[LucaCappelletti94]

The parser does not recognize DROP RULE, so migration scripts and schema dumps that remove rewrite rules fail to parse. This is the companion of #2433, and the two are independent and can land in either order.

Doc: https://www.postgresql.org/docs/current/sql-droprule.html

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 82.87%. Comparing base (94fa700) to head (5c50f9a).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2434      +/-   ##
==========================================
+ Coverage   82.85%   82.87%   +0.01%     
==========================================
  Files          42       42              
  Lines       35935    35968      +33     
  Branches    35935    35968      +33     
==========================================
+ Hits        29774    29807      +33     
  Misses       2654     2654              
  Partials     3507     3507              

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[s5dsn-eqee]

Like it, Span especially

Could you add a few more red tests, e.g.:

#[test]
fn parse_drop_rule_rejects_lists_and_extra_options() {
    for (sql, expected) in [
        ("DROP RULE r1, r2 ON t", "Expected: ON, found: ,"),
        ("DROP RULE s.r ON t", "Expected: ON, found: ."),
        ("DROP RULE 1 ON t", "Expected: identifier, found: 1"),
        ("DROP RULE IF EXISTS", "Expected: identifier, found: EOF"),
        ("DROP RULE IF NOT EXISTS r ON t", "Expected: ON, found: NOT"),
        ("DROP RULE r t", "Expected: ON, found: t"),
        ("DROP RULE r ON (t)", "Expected: identifier, found: ("),
        ("DROP RULE r ON t1, t2", "Expected: end of statement, found: ,"),
        ("DROP RULE r ON t ON t", "Expected: end of statement, found: ON"),
        ("DROP RULE r ON t CASCADE RESTRICT", "Expected: end of statement, found: RESTRICT"),
        ("DROP RULE r ON t RESTRICT CASCADE", "Expected: end of statement, found: CASCADE"),
        ("DROP RULE r ON t PURGE", "Expected: end of statement, found: PURGE"),
    ] {
        let err = all_dialects().parse_sql_statements(sql).unwrap_err().to_string();
        assert!(err.starts_with(&format!("sql parser error: {expected}")), "{sql}: {err}");
    }
}

[s5dsn-eqee]

👍

[LucaCappelletti94]

Added all the tests, but I have not understood what is there to like in particular of the Span? Unfortunately most of the spans are objectively to be refactored, likely including this one, but that is something I will only touch at some point after the tokenizer refactoring

[s5dsn-eqee]

Mainly just that Spans weren’t forgotten 😅
Agreed that Spans and error handling both have room for improvement

[s5dsn-eqee]

LGTM

[LucaCappelletti94]

Thanks @s5dsn-eqee and @sabir-akhadov-localstack for your reviews, I will be merging this now.

cc @alamb