contrib/ivorysql_ora: refuse non-UTF8 test encodings up front

[muzimu217]

Fixes #1883.

Root cause

The five failures are not product bugs — every one of them is correct behaviour for a single-byte test database. The expected results encode UTF8-specific facts:

• ora_character truncates multibyte literals against VARCHAR2 byte limits (lengthb(cast('中文' AS varchar2(5 byte))) = 3); on LATIN1 a byte is a character, so the truncation point legitimately differs.
• ora_character_datatype_functions checks length/lengthc character counts, chr() codepoints that only exist in Unicode (chr(769), chr(2392), chr(119070) → "requested character too large"), UTF8-only normalization (compose/decompose → "Unicode normalization can only be performed if server encoding is UTF8"), unistr() and to_multi_byte(), plus psql column alignment that depends on character counts.
• ora_misc_functions exercises compose/decompose the same way.
• datatype_and_func_bugs inserts '测试' into char(3 char); under LATIN1 the literal is six characters, so "value too long" is the correct answer.
• utl_url escapes multibyte input and converts between character sets, so its multibyte expectations are UTF8-bound by nature (and a LATIN1 → GB18030 conversion does not exist in PostgreSQL).

A single-byte Oracle database could not store these literals either, so there is no LATIN1 behaviour to pin down: the tests are about multibyte semantics, and those only exist under UTF8.

Fix

Since the suite is inherently UTF8-bound, make that requirement explicit at the suite level instead of letting a non-UTF8 run surface as five seemingly unrelated failures. contrib/ivorysql_ora/Makefile already pins ENCODING = UTF8 by default; this adds a guard so an explicit non-UTF8 override fails up front with a clear message rather than running (and failing) the suite:

Makefile:78: *** ivorysql_ora regression tests require ENCODING=UTF8 (got 'LATIN1'); the expected outputs are only valid for a UTF8 test database.  Stop.

Targets that do not run the regression suite (all, install, ...) are unaffected.

Verification

oracle-check on a clean tree (Linux aarch64):

invocation	result
make -C contrib/ivorysql_ora oracle-check ENCODING=LATIN1	refused up front with the message above (previously: 5 of 28 tests failed)
LC_ALL=en_US.UTF-8 make -C contrib/ivorysql_ora oracle-check	All 28 tests passed
LC_ALL=C make -C contrib/ivorysql_ora oracle-check	All 28 tests passed

Summary by CodeRabbit

• Bug Fixes
  • Regression tests now consistently use UTF-8 encoding.
  • Invalid non-UTF-8 encoding overrides are rejected for Oracle compatibility test commands, preventing misleading test results.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Advanced

Run ID: 9e29c21a-4110-4659-abc2-218acc23e5ba

📥 Commits

Reviewing files that changed from the base of the PR and between 481c238 and 3685363.

📒 Files selected for processing (1)

• contrib/ivorysql_ora/Makefile

Included review availability: Your plan provides up to 4 included reviews per hour; 3 remain after this review.

---

📝 Walkthrough

Walkthrough

The Oracle regression Makefile now pins the test database encoding to UTF8 and rejects non-UTF8 overrides for the oracle-check and oracle-installcheck goals.

Changes

Oracle regression encoding

Layer / File(s)	Summary
Encoding pin and validation contrib/ivorysql_ora/Makefile	The Makefile sets ENCODING = UTF8, documents the UTF8-dependent tests, and rejects non-UTF8 overrides for the Oracle regression goals.

Priority: ➖ Normal

Estimated code review effort: 1 (Trivial) | ~5 minutes

Severity of issue fixed: Medium

Merge Risk: ⚪ Minimal · up to 36853

Oracle regression runs now consistently use UTF8 and clearly reject unsupported encoding overrides, preventing encoding-dependent test failures without affecting other targets.

Suggested reviewers: bigplaice, jiaoshuntian, hs-liuxh

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the main change: rejecting non-UTF8 test encodings in the contrib/ivorysql_ora regression suite.
Linked Issues check	✅ Passed	The changes satisfy issue #1883 by enforcing the suite's UTF8 requirement and rejecting explicit non-UTF8 overrides before regression tests run. The default UTF8 behavior and unrelated make targets re…
Out of Scope Changes check	✅ Passed	The changes are limited to the contrib/ivorysql_ora Makefile and directly support the linked issue. No unrelated code changes are present.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0…

✨ 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.}

[hanjianqiao]

Thanks for the PR!