Release DataFusion 52.5.0 (minor) Release (Apr 2026)

[alamb]

Is your feature request related to a problem or challenge?

• Part of [EPIC] DataFusion Releases #19783
• Follow on to Release DataFusion 52.4.0 (minor/) Release (Mar 2026) #20855

I strongly suspect we'll end up with a few more bugs on the branch-52 release, so I am proactively opening a ticket to propose / track such a release

Steps:

• Update version and changelog: [branch-52] Update version to 52.5.0 and add changelog #21407
• Voting Thread: https://lists.apache.org/thread/scnnmsw6g200ckjj1rgx58oj9q6xzb2h
• Publish to crates.io: https://lists.apache.org/thread/kl19synnmoj4z7mhlf3mywox9pvnn0w8
• Port changes / changelog to main: port 52.5.0 changelog to main #21553

Content:

Requested Bug Fixes. Format: Issue / Requester / Backport PR

• Incorrect ORDER BY in unparser after DF 52 upgrade #20905 @alamb @krinart / [branch-52] Restore Sort unparser guard for correct ORDER BY placement (#20658) #21358
• Filter pushdown past children with fetch limits causes correctness bug #21063 / @shivbhatia10 / [branch-52] Fix push_down_filter for children with non-empty fetch fields (#21057) #21141
• Substrait join consumer produces incorrect logical plan #21124 / @gabotechs / [branch-52] Substrait join consumer should not merge nullability of join keys (#21121) #21161
• Fix incorrect regex pattern in regex_replace_posix_groups #19827 / @alamb @GaneshPatil7517 / [branch-52] Fix incorrect regex pattern in regex_replace_posix_groups (#19827) #21396
• fix: Fix panic in regexp_like() #20200 / @alamb @neilconway / [branch-52] fix: Fix panic in regexp_like() (#20200) #21397
• fix: use spill writer's schema instead of the first batch schema for spill files #21293 / @alamb @gruuya / [branch-52] fix: use spill writer's schema instead of the first batch schema for spill files (#21293) #21403
• FFI schema providers fail when views are registered #21155 / @timsaucer / [BRANCH-52] fix: foreign inner ffi types #21439

[hareshkh]

@alamb : Can we add the following issues to this release please?

• Filter pushdown past children with fetch limits causes correctness bug #21063 (cc @shivbhatia10)
  • Backport PR - [branch-52] Fix push_down_filter for children with non-empty fetch fields (#21057) #21141
• Substrait join consumer produces incorrect logical plan #21124 (cc @gabotechs)
  • Backport PR - [branch-52] Substrait join consumer should not merge nullability of join keys (#21121) #21161

[alamb]

@alamb : Can we add the following issues to this release please?

Done -- thank you @hareshkh

[alamb]

I'll plan to make a RC next week hopefully

[alamb]

I went through the list of commits on main and found some panics / regressions i think would be good to fix on branch-52 and added them to this list.