Rails 8.1 support by andynu · Pull Request #2480 · rsim/oracle-enhanced

andynu · 2025-10-23T20:10:03Z

This PR brings full Rails 8.1 compatibility to the Oracle Enhanced adapter, building on excellent work by @akostadinov (Aleksandar N. Kostadinov) in PR #2471.

I've added only two additional fixes:

CLOB insertion with disabled prepared statements,
removal of the obsolete ExplainSubscriber reference that Rails 8.1 refactored away.

Credits

The vast majority of this work was done by Aleksandar N. Kostadinov (@akostadinov) and Daria Mayorova (@mayorova). Their comprehensive compatibility updates fixed nearly all Rails 8.1 issues.

Changes from @akostadinov's PR #2471

Core Rails 8.1 API Compatibility

Column constructor signature change (Eliminate queries loading dumped model schema on Postgres rails/rails#54333)
- Updated OracleEnhanced::Column.new to accept cast_type as second parameter
- Modified schema_statements.rb and schema_creation.rb to call lookup_cast_type and pass cast types to column
  constructors
- Updated tests to use new column constructor signature
Schema dumping method rename
- Changed dump_schema_information → dump_schema_versions throughout codebase
- Updated method calls and comments in schema_statements.rb and test specs
Native database types refactoring
- Moved native_database_types to class method to support Rails 8.1's static type resolution
- Instance method now delegates to class method
- Maintains backward compatibility with emulate_booleans_from_strings setting
Distinct column API changes (by @mayorova)
- Updated distinct_relation_for_primary_key to use Rails 7.1+ distinct column API
- Changed internal method call from previous API to distinct(column_name)

Test Infrastructure & Coverage

Rails 7.1 composite primary key tests
- Added comprehensive test suite for composite primary keys (composite_spec.rb)
- Tests unique constraints, validations, and composite key behavior
- 84 lines of new test coverage
Test reliability improvements
- Fixed emulate_booleans test to properly clear type map and re-establish connections
- Changed from instance variable @conn to leased connections for thread safety
- Added proper connection cleanup in test teardown
Schema enhancement follow-up (Remove `OracleEnhancedAdapter.default_tablespaces[native_database_typ… #1544)
- Updated create_sequence call to use proper keyword arguments
- Modernized sequence creation syntax

My Contributions (this PR)

CLOB Insertion Fix

Fixed CLOB/BLOB insertion when prepared_statements: false (commit 0a4ad58)
- Added before_create and after_create callbacks to LOB module
- Previously only had after_update callbacks, causing data loss on insert
- When prepared_statements is disabled, SQL uses empty_clob()/empty_blob() literals
- Now properly writes LOB data after record creation
- Added 51 lines of test coverage in type/text_spec.rb
- Fixes CLOB/BLOB Inserts Fail When prepared_statements: false (which happens when config.active_record.query_log_tags_enabled = true) #2477
- ~~Uses ROWID for clob update statements to avoid issues with missing primary keys and/or composite primary keys. (e4b0156)~~

Rails 8.1 ExplainSubscriber Removal

Removed obsolete ActiveRecord::ExplainSubscriber reference (commit 85618af)
- Rails 8.1 removed ExplainSubscriber class
  (rails/rails@f488878)
- ExplainSubscriber functionality refactored into ExplainRegistry with lazy subscription
- Rails 8.1 now subscribes only when .explain is first called
- Removed manual subscription line from spec_helper.rb that was causing NameError

Testing

All tests pass against Rails 8.1 (8-1-stable branch).

yahonda · 2025-12-09T12:10:46Z

Would you drop these commits from this pull request? They should be for 8.0, not 8.1.

f34d8d6
0235b57
1153e51
b898c80

andynu · 2025-12-09T14:55:25Z

Certainly. Dropped.

andynu · 2025-12-09T15:34:06Z

Better. Rather than dropping, I just rebased this on top of master again. Since those commits were from earlier.

akostadinov · 2025-12-21T17:31:36Z

Amazing work @andynu!

I just want to point your attention also to #2483 as a big improvement of performance and memory usage for workloads involving LOBs. Basically I remove this write_lobs stuff altogether because it is completely redundant. This also allows much more efficient LOB reading from the server. IDK if you would want to amend your LOB fixes or maybe it can be merged/discussed further after this PR is merged. But wanted to point out to you. Or if you prefer I can create a pull request to your branch with it. Will be happy to get your feedback in either case.

akostadinov · 2025-12-30T16:23:06Z

          # After setting large objects to empty, select the OCI8::LOB
          # and write back the data.
+          before_create :record_lobs_for_create
+          after_create :enhanced_write_lobs


For create, shouldn't it just write all non-empty and non-null lobs instead of tracking them?

