Deprecate is_FiniteField etc., make sage.rings.finite_rings a namespace package

Question

Deprecate is_FiniteField etc., make sage.rings.finite_rings a namespace package

Closed this issue 3 years ago · 30 comments

Part of meta-ticket #32414.

Deprecating is_FiniteField, is_PrimeFiniteField, and is_FiniteFieldElement is done for uniformity but is not needed to support modularization. #32432 shows that sage.rings.finite_rings.finite_field_base (which defines the ABC FiniteField) can be imported without pulling in implementation classes.

sage.rings.finite_rings must become a namespace package because element implementations depend on various libraries.

CC: @dimpase @kwankyu @fchapoton @tscrim

Component: refactoring

Author: Matthias Koeppe

Issue created by migration from https://trac.sagemath.org/ticket/32664

Answer 1 · 2021-10-22T00:08:37.000Z

Description changed:

--- 
+++ 
@@ -1 +1,4 @@
 Part of meta-ticket #32414.
+
+This is done for uniformity but is not needed to support modularization. #32432 shows that `finite_field_base` (which defines `is_FiniteField`) can be imported without pulling in implementation classes.
+

Answer 2 · 2022-12-05T21:33:30.000Z

Changed dependencies from #32612 to none

Answer 3 · 2022-12-05T21:33:30.000Z

Description changed:

--- 
+++ 
@@ -1,4 +1,5 @@
 Part of meta-ticket #32414.
 
-This is done for uniformity but is not needed to support modularization. #32432 shows that `finite_field_base` (which defines `is_FiniteField`) can be imported without pulling in implementation classes.
+Deprecating `is_FiniteField` is done for uniformity but is not needed to support modularization. #32432 shows that `sage.finitefinite_field_base` (which defines the ABC `FiniteField`) can be imported without pulling in implementation classes.
 
+`sage.rings.finite_rings` must become a namespace package because element implementations depend on various libraries.

Answer 4 · 2022-12-05T21:35:15.000Z

Description changed:

--- 
+++ 
@@ -1,5 +1,5 @@
 Part of meta-ticket #32414.
 
-Deprecating `is_FiniteField` is done for uniformity but is not needed to support modularization. #32432 shows that `sage.finitefinite_field_base` (which defines the ABC `FiniteField`) can be imported without pulling in implementation classes.
+Deprecating `is_FiniteField` is done for uniformity but is not needed to support modularization. #32432 shows that `sage.rings.finite_rings.finite_field_base` (which defines the ABC `FiniteField`) can be imported without pulling in implementation classes.
 
 `sage.rings.finite_rings` must become a namespace package because element implementations depend on various libraries.

Answer 5 · 2022-12-05T22:21:30.000Z

Branch: u/mkoeppe/add_sage_rings_abc_finitefield__deprecate_is_finitefield

Answer 6 · 2022-12-05T22:22:01.000Z

Description changed:

--- 
+++ 
@@ -1,5 +1,5 @@
 Part of meta-ticket #32414.
 
-Deprecating `is_FiniteField` is done for uniformity but is not needed to support modularization. #32432 shows that `sage.rings.finite_rings.finite_field_base` (which defines the ABC `FiniteField`) can be imported without pulling in implementation classes.
+Deprecating `is_FiniteField` and `is_PrimeFiniteField` is done for uniformity but is not needed to support modularization. #32432 shows that `sage.rings.finite_rings.finite_field_base` (which defines the ABC `FiniteField`) can be imported without pulling in implementation classes.
 
 `sage.rings.finite_rings` must become a namespace package because element implementations depend on various libraries.

Answer 7 · 2022-12-05T22:22:01.000Z

Commit: 8af57c4

Answer 8 · 2022-12-05T22:22:01.000Z

New commits:

`8af57c4`	`sage.rings.finite_rings: Deprecate is_FiniteField, is_PrimeFiniteField`

Answer 9 · 2022-12-05T23:00:16.000Z

Author: Matthias Koeppe

Answer 10 · 2022-12-05T23:23:18.000Z

Description changed:

--- 
+++ 
@@ -1,5 +1,5 @@
 Part of meta-ticket #32414.
 
