coerce_binop errors with keyword arguments

[lftabera]

On a coerced binary operator

@coerce_binop
__op__(x,y,**kwds)
}}

if y has to be coerced but x does not change its parent, the keyword arguments are not correctly passed.

A real example

{{{
sage: R1=QQ['x,y']
sage: R2=QQ['x,y,z']
sage: f=R1(1)
sage: h=R2(1)
sage: h.gcd(f,algorithm='modular')
TypeError                                 Traceback (most recent call last)
...
TypeError: algorithm algorithm not supported
}}}

Component: coercion

Keywords: coerce_binop, keyword

Author: Luis Felipe Tabera Alonso

Branch/Commit: u/lftabera/ticket/14186 @ 0cb642b

Reviewer: Robert Bradshaw, Travis Scrimshaw

Issue created by migration from https://trac.sagemath.org/ticket/14186

[lftabera]

comment:1

It is a one-line fix.

By the way, if a binary method with keyword values that is not decorated

__op__(self, y, method='z')

One can call it with x.__op__(y, 'preferred_method')

But once it is decorated this is not possible and key=value arguments are mandatory. Is this inteded?

[lftabera]

Author: Luis Felipe Tabera Alonso

[lftabera]

comment:3

Attachment: coer_binop_keywords.patch.gz

Update patch to new coercion framework.

[robertwb]

Reviewer: Robert Bradshaw

[robertwb]

comment:4

Looks good to me.

[robertwb]

comment:6

Lets make sure the tests pass though.

[lftabera]

comment:8

Robert,

Could you please take a look to the ticket? It is a one liner, you claimed that looked good and after seven months, patchbot has not complained.

[lftabera]

Branch: u/lftabera/ticket/14186

[sagetrac-git]

Commit: 0cb642b

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. New commits:

0cb642b

Trac #14186 coerce_binop errors with keyword arguments

[tscrim]

comment:11

Looks good to me.

[lftabera]

comment:13

Do not forget to add yoursef to the reviewers field

[lftabera]

Changed reviewer from Robert Bradshaw to Robert Bradshaw, Travis Scrimshaw