[LucaCappelletti94]

CREATE FUNCTION rejected the SQL-standard body that PostgreSQL 14 added. CREATE FUNCTION f() RETURNS INT LANGUAGE sql BEGIN ATOMIC SELECT 1; END failed with Expected: end of statement, found: BEGIN under both PostgreSqlDialect and GenericDialect, and pg_dump writes SQL functions with such bodies in exactly this form.

The PostgreSQL CREATE FUNCTION option loop now reads BEGIN ATOMIC and the statements up to END, and stores them in a new CreateFunctionBody::BeginAtomic variant that reuses BeginEndStatements. PostgreSQL writes no AS before the block, so a separate variant keeps it apart from the SQL Server AS BEGIN ... END body. A RETURN expr; inside the block parses as the existing Statement::Return.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 82.41%. Comparing base (e15e9ab) to head (a710be2).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2728      +/-   ##
==========================================
+ Coverage   82.40%   82.41%   +0.01%     
==========================================
  Files          42       42              
  Lines       35342    35362      +20     
  Branches    35342    35362      +20     
==========================================
+ Hits        29122    29145      +23     
+ Misses       2678     2675       -3     
  Partials     3542     3542              

☔ 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.

[sabir-akhadov-localstack]

PostgreSQL accepts the same body for procedures (CREATE PROCEDURE p() LANGUAGE sql BEGIN ATOMIC INSERT INTO t VALUES (1); END), which parse_create_procedure still rejects at AS. Fine as a follow-up.