feat: Add SQL support for multi-condition joins#698
feat: Add SQL support for multi-condition joins#698makarandhinge wants to merge 3 commits intoapache:mainfrom
Conversation
- Enhanced WayangMultiConditionJoinVisitor with SQL implementation - Updated JdbcJoinOperator to generate AND clauses for multi-column joins - Added test coverage for multi-condition join SQL generation - Maintains backward compatibility with single-condition joins
|
Hi, it doesn't seem like your test case tests the implementation |
|
Thank you @mspruc for the feedback! You’re right |
- Execute generated SQL against HSQLDB to verify actual join behavior - Validate row count (2 matching rows expected) - Verify data integrity of joined results - Test edge case: order with mismatched customer_id is correctly excluded - Confirm multi-condition AND clause works end-to-end
|
Hi @mspruc, Can you review this, if any improvement or changes required please mention. |
|
Hi, |
- Add sqlApiMultiConditionJoinGeneratesJdbcSql test - Validates SQL parsing through WayangMultiConditionJoinVisitor - Verifies withSqlImplementation() sets correct table names and field names - Confirms comma-separated field format for multi-condition joins - Proves end-to-end flow: SQL → Wayang plan → JDBC operators
|
Hi @mspruc, Thanks for your patience! I’ve added sqlApiMultiConditionJoinGeneratesJdbcSql in SqlToWayangRelTest.java to cover SQL API parsing, multi-condition JOIN, table names, and field names. As a new contributor, I really appreciate your guidance - please let me know if any changes are needed. Thanks again! |
|
Hi @mspruc, Can you review this, if any improvement or changes required please mention. |
|
Hi, why have you changed several |
|
Hi @mspruc , Thanks for pointing this out. I changed My intention was to avoid any potential incompatibility with Calcite’s internal handling of lists (especially since Calcite relies heavily on Guava’s That said, there is no strong technical constraint that strictly requires Thanks |
Todo #545