-Deprecating `is_FiniteField` and `is_PrimeFiniteField` is done for uniformity but is not needed to support modularization. #32432 shows that `sage.rings.finite_rings.finite_field_base` (which defines the ABC `FiniteField`) can be imported without pulling in implementation classes.
+Deprecating `is_FiniteField`, `is_PrimeFiniteField`, and `is_FiniteFieldElement` is done for uniformity but is not needed to support modularization. #32432 shows that `sage.rings.finite_rings.finite_field_base` (which defines the ABC `FiniteField`) can be imported without pulling in implementation classes.
 
 `sage.rings.finite_rings` must become a namespace package because element implementations depend on various libraries.

Answer 11 · 2022-12-05T23:29:30.000Z

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

`98b433c`	`sage.rings.finite_rings: Make it a namespace package`
`6323775`	`src/sage/rings/finite_rings/finite_field_constructor.py: Restore re-export of is_FiniteField`
`473e9d4`	`src/sage/rings/finite_rings/hom_finite_field.pyx: Import ABC as FiniteField_base`
`b82fb1e`	`sage.rings.finite_rings: Deprecate is_FiniteFieldElent`

Answer 12 · 2022-12-05T23:29:30.000Z

Changed commit from 8af57c4 to b82fb1e

Answer 13 · 2022-12-06T00:30:19.000Z

Changed commit from b82fb1e to 106b22e

Answer 14 · 2022-12-06T00:30:19.000Z

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

`106b22e`	`Fixups`

Answer 15 · 2022-12-06T01:58:13.000Z

Changed commit from 106b22e to de5cfad

Answer 16 · 2022-12-06T01:58:13.000Z

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

`de5cfad`	`Fixups`

Answer 17 · 2022-12-06T03:04:08.000Z

Changed commit from de5cfad to 429b372

Answer 18 · 2022-12-06T03:04:08.000Z

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

`429b372`	`Fixup: Elements of finite fields are not instances of FieldElement`

Answer 19 · 2022-12-06T03:26:14.000Z

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

`e633d8f`	`src/sage/crypto/mq/rijndael_gf.py: Fix imports`

Answer 20 · 2022-12-06T03:26:14.000Z

Changed commit from 429b372 to e633d8f

Answer 21 · 2022-12-06T04:30:38.000Z

Changed commit from e633d8f to 886bc1a

Answer 22 · 2022-12-06T04:30:38.000Z

Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:

`886bc1a`	`src/sage/crypto/mq/rijndael_gf.py: Fix imports`

Answer 23 · 2022-12-06T12:03:23.000Z

comment:16

shouldn't this branch be based on #34804 ?

Answer 24 · 2022-12-18T21:41:15.000Z

Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:

`53ede8c`	`sage.rings.finite_rings: Deprecate is_FiniteField, is_PrimeFiniteField`
`8d4d8e5`	`sage.rings.finite_rings: Make it a namespace package`
`6ce26bc`	`src/sage/rings/finite_rings/finite_field_constructor.py: Restore re-export of is_FiniteField`
`98a5484`	`src/sage/rings/finite_rings/hom_finite_field.pyx: Import ABC as FiniteField_base`
`2e6b5ef`	`sage.rings.finite_rings: Deprecate is_FiniteFieldElent`
`3be7c97`	`Fixups`
`97a363e`	`Fixups`
`0c01403`	`Fixup: Elements of finite fields are not instances of FieldElement`
`9f4d1c4`	`src/sage/crypto/mq/rijndael_gf.py: Fix imports`

Answer 25 · 2022-12-18T21:41:15.000Z

Changed commit from 886bc1a to 9f4d1c4

Answer 26 · 2022-12-18T21:42:22.000Z

comment:18

Replying to Dima Pasechnik:

shouldn't this branch be based on #34804 ?

There's no relation. #34804 merges cleanly with this branch

Answer 27 · 2023-01-07T02:44:20.000Z

Changed commit from 9f4d1c4 to 9ebadcf

Answer 28 · 2023-01-07T02:44:20.000Z

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

`9ebadcf`	`Merge tag '9.8.beta6' into t/32664/add_sage_rings_abc_finitefield__deprecate_is_finitefield`

Answer 29 · 2023-02-10T02:13:18.000Z

Tests ran successfully in https://github.com/sagemath/sagetrac-mirror/actions/runs/3860089783 (before the migration).

Needs review

Answer 30 · 2023-02-13T19:37:31.000Z

Removed branch from issue description; replaced by PR #35119