Skip to the content
Know your tech before you buy it. Makers: get listed from $49/mo

SourcererCC

github.com

  • Free plan
  • Linux
  • Self-hosted
7.1 Club score Visit github.com

What SourcererCC is

SourcererCC is a free, GPL-3.0 token-based detector for code clones in very large codebases and Internet-scale repositories. It can compare files, methods, statements, or blocks, and its tokenizer can be configured for any language through language settings and file extensions. Source code must be tokenized before detection. The documented default similarity threshold is 80%, and users can change it. File-level configuration ignores files below 65 tokens or above 500,000 tokens by default. A controller distributes detection across multiple nodes and aggregates the results. The project provides a virtual-machine artifact with prepared steps from raw source to a database of clone mappings. It also describes DéjàVu, a web tool for quick clone analysis. SourcererCC is intended for Linux and self-hosted deployment.

Who it suits

It suits teams analyzing large repositories that can run Linux or self-host the tooling. Users should be comfortable configuring tokenization and distributed processing.

Good to know, and what to watch

  • Detects clones at four code granularities
  • Configurable for any programming language
  • Adjustable similarity threshold
  • Virtual machine eases setup
  • GPL-3.0 open-source license
  • Source must be tokenized first
  • Default file bounds exclude very small or huge files
  • Mac multiprocessing requires Python 3.6
  • Designed for Linux and self-hosting

The verdict

SourcererCC offers configurable clone detection across large repositories and multiple code levels. Its tokenization workflow, distributed processing, and Mac limitation matter when planning deployment.

What the maker publishes

Detection scope
It can detect clones between files, methods, statements, or blocks, in any language. source
DéjàVu
The project says SourcererCC supports DéjàVu, a large-scale study of code cloning on GitHub, and describes DéjàVu as a web tool for clone analysis. source
File size bounds
The documented file-level clone detection configuration ignores files below 65 tokens or above 500000 tokens by default. source
Getting started
The project provides a virtual machine with instructions and programs for processing source code into a database with clone mappings. source
Granularity
It can detect clones between files, methods, statements, or blocks. source
Input preparation
Source code must first be tokenized, and the documented tokenizer accepts configured file extensions and project paths. source
Input size limits
The documented configuration ignores files below 65 tokens or above 500,000 tokens. source
Languages
The tokenizer can be configured with language comment syntax and a list of file extensions to analyze. source
Language support
The README says it can detect clones in any language, with tokenizer language settings and file extensions configured by the user. source
License
The repository identifies its license as GPL-3.0. source
License
The GitHub repository identifies its license as GPL-3.0. source
Notable platform limit
The README says the tokenizer's multiprocessing features do