lawrence.michael at gene.com> wrote:
> >>>
> >>> Thanks for pointing out these issues. I have a fix that I will
commit
> >>> soon.
> >>>
> >>> Btw, I would never have seen the post on Stack Overflow.
It's best to
> >>> report bugs on the bugzilla.
> >>>
> >>> Michael
> >>>
> >>> On Thu, Apr 20, 2017 at 8:30 AM, Andrzej Ole? <andrzej.oles
at gmail.com>
> >>> wrote:
> >>> > Hi all,
> >>> >
> >>> > I recently encountered some unexpected behavior with S4
generics
> >>> > dispatching on `...`, which I described in
> >>> >
> >>> >
http://stackoverflow.com/questions/43499203/use-callnextmethod-with-
> dotsmethods
> >>> >
> >>> > TL;DR: `callNextMethod()` doesn't work in methods
dispatching on
> `...`,
> >>> > and
> >>> > arguments of such methods are resolved differently than
the arguments
> >>> > of
> >>> > methods dispatching on formal arguments.
> >>> >
> >>> > Could this indicate a potential problem with the
implementation of
> the
> >>> > `...` dispatch?
> >>> >
> >>> > Cheers,
> >>> > Andrzej
> >>> >
> >>> >         [[alternative HTML version deleted]]
> >>> >
> >>> > ______________________________________________
> >>> > R-devel at r-project.org mailing list
> >>> > https://stat.ethz.ch/mailman/listinfo/r-devel
> >>
> >>
> >
>
	[[alternative HTML version deleted]]

Thank you Michael for updating the 3.4 branch, the `callNextMethod()` now
works for `...` methods as expected. However, I'm still missing your other
patch fixing the handling of arguments in `...` methods. It would be really
great if this bugfix could be integrated into the 3.4 branch as well, such
that the following code doesn't result in an error.

Cheers,
Andrzej


    f = function(x, ..., a = b) {
      b = "missing 'a'"
      print(a)
    }

    f()
    ## [1] missing 'a'

    f(a = 1)
    ## [1] 1

    setGeneric("f", signature = "x")

    # works as the non-generic version
    f()
    ## [1] missing 'a'

    setGeneric("f", signature = "...")

    # unexpectedly fails to find 'b'
    f()
    ## Error in print(a) : object 'b' not found


On Fri, Jul 28, 2017 at 9:15 PM, Michael Lawrence <lawrence.michael at
gene.com
> wrote:
> I pushed the patch to the 3.4 branch. Feel free to test.
>
> Michael
>
> On Wed, Jul 26, 2017 at 4:02 AM, Andrzej Ole? <andrzej.oles at
gmail.com>
> wrote:
> > Hi Michael,
> >
> > it seems that your patch to S4 generics dispatching on `...` is still
> > available only in R-devel, and was not included in the minor R-3.4.1
> > release. I was wondering what is the policy of incorporating bug fixes
> from
> > the devel branch into release, and whether there is any chance that
the
> > broken `...` dispatch is fixed before R-3.5.0?
> >
> > Cheers,
> > Andrzej
> >
> >
> > On Tue, Apr 25, 2017 at 4:15 PM, Andrzej Ole? <andrzej.oles at
gmail.com>
> > wrote:
> >>
> >> You're right, I must have mixed up my R versions when running
the
> example,
> >> as the problem seems to be resolved in R-devel.
> >>
> >> Sorry for the noise and thanks again for fixing this.
> >>
> >> Andrzej
> >>
> >> On Tue, Apr 25, 2017 at 3:55 PM, Michael Lawrence
> >> <lawrence.michael at gene.com> wrote:
> >>>
> >>> I attempted to fix it, and that example seems to work for me.
It's
> >>> also a (passing) regression test in R. Are you sure you're
using a new
> >>> enough R-devel?
> >>>
> >>>
> >>> On Tue, Apr 25, 2017 at 2:34 AM, Andrzej Ole? <andrzej.oles
at gmail.com>
> >>> wrote:
> >>> > Hi Michael,
> >>> >
> >>> > thanks again for your patch! I've tested it and
I'm happy to confirm
> >>> > that
> >>> > `callNextMethod()` works with methods dispatching on
`...`.
> >>> >
> >>> > However, the second issue I reported still seems to be
unresolved.
> >>> > Consider
> >>> > the following toy example, where the `f()` calls differ
in result
> >>> > depending
> >>> > on whether the dispatch happens on a formal argument or
the `...`
> >>> > argument.
> >>> >
> >>> >
> >>> >     f = function(x, ..., a = b) {
> >>> >       b = "missing 'a'"
> >>> >       print(a)
> >>> >     }
> >>> >
> >>> >     f()
> >>> >     ## [1] missing 'a'
> >>> >
> >>> >     f(a = 1)
> >>> >     ## [1] 1
> >>> >
> >>> >     setGeneric("f", signature = "x")
> >>> >
> >>> >     # works as the non-generic version
> >>> >     f()
> >>> >     ## [1] missing 'a'
> >>> >
> >>> >     setGeneric("f", signature =
"...")
> >>> >
> >>> >     # unexpectedly fails to find 'b'
> >>> >     f()
> >>> >     ## Error in print(a) : object 'b' not found
> >>> >
> >>> >
> >>> > Any chances of fixing this?
> >>> >
> >>> > Cheers,
> >>> > Andrzej
> >>> >
> >>> >
> >>> >
> >>> > On Fri, Apr 21, 2017 at 11:40 AM, Andrzej Ole? <
> andrzej.oles at gmail.com>
> >>> > wrote:
> >>> >>
> >>> >> Great, thanks Michael for you quick response!
> >>> >>
> >>> >> I started off with a question on SO because I was not
sure whether
> >>> >> this
> >>> >> was an actual bug or I was just missing something
obvious. I'm
> looking
> >>> >> forward to the patch.
> >>> >>
> >>> >> Cheers,
> >>> >> Andrzej
> >>> >>
> >>> >>
> >>> >> On Thu, Apr 20, 2017 at 10:28 PM, Michael Lawrence
> >>> >> <lawrence.michael at gene.com> wrote:
> >>> >>>
> >>> >>> Thanks for pointing out these issues. I have a
fix that I will
> commit
> >>> >>> soon.
> >>> >>>
> >>> >>> Btw, I would never have seen the post on Stack
Overflow. It's best
> to
> >>> >>> report bugs on the bugzilla.
> >>> >>>
> >>> >>> Michael
> >>> >>>
> >>> >>> On Thu, Apr 20, 2017 at 8:30 AM, Andrzej Ole?
> >>> >>> <andrzej.oles at gmail.com>
> >>> >>> wrote:
> >>> >>> > Hi all,
> >>> >>> >
> >>> >>> > I recently encountered some unexpected
behavior with S4 generics
> >>> >>> > dispatching on `...`, which I described in
> >>> >>> >
> >>> >>> >
> >>> >>> >
http://stackoverflow.com/questions/43499203/use-
> callnextmethod-with-dotsmethods
> >>> >>> >
> >>> >>> > TL;DR: `callNextMethod()` doesn't work
in methods dispatching on
> >>> >>> > `...`,
> >>> >>> > and
> >>> >>> > arguments of such methods are resolved
differently than the
> >>> >>> > arguments
> >>> >>> > of
> >>> >>> > methods dispatching on formal arguments.
> >>> >>> >
> >>> >>> > Could this indicate a potential problem with
the implementation
> of
> >>> >>> > the
> >>> >>> > `...` dispatch?
> >>> >>> >
> >>> >>> > Cheers,
> >>> >>> > Andrzej
> >>> >>> >
> >>> >>> >         [[alternative HTML version deleted]]
> >>> >>> >
> >>> >>> >
______________________________________________
> >>> >>> > R-devel at r-project.org mailing list
> >>> >>> >
https://stat.ethz.ch/mailman/listinfo/r-devel
> >>> >>
> >>> >>
> >>> >
> >>
> >>
> >
>
	[[alternative HTML version deleted]]

I ported that over.

On Tue, Aug 1, 2017 at 5:50 AM, Andrzej Ole? <andrzej.oles at gmail.com>
wrote:
> Thank you Michael for updating the 3.4 branch, the `callNextMethod()` now
> works for `...` methods as expected. However, I'm still missing your
other
> patch fixing the handling of arguments in `...` methods. It would be really
> great if this bugfix could be integrated into the 3.4 branch as well, such
> that the following code doesn't result in an error.
>
> Cheers,
> Andrzej
>
>
>     f = function(x, ..., a = b) {
>       b = "missing 'a'"
>       print(a)
>     }
>
>     f()
>     ## [1] missing 'a'
>
>     f(a = 1)
>     ## [1] 1
>
>     setGeneric("f", signature = "x")
>
>     # works as the non-generic version
>     f()
>     ## [1] missing 'a'
>
>     setGeneric("f", signature = "...")
>
>     # unexpectedly fails to find 'b'
>     f()
>     ## Error in print(a) : object 'b' not found
>
>
> On Fri, Jul 28, 2017 at 9:15 PM, Michael Lawrence <
> lawrence.michael at gene.com> wrote:
>
>> I pushed the patch to the 3.4 branch. Feel free to test.
>>
>> Michael
>>
>> On Wed, Jul 26, 2017 at 4:02 AM, Andrzej Ole? <andrzej.oles at
gmail.com>
>> wrote:
>> > Hi Michael,
>> >
>> > it seems that your patch to S4 generics dispatching on `...` is
still
>> > available only in R-devel, and was not included in the minor
R-3.4.1
>> > release. I was wondering what is the policy of incorporating bug
fixes
>> from
>> > the devel branch into release, and whether there is any chance
that the
>> > broken `...` dispatch is fixed before R-3.5.0?
>> >
>> > Cheers,
>> > Andrzej
>> >
>> >
>> > On Tue, Apr 25, 2017 at 4:15 PM, Andrzej Ole? <andrzej.oles at
gmail.com>
>> > wrote:
>> >>
>> >> You're right, I must have mixed up my R versions when
running the
>> example,
>> >> as the problem seems to be resolved in R-devel.
>> >>
>> >> Sorry for the noise and thanks again for fixing this.
>> >>
>> >> Andrzej
>> >>
>> >> On Tue, Apr 25, 2017 at 3:55 PM, Michael Lawrence
>> >> <lawrence.michael at gene.com> wrote:
>> >>>
>> >>> I attempted to fix it, and that example seems to work for
me. It's
>> >>> also a (passing) regression test in R. Are you sure
you're using a new
>> >>> enough R-devel?
>> >>>
>> >>>
>> >>> On Tue, Apr 25, 2017 at 2:34 AM, Andrzej Ole?
<andrzej.oles at gmail.com
>> >
>> >>> wrote:
>> >>> > Hi Michael,
>> >>> >
>> >>> > thanks again for your patch! I've tested it and
I'm happy to confirm
>> >>> > that
>> >>> > `callNextMethod()` works with methods dispatching on
`...`.
>> >>> >
>> >>> > However, the second issue I reported still seems to
be unresolved.
>> >>> > Consider
>> >>> > the following toy example, where the `f()` calls
differ in result
>> >>> > depending
>> >>> > on whether the dispatch happens on a formal argument
or the `...`
>> >>> > argument.
>> >>> >
>> >>> >
>> >>> >     f = function(x, ..., a = b) {
>> >>> >       b = "missing 'a'"
>> >>> >       print(a)
>> >>> >     }
>> >>> >
>> >>> >     f()
>> >>> >     ## [1] missing 'a'
>> >>> >
>> >>> >     f(a = 1)
>> >>> >     ## [1] 1
>> >>> >
>> >>> >     setGeneric("f", signature =
"x")
>> >>> >
>> >>> >     # works as the non-generic version
>> >>> >     f()
>> >>> >     ## [1] missing 'a'
>> >>> >
>> >>> >     setGeneric("f", signature =
"...")
>> >>> >
>> >>> >     # unexpectedly fails to find 'b'
>> >>> >     f()
>> >>> >     ## Error in print(a) : object 'b' not
found
>> >>> >
>> >>> >
>> >>> > Any chances of fixing this?
>> >>> >
>> >>> > Cheers,
>> >>> > Andrzej
>> >>> >
>> >>> >
>> >>> >
>> >>> > On Fri, Apr 21, 2017 at 11:40 AM, Andrzej Ole? <
>> andrzej.oles at gmail.com>
>> >>> > wrote:
>> >>> >>
>> >>> >> Great, thanks Michael for you quick response!
>> >>> >>
>> >>> >> I started off with a question on SO because I was
not sure whether
>> >>> >> this
>> >>> >> was an actual bug or I was just missing something
obvious. I'm
>> looking
>> >>> >> forward to the patch.
>> >>> >>
>> >>> >> Cheers,
>> >>> >> Andrzej
>> >>> >>
>> >>> >>
>> >>> >> On Thu, Apr 20, 2017 at 10:28 PM, Michael
Lawrence
>> >>> >> <lawrence.michael at gene.com> wrote:
>> >>> >>>
>> >>> >>> Thanks for pointing out these issues. I have
a fix that I will
>> commit
>> >>> >>> soon.
>> >>> >>>
>> >>> >>> Btw, I would never have seen the post on
Stack Overflow. It's
>> best to
>> >>> >>> report bugs on the bugzilla.
>> >>> >>>
>> >>> >>> Michael
>> >>> >>>
>> >>> >>> On Thu, Apr 20, 2017 at 8:30 AM, Andrzej Ole?
>> >>> >>> <andrzej.oles at gmail.com>
>> >>> >>> wrote:
>> >>> >>> > Hi all,
>> >>> >>> >
>> >>> >>> > I recently encountered some unexpected
behavior with S4 generics
>> >>> >>> > dispatching on `...`, which I described
in
>> >>> >>> >
>> >>> >>> >
>> >>> >>> >
http://stackoverflow.com/questions/43499203/use-callnextmeth
>> od-with-dotsmethods
>> >>> >>> >
>> >>> >>> > TL;DR: `callNextMethod()` doesn't
work in methods dispatching on
>> >>> >>> > `...`,
>> >>> >>> > and
>> >>> >>> > arguments of such methods are resolved
differently than the
>> >>> >>> > arguments
>> >>> >>> > of
>> >>> >>> > methods dispatching on formal arguments.
>> >>> >>> >
>> >>> >>> > Could this indicate a potential problem
with the implementation
>> of
>> >>> >>> > the
>> >>> >>> > `...` dispatch?
>> >>> >>> >
>> >>> >>> > Cheers,
>> >>> >>> > Andrzej
>> >>> >>> >
>> >>> >>> >         [[alternative HTML version
deleted]]
>> >>> >>> >
>> >>> >>> >
______________________________________________
>> >>> >>> > R-devel at r-project.org mailing list
>> >>> >>> >
https://stat.ethz.ch/mailman/listinfo/r-devel
>> >>> >>
>> >>> >>
>> >>> >
>> >>
>> >>
>> >
>>
>
>
	[[alternative HTML version deleted]]