Add object functions, modernize CI, and remove reg* pseudotypes

[jnasbyupgrade]

• Add object__cleanup()
• Add functions to retrieve object info
• Disallow tracking objects in temp schemas
• Mark partitioned table/index as unsupported
• Stop using reg* pseudotypes in tables
• Add README documenting the framework and API
• Switch to pgxn-tools based testing
• Update pgxntool

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: e6c254c5-038a-4f81-863c-e06082f26f0c

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[jnasbyupgrade]

@claude please review this PR.

[claude]

Claude finished @jnasbyupgrade's task in 7m 3s —— View job

---

Review of this PR

Todo list

• Gather context (diff vs origin/master, read changed SQL/README/CI files)
• Review SQL changes (object__cleanup, new functions, reg* pseudotype removal, partitioned table/index handling, temp schema exclusion)
• Review CI/build changes (ci.yml, pgxntool updates, Makefile)
• Review README documentation
• Post final review feedback

This is a solid PR overall — the object info functions (object__describe, object__identity), the temp-schema rejection check, the object__cleanup() trigger, and the reg* pseudotype removal are all clean, well-tested changes, and the new README is a good addition. A few issues worth fixing before merge:

1. object_reference.control still requires count_nulls (bug)

object_reference.control:5 still has:

requires = 'cat_tools, count_nulls'

This PR removes the count_nulls dependency from sql/object_reference.sql (the search_path DO block and the null_count trigger using not_null_count_trigger are both gone), and META.json/META.in.json correctly list only cat_tools under runtime requires. The control file wasn't updated to match, so CREATE EXTENSION object_reference will still try to pull in count_nulls unnecessarily.
Fix this →

2. CI failure detection likely never triggers (bug)

.github/workflows/ci.yml:17-26:

- name: Test on PostgreSQL ${{ matrix.pg }}
  run: make test PGUSER=postgres
  continue-on-error: true
- name: Output failed results
  run: |
    if [ -e test/regression.out ]; then
      ...
      exit 1

pgxntool/base.mk:122 (and pg_regress itself) writes test failures to regression.diffs, not regression.out. Since test/regression.out is not the file pg_regress uses to signal failure, this existence check will likely never be true, and combined with continue-on-error: true on the test step, the job will report green even when tests actually fail. Given "switch to pgxn-tools based testing" is a stated goal of this PR, this is worth double-checking — the check should probably look at test/regression.diffs instead.
Fix this →

3. _object_oid.object_oid is now a redundant duplicate column (simplification)

After removing the reg* pseudotype columns, _object_reference._object_oid has:

objid       oid NOT NULL
...
CONSTRAINT objid_must_match CHECK( objid IS NOT DISTINCT FROM object_oid )
...
object_oid  oid NOT NULL

and every insert path (_object_oid__add, _object_v__for_update) sets object_oid to the same value as objid. With the reg* columns gone, object_oid no longer serves a distinguishing purpose — it's always identical to objid. Worth dropping the column (and simplifying _object_v/_object_v__for_update) unless there's a forward-looking reason to keep it.
Fix this →

4. Pre-existing bug spotted nearby: undefined r_object in fix_refs()

Not introduced by this PR, but adjacent to the object-cleanup/group changes: in _object_reference.fix_refs() (sql/object_reference.sql), the WHEN r_object_v.ids_exist THEN branch references r_object.object_id in both its RAISE WARNING and RAISE statements, but only r_object_v is declared — there is no r_object variable. This branch will error at runtime ("missing FROM-clause entry for table r_object") whenever it's hit. Since you're already touching this file, might be worth a quick fix.
Fix this →

---

Note: this environment only allows a curated set of Bash commands without interactive approval (some git subcommands, Read/Grep/Glob), so I wasn't able to actually spin up PostgreSQL and run make test to confirm the CI-detection issue empirically — the analysis above is based on static reading of pgxntool/base.mk, ci.yml, and PostgreSQL/pg_regress conventions. Worth confirming by watching an actual CI run with a deliberately failing test.