Improved Kash interface

[BrentBaccala]

The current interface to the optional kash package has limited utility. No way is provided to convert Kash expressions to Sage expressions, and Kash names like sage1 might be re-used even if there are outstanding references to subexpressions like sage1.1 (so this is also a bug fix, in addition to an enhancement).

This ticket fixes several bugs in the Kash interface, and allows a translation dictionary to be passed to a Kash object's sage() method, mapping Kash objects to Sage objects, allowing conversions like this:

    sage: kQ = kash.RationalField()
    sage: kR = kQ.PolynomialAlgebra()

    sage: R.<x> = QQ[]
    sage: ka = (x^2+x).subs({x : kR.1})
    sage541.1^2 + sage541.1
    sage: ka.sage({kR.1: x})
    x^2 + x

Also, keyword arguments are now passed correctly to Kash functions.

Component: interfaces: optional

Keywords: kash

Author: Brent Baccala

Branch/Commit: c8f01e7

Reviewer: Dima Pasechnik

Issue created by migration from https://trac.sagemath.org/ticket/25488

[BrentBaccala]

comment:1

In the course of preparing this ticket, I realized that the default bool() routine in interface/interface.py is almost certainly wrong and would always need to be overridden.

Why? Since the default implementation of is_zero() (in structure/element.pyx) returns not self, almost everything needs to type-convert to a boolean True, except for zero, which should type-convert to False.

interface.py's default bool() routine has it backwards - it tests for equality to _true_symbol() (and everything else returns False), instead of inequality to _false_symbol() (and everything else returns True).

Maybe we should open another ticket for that. For Kash, where we have separate boolean and integer types, we need to compare to both FALSE and 0, so I think we need to override the method no matter what.

---

New commits:

00463de	improve Kash interface
5768ed4	improve Kash interface
16289fc	Trac #25488: fix boolean conversion in kash interface

[BrentBaccala]

Commit: 16289fc

[BrentBaccala]

Author: Brent Baccala

[BrentBaccala]

Branch: public/25488

[sagetrac-git]

Changed commit from 16289fc to 84e58af

[sagetrac-git]

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

84e58af

Trac #25488: evade bizarre dictionary problem

[sagetrac-git]

Changed commit from 84e58af to c629174

[sagetrac-git]

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

c629174

Trac #25488: fix kash _contains() method

[dimpase]

comment:4

We are in fact thinking about removing support for kash completely.
The last release of kash was 10 years ago, and as it's only binary releases, there are increasingly fewer computers out there capable of running it.

[BrentBaccala]

comment:5

Replying to @dimpase:

We are in fact thinking about removing support for kash completely.
The last release of kash was 10 years ago, and as it's only binary releases, there are increasingly fewer computers out there capable of running it.

Removing kash makes a lot of sense.

The problem is that it's the only thing we've got right now to compute bases for Riemann-Roch spaces over QQ.

Kwankyu Lee has written some new function field code that can handle Riemann-Roch calculations over finite fields, but even that code is taking forever to move into the released code base.

Somebody (maybe me) needs to expand the function field code to handle QQ, and get it into master, before we remove kash.

Although, I admit that next to nobody knows how to use kash to do those calculations.

[sagetrac-git]

Changed commit from c629174 to b557a97

[sagetrac-git]

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

b557a97

Trac #25488: allow optional keyword arguments to be passed to Kash functions

[sagetrac-git]

Changed commit from b557a97 to a802923

[sagetrac-git]

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

a802923

Trac #25488: handle Sage values (like elements of QQbar) whose print representation can't be parsed

[fchapoton]

comment:8

this is missing the optional kash tag:

kash.function_call('ComplexToPolar', [1+I], {'Results' : 1})

[sagetrac-git]

Changed commit from a802923 to f2c1a1f

[sagetrac-git]

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

2b1bd8d	Merge tag '8.7' into public/25488
f2c1a1f	Trac #25488: add optional kash tag to test case

[embray]

comment:11

Moving tickets from the Sage 8.8 milestone that have been actively worked on in the last six months to the next release milestone (optimistically).

[embray]

comment:12

Ticket retargeted after milestone closed

[mkoeppe]

comment:13

Batch modifying tickets that will likely not be ready for 9.1, based on a review of the ticket title, branch/review status, and last modification date.

[mkoeppe]

comment:14

The patchbot complains about some docstring issues

[sagetrac-git]

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

4a8a8a6	Merge branch 'public/25488' in 9.2.rc2
c8f01e7	flake8 refreshment for kash interface

[sagetrac-git]

Changed commit from f2c1a1f to c8f01e7

[fchapoton]

comment:16

patchbot is now green

[dimpase]

Reviewer: Dima Pasechnik

[vbraun]

Changed branch from public/25488 to c8f01e7