[Merged by Bors] - refactor: rename `Rel` to `SetRel`, restore the old `Rel` by eric-wieser · Pull Request #27447 · leanprover-community/mathlib4

eric-wieser · 2025-07-24T23:13:03Z

Many users are confused that Rel no longer means what it used to. Since the design is not obvious (and the implementation is explicitly part of the interface thanks to abbrev), let's give it a name that makes the choice obvious.

This then restores the old definition of Rel, as an abbrev; it's conceivable that many users were just using this as a shorthand, and this stops them being broken by #25587. Such users will of course be broken if they used the old API.

This way, users bumping from 4.21.0 to 4.22.0 and using Rel only as a shorthand will not be broken.

#mathlib4 > Changing `Rel` to `Set (α × β)`, to help uniform spaces @ 💬

Many users are confused that Rel no longer means what it used to. Since the design is not obvious, lets put it in a namespace that makes the choice obvious.

github-actions · 2025-07-24T23:14:13Z

PR summary 1b235f6e15

Import changes for modified files

No significant changes to the import graph

Import changes for all files

Files	Import difference

Declarations diff

+ SetRel.FiniteDimensional.inv
+ SetRel.InfiniteDimensional.inv
+ SetRel.IsWellFounded.inv_of_finiteDimensional
+ SetRel.IsWellFounded.of_finiteDimensional
+ SetRel.exists_graph_eq_iff
+ SetRel.finiteDimensional_inv
+ SetRel.finiteDimensional_or_infiniteDimensional
+ SetRel.finiteDimensional_swap_iff
+ SetRel.infiniteDimensional_inv
+ SetRel.infiniteDimensional_swap_iff
+ SetRel.not_finiteDimensional_iff
+ SetRel.not_infiniteDimensional_iff
+ SetRel.wellFounded_of_finiteDimensional
+ SetRelation.is_graph_iff
+ instance {R : α → α → Prop} [IsIrrefl α R] : SetRel.IsIrrefl {(a, b) | R a b} := ‹_›
+ instance {R : α → α → Prop} [IsTrans α R] : SetRel.IsTrans {(a, b) | R a b} := ‹_›
+ instance {α : Type u} {β : Type v} [DecidableEq β] (R : SetRel α β)
+++ SetRel
+--- Rel
- Rel.FiniteDimensional.inv
- Rel.InfiniteDimensional.inv
- Rel.IsWellFounded.inv_of_finiteDimensional
- Rel.IsWellFounded.of_finiteDimensional
- Rel.exists_graph_eq_iff
- Rel.finiteDimensional_inv
- Rel.finiteDimensional_or_infiniteDimensional
- Rel.finiteDimensional_swap_iff
- Rel.infiniteDimensional_inv
- Rel.infiniteDimensional_swap_iff
- Rel.not_finiteDimensional_iff
- Rel.not_infiniteDimensional_iff
- Rel.wellFounded_of_finiteDimensional
- Relation.is_graph_iff
- instance {R : α → α → Prop} [IsIrrefl α R] : Rel.IsIrrefl {(a, b) | R a b} := ‹_›
- instance {R : α → α → Prop} [IsTrans α R] : Rel.IsTrans {(a, b) | R a b} := ‹_›
- instance {α : Type u} {β : Type v} [DecidableEq β] (R : Rel α β)

You can run this locally as follows

## summary with just the declaration names:
./scripts/declarations_diff.sh <optional_commit>

## more verbose report:
./scripts/declarations_diff.sh long <optional_commit>

The doc-module for script/declarations_diff.sh contains some details about this script.

No changes to technical debt.

You can run this locally as

./scripts/technical-debt-metrics.sh pr_summary

The relative value is the weighted sum of the differences with weight given by the inverse of the current value of the statistic.
The absolute value is the relative value divided by the total sum of the inverses of the current values (i.e. the weighted average of the differences).

YaelDillies · 2025-07-25T13:08:59Z

Would SetRel be better to avoid nameclash?

eric-wieser · 2025-07-25T14:54:45Z

Perhaps; can you create a zulip poll?

YaelDillies · 2025-08-05T14:32:34Z

SetRel is the clear winner of the poll

…to-set-rel

kim-em · 2025-08-13T00:27:54Z

bors d+

mathlib-bors · 2025-08-13T00:27:56Z

✌️ eric-wieser can now approve this pull request. To approve and merge a pull request, simply reply with bors r+. More detailed instructions are available here.

eric-wieser · 2025-08-13T00:55:24Z

bors merge

Many users are confused that Rel no longer means what it used to. Since the design is not obvious (and the implementation is explicitly part of the interface thanks to `abbrev`), let's give it a name that makes the choice obvious. This then restores the old definition of `Rel`, as an abbrev; it's conceivable that many users were just using this as a shorthand, and this stops them being broken by #25587. Such users will of course be broken if they used [the old API](https://github.com/leanprover-community/mathlib4/blob/59bd63c72671849996d64c5d3b5f24a36333483c%5E/Mathlib/Data/Rel.lean). This way, users bumping from 4.21.0 to 4.22.0 and using `Rel` only as a shorthand will not be broken. [#mathlib4 > Changing `Rel` to `Set (α × β)`, to help uniform spaces @ 💬](https://leanprover.zulipchat.com/#narrow/channel/287929-mathlib4/topic/Changing.20.60Rel.60.20to.20.60Set.20.28.CE.B1.20.C3.97.20.CE.B2.29.60.2C.20to.20help.20uniform.20spaces/near/528531852)

mathlib-bors · 2025-08-13T01:25:08Z

Pull request successfully merged into master.

