finding-duplicate-functions

To find semantic duplicate functions in a codebase, you can run a script that extracts a function catalog, clusters functions by intent using LLM categorization, and generates a consolidation plan. This surfaces implementations that share intent but differ in naming or implementation.

Semantic duplication detection identifies functions that perform the same task but are named or implemented differently. It is applied to multi-contributor or LLMed codebases to reduce redundancy by categorizing extracted function catalogs by domain intent.

You audit a large project by extracting a function catalog, categorizing it by domain, and performing semantic-duplicate detection. This produces categorized groups and a final consolidation report to help merge multiple implementations onto a single, well-tested function.

Yes, duplicate function analysis generates traceable outputs including catalog.json, categorized.json, category-level outputs under a categories directory, and a final report to document the semantic duplication detection process.

Yes, semantic duplicate detection is designed for use before a major rewrite. It identifies multiple implementations of the same business logic across a codebase, allowing you to consolidate them onto a single, well-tested function.

LLM clustering for function analysis relies on extracting a function catalog and categorizing by domain, which may require manual review of the generated consolidation plan and category-level outputs to verify intent accuracy before refactoring.