Distributed Grep with the MapReduce Framework

This repository already contains a coordinator/worker MapReduce runtime (src/mr). The new mrapps/grep.go plugin lets you run a distributed grep that scans many input files in parallel.

Quick start (from the src directory)

• Build the grep plugin: go build -buildmode=plugin -o mrapps/grep.so mrapps/grep.go
• Start the coordinator with your input files (choose any pg-*.txt or your own): go run main/mrcoordinator.go main/pg-*.txt
• In one or more other shells, launch workers with the pattern to search for: MR_GREP_PATTERN="Sherlock Holmes" go run main/mrworker.go mrapps/grep.so
• When the workers finish, inspect the results: cat mr-out-*

How it works

• Workers read each file line by line and emit matches where the line contains MR_GREP_PATTERN.
• Each match is recorded as <filename>:<line number>:<line contents>; the reduce step sorts all matches for stable output.
• MR_GREP_PATTERN must be set for every worker process (coordinator does not need it).