Rectangle Decomposition Primer

Three strategies for handling overlapping 2D rectangles.

The Core Problem

When storing overlapping 2D rectangles, you must decide how to handle intersections. Three strategies exist with different trade-offs in insertion complexity, query complexity, and storage characteristics.

Strategy Comparison

Approach	Insert	Query	Storage (example)	Best for
LWW	Simple	Simple	2 ranges	Single property only
Shallow Merge	Complex	Simple	4 ranges	Properties always together
Spatial Join	Simple	Join	3 ranges (across 2 indexes)	Independent properties

Strategy Details

1. Last-Writer-Wins (LWW)

New values replace old values in overlaps.

✅ Pro: Simple inserts, minimal storage
❌ Con: Single value per coordinate
📖 Read more: rectangle-decomposition-lww.md

Use when: Single property per cell.

2. Shallow Merge

Merge object properties in overlaps using spread operator.

✅ Pro: Multiple properties per coordinate
❌ Con: More complex inserts, more storage fragments
📖 Read more: rectangle-decomposition-merge.md

Use when: Properties always updated together.

3. Spatial Join

Separate index per property, combine at query time.

✅ Pro: Simple inserts, independent property updates
❌ Con: Query-time join operation required
📖 Read more: rectangle-decomposition-spatial-join.md

Use when: Properties updated independently.