But anyway, I think this LOB callback writing stuff needs to go unless there are non-easily adaptable use-cases for disabling prepared statements as discussed in #2483

P.S. adding this here, doesn't it mean write_lobs was not used for a long time on create already? And that people just use prepared statements and don't need this?

yahonda

ff45875

Could you please move this commit to a separate pull request? This appears to be an issue with the Active Record Oracle enhanced adapter 8.0, which is independent from the 8.1 support work.

yahonda

7cbb567

Could you move this commit to a separate pull request? Regarding the content itself, I think the troubleshooting note might benefit from more specific guidance—as written, it feels a bit general to be actionable.

yahonda

e3e858a

Could you move this commit to a separate pull request? It doesn't seem required for Rails 8.1 support.

yahonda

3c15a11

Could you reference the Rails commit or changelog entry for this "activerecord 7.1 distinct column api change"?

andynu · 2026-01-26T19:28:56Z

Updated per review feedback. This PR is now stacked on:

Fix CLOB/BLOB insertion when prepared_statements is false #2485 — Fix CLOB/BLOB insertion when prepared_statements: false
Support ROWID-based LOB writes for tables without primary keys #2486 — Support ROWID-based LOB writes for tables without primary keys

I'm actively using this branch in production, and need those fixes, I could conceivably branch again for me needs, and flip the order of these PRs, but my preference would be for a review and decision on those two PRs before merging this one if at all possible. Also both stem from a rails 8.1 change (query_log_tags_enabled) forcing prepared_statements: false; So while the functionality under discussion there could have affected folks prior to rails 8.1, I think people will be more likely to run into it with rails 8.1.

Changes in this update:

Removed LOB-related commits (moved to Fix CLOB/BLOB insertion when prepared_statements is false #2485 and Support ROWID-based LOB writes for tables without primary keys #2486)
Removed README troubleshooting note (too vague to be useful)
Added Rails PR reference to the columns_for_distinct commit (Eager load composite primary keys models rails/rails#48490)

yahonda · 2026-01-29T07:52:47Z

Thank you for your work on this.

My request would be to keep this pull request minimal for Rails 8.1 support and separate those fixes　#2485 #2486 into their own pull requests, so that we can backport them to address issues occurring in 8.0 as well. If those fixes are included in this Rails 8.1 pull request, it would be difficult to cherry-pick them back to 8.0.

With that in mind, I'd prefer to first merge the minimal Rails 8.1 support, and then review those two pull requests separately afterward. I hope this approach works for you.

yahonda · 2026-01-29T15:38:09Z

In #2489, I temporarily changed the Rails branch in Oracle enhanced adapter's master from main to 8-0-stable. This should allow #2485 and #2486 to be reviewed before #2480.

andynu · 2026-01-29T15:49:59Z

Yasuo, Thank you. That is a kindness. I was going to attempt to re-stack the PRs with rails-8-1-support in front after work today; And still can, but if those other two PRs are acceptable, I do think they would go nicely with the rails-8-1-support PR, and would still allow for back-porting.

…

On Thu, Jan 29, 2026 at 10:38 AM Yasuo Honda ***@***.***> wrote: *yahonda* left a comment (rsim/oracle-enhanced#2480) <#2480 (comment)> In #2489 <#2489>, I temporarily changed the Rails branch in Oracle enhanced adapter's master from main to 8-0-stable. This should allow #2485 <#2485> and #2486 <#2486> to be reviewed before #2480 <#2480>. — Reply to this email directly, view it on GitHub <#2480 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAONGMJ35TUEP5Z5R5MPUD4JISPRAVCNFSM6AAAAACKBOBP22VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQMJYGQ3DKNZZHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

Add ability to write LOB data to tables that don't have a primary key by: 1. Exposing cursor.rowid method in OCI connection wrapper 2. Capturing ROWID after INSERT in exec_insert (@last_insert_rowid) 3. Using ROWID in write_lobs WHERE clause when no PK is available 4. Supporting composite primary keys (Array) in write_lobs The ROWID approach works because: - Ruby-oci8's cursor.rowid returns the ROWID of the last inserted row - ROWID uniquely identifies any row regardless of table structure - The after_create callback fires immediately after INSERT on same connection Also includes ORA-01741 diagnostic logging for empty column detection.

…ents This spec clearly demonstrates the critical difference between LOB handling with prepared statements (bind parameters) vs without (raw SQL with empty_clob()). Key points demonstrated: - With prepared_statements: true, LOB data flows through type_cast() which creates OCI8::CLOB temp LOBs. Data is populated BEFORE INSERT. - With prepared_statements: false, SQL contains empty_clob() literals. The write_lobs callback is REQUIRED to populate LOB data after INSERT. This test suite will FAIL if the lob.rb callbacks are removed, proving they are necessary for backwards compatibility with prepared_statements: false. Related to: rsim#2483

Rails 8.1 removed the ExplainSubscriber class (commit f488878f1bc), causing test failures with: NameError: uninitialized constant ActiveRecord::ExplainSubscriber The ExplainSubscriber functionality was refactored into ExplainRegistry with lazy subscription - instead of subscribing at initialization, Rails 8.1 now subscribes only when .explain is first called via ExplainRegistry.start. The spec_helper.rb line that manually subscribed to ExplainSubscriber is no longer needed or valid. Rails 8.1 handles the subscription automatically and lazily. Related Rails commit: - f488878f1bc "Refactor ExplainRegistry to only be subscribed once used" - Author: Jean Boussier - Date: Thu Sep 25 10:37:09 2025 +0200 - Link: rails/rails@f488878

Switch from pre-release alpha constraint to stable release constraint for ActiveRecord 8.1.0 dependency, following Rails 8.1 stable release.