Build succeeded:

CI Success

…-community#27447) Many users are confused that Rel no longer means what it used to. Since the design is not obvious (and the implementation is explicitly part of the interface thanks to `abbrev`), let's give it a name that makes the choice obvious. This then restores the old definition of `Rel`, as an abbrev; it's conceivable that many users were just using this as a shorthand, and this stops them being broken by leanprover-community#25587. Such users will of course be broken if they used [the old API](https://github.com/leanprover-community/mathlib4/blob/59bd63c72671849996d64c5d3b5f24a36333483c%5E/Mathlib/Data/Rel.lean). This way, users bumping from 4.21.0 to 4.22.0 and using `Rel` only as a shorthand will not be broken. [#mathlib4 > Changing &leanprover-community#96;Rel&leanprover-community#96; to &leanprover-community#96;Set (α × β)&leanprover-community#96;, to help uniform spaces @ 💬](https://leanprover.zulipchat.com/#narrow/channel/287929-mathlib4/topic/Changing.20.60Rel.60.20to.20.60Set.20.28.CE.B1.20.C3.97.20.CE.B2.29.60.2C.20to.20help.20uniform.20spaces/near/528531852)

refactor: move Rel within the Set namespace

46893fd

Many users are confused that Rel no longer means what it used to. Since the design is not obvious, lets put it in a namespace that makes the choice obvious.

eric-wieser added 2 commits July 24, 2025 23:24

add back the old rel as an abbrev

d33f82a

fix

c278223

eric-wieser changed the title ~~refactor: move Rel within the Set namespace~~ refactor: rename Rel to Set.Rel, restore the old Rel Jul 25, 2025

eric-wieser added the t-order Order theory label Jul 25, 2025

eric-wieser marked this pull request as ready for review July 25, 2025 10:25

eric-wieser requested review from YaelDillies and digama0 July 25, 2025 10:28

leanprover-community-bot-assistant assigned YaelDillies Aug 5, 2025

YaelDillies added the awaiting-author A reviewer has asked the author a question or requested changes. label Aug 5, 2025

eric-wieser added 3 commits August 12, 2025 23:25

Merge remote-tracking branch 'origin/master' into eric-wieser/rename-…

a70476e

…to-set-rel

rename again

a1c5bff

fix

20d2eb6

eric-wieser changed the title ~~refactor: rename Rel to Set.Rel, restore the old Rel~~ refactor: rename Rel to SetRel, restore the old Rel Aug 12, 2025

eric-wieser added 4 commits August 12, 2025 23:33

fix

a715d86

fix

497e28f

fix

2508487

fix

6ac21b2

eric-wieser added awaiting-CI This PR does not pass CI yet. This label is automatically removed once it does. and removed awaiting-author A reviewer has asked the author a question or requested changes. labels Aug 12, 2025

github-actions bot removed the awaiting-CI This PR does not pass CI yet. This label is automatically removed once it does. label Aug 13, 2025

ghost added the delegated This pull request has been delegated to the PR author (or occasionally another non-maintainer). label Aug 13, 2025

ghost added the ready-to-merge This PR has been sent to bors. label Aug 13, 2025

mathlib-bors bot changed the title ~~refactor: rename Rel to SetRel, restore the old Rel~~ [Merged by Bors] - refactor: rename Rel to SetRel, restore the old Rel Aug 13, 2025

mathlib-bors bot closed this Aug 13, 2025

mathlib-bors bot deleted the eric-wieser/rename-to-set-rel branch August 13, 2025 01:25

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Merged by Bors] - refactor: rename `Rel` to `SetRel`, restore the old `Rel`#27447

[Merged by Bors] - refactor: rename `Rel` to `SetRel`, restore the old `Rel`#27447
eric-wieser wants to merge 10 commits intomasterfrom
eric-wieser/rename-to-set-rel

eric-wieser commented Jul 24, 2025 •

edited

Loading

Uh oh!

github-actions bot commented Jul 24, 2025 •

edited

Loading

Declarations diff

Uh oh!

YaelDillies commented Jul 25, 2025

Uh oh!

eric-wieser commented Jul 25, 2025

Uh oh!

YaelDillies commented Aug 5, 2025

Uh oh!

kim-em commented Aug 13, 2025

Uh oh!

mathlib-bors bot commented Aug 13, 2025

Uh oh!

eric-wieser commented Aug 13, 2025

Uh oh!

mathlib-bors bot commented Aug 13, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

eric-wieser commented Jul 24, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions bot commented Jul 24, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR summary 1b235f6e15

Import changes for modified files

Declarations diff

Uh oh!

YaelDillies commented Jul 25, 2025

Uh oh!

eric-wieser commented Jul 25, 2025

Uh oh!

YaelDillies commented Aug 5, 2025

Uh oh!

kim-em commented Aug 13, 2025

Uh oh!

mathlib-bors bot commented Aug 13, 2025

Uh oh!

eric-wieser commented Aug 13, 2025

Uh oh!

mathlib-bors bot commented Aug 13, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

eric-wieser commented Jul 24, 2025 •

edited

Loading

github-actions bot commented Jul 24, 2025 •

edited

Loading