Stage 5: Group By & Having

Stage Objective

Implement the GROUP BY clause and the HAVING filter using a Hash Aggregation approach.

Conceptual Algorithms

• Extend the SQL Parser to recognize GROUP BY <columns> and HAVING <condition>.
• In the Query Planner, introduce a HashAggregationNode that sits above the data scan.
• The HashAggregationNode will use an in-memory Hash Table to group rows.
• The hashing key is the concatenated bytes of the grouped columns.
• For each row from the child node, probe the hash table. If the group exists, update the running aggregates. If it doesn't, insert a new entry.
• Once the child node is fully consumed, the HashAggregationNode begins yielding the grouped rows, applying the HAVING condition before yielding.

Implementation Checklist

• Add GROUP BY and HAVING tokens to Lexer.
• Parse GROUP BY and HAVING into the AST.
• Update explain to show HashAggregationNode.
• Implement a generic Hash Table structure for memory.
• Implement HashAggregation execution logic.
• Verify execution outputs correctly grouped data.

Expected Contract

When a GROUP BY statement is executed, the engine should group results and apply aggregates per group, and filter groups using HAVING.