Document Rails 8.1 support with installation instructions, including the gem version constraint for activerecord-oracle_enhanced-adapter 8.1.

andynu · 2026-02-16T16:19:46Z

After discussion with @yahonda on #2486, I've removed e3e858a "ROWID + composite key fix in database_statements.rb + oci_connection.rb" from this PR.

andynu · 2026-03-26T13:59:31Z

@yahonda Is there anything I can help with that might unblock a release?

andynu force-pushed the rails-8-1-support branch from f942020 to e834b46 Compare October 23, 2025 21:16

andynu force-pushed the rails-8-1-support branch 2 times, most recently from 72a17f3 to e4b0156 Compare December 2, 2025 19:29

andynu added the dependencies Pull requests that update a dependency file label Dec 3, 2025

This was referenced Dec 3, 2025

Fix CLOB/BLOB insert failure when prepared_statements disabled (release80) #2479

Closed

Fix CLOB/BLOB insert failure when prepared_statements disabled (master) #2478

Closed

andynu force-pushed the rails-8-1-support branch from e4b0156 to 29a404c Compare December 9, 2025 14:52

andynu force-pushed the rails-8-1-support branch 3 times, most recently from e458c74 to e3e858a Compare December 9, 2025 15:29

This was referenced Dec 16, 2025

Preliminary Rails 8.1 support #2471

Closed

No double writing of lobs #2483

Closed

akostadinov reviewed Dec 30, 2025

View reviewed changes

akostadinov mentioned this pull request Jan 12, 2026

columns_for_distinct might need some adjustments #2482

Closed

yahonda reviewed Jan 26, 2026

View reviewed changes

andynu force-pushed the rails-8-1-support branch from 151e629 to 942c50e Compare January 26, 2026 19:27

andynu force-pushed the rails-8-1-support branch from 942c50e to fdb161d Compare January 26, 2026 19:36

This was referenced Jan 30, 2026

Rails 8.1 support (without CLOB fixes) #2491

Merged

Support ROWID-based LOB writes for tables without primary keys #2486

Closed

andynu force-pushed the rails-8-1-support branch 2 times, most recently from bc88c65 to 151e629 Compare February 4, 2026 15:03

andynu and others added 13 commits February 4, 2026 10:05

Switch to Rails 8.1 (8-1-stable branch) for compatibility testing

847d100

Rails 8.1 API compatibility with rails/rails#54333

399aa7f

activerecord 7.1 distinct column api change

7c225d6

test rails 7.1 composite indices

c63ef65

follow through on rsim#1544

05dd78f

Rails compat - #native_database_types

0580832

Rails #dump_schema_information -> #dump_schema_versions

16fa631

emulate_booleans test reliable on newer Rails

4cc4c9f

Update ActiveRecord dependency to stable 8.1.0 release

45be32a

Switch from pre-release alpha constraint to stable release constraint for ActiveRecord 8.1.0 dependency, following Rails 8.1 stable release.

Add Rails 8.1 installation section to README

81e1a6a

Document Rails 8.1 support with installation instructions, including the gem version constraint for activerecord-oracle_enhanced-adapter 8.1.

andynu force-pushed the rails-8-1-support branch from 151e629 to 81e1a6a Compare February 4, 2026 15:05

Conversation

andynu commented Oct 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Credits

Changes from @akostadinov's PR #2471

My Contributions (this PR)

Uh oh!

yahonda commented Dec 9, 2025

Uh oh!

andynu commented Dec 9, 2025

Uh oh!

andynu commented Dec 9, 2025

Uh oh!

akostadinov commented Dec 21, 2025

Uh oh!

akostadinov Dec 30, 2025

Choose a reason for hiding this comment

Uh oh!

yahonda left a comment

Choose a reason for hiding this comment

Uh oh!

yahonda left a comment

Choose a reason for hiding this comment

Uh oh!

yahonda left a comment

Choose a reason for hiding this comment

Uh oh!

yahonda left a comment

Choose a reason for hiding this comment

Uh oh!

andynu commented Jan 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

yahonda commented Jan 29, 2026

Uh oh!

yahonda commented Jan 29, 2026

Uh oh!

andynu commented Jan 29, 2026 via email

Uh oh!

andynu commented Feb 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

andynu commented Mar 26, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

andynu commented Oct 23, 2025 •

edited

Loading

andynu commented Jan 26, 2026 •

edited

Loading

andynu commented Feb 16, 2026 •

edited

Loading