Merger Module¶

API reference for merge strategy implementations.

Merger Strategies¶

Each merge strategy class inherits from either BaseMerger or BaseLLMMerger.

Always use the incoming (latest) data.

from ontomem.merger import KeepIncomingMerger

merger = KeepIncomingMerger()
result = merger.merge(existing, incoming)  # Returns incoming

Always preserve the first (existing) data.

from ontomem.merger import KeepExistingMerger

merger = KeepExistingMerger()
result = merger.merge(existing, incoming)  # Returns existing

Non-null overwrites, lists append (default).

from ontomem.merger import FieldMerger

merger = FieldMerger()
result = merger.merge(existing, incoming)  # Smart merge

Require llm_client for intelligent synthesis.

LLM synthesis preferring new data on conflicts.

LLM synthesis preferring existing data on conflicts.

LLM synthesis balancing both perspectives.

See Merge Strategies Guide for usage examples.