Module map_reduce 

Map-reduce pattern for large diff analysis

When diffs exceed the token threshold, this module splits analysis across files, then synthesizes results for accurate classification.

Structs

FileObservation

Observation from a single file during map phase

Functions

reduce_phase

Reduce phase: synthesize all observations into final analysis

run_map_reduce

Run full map-reduce pipeline for large diffs

should_use_map_reduce

Check if map-reduce should be used Always use map-reduce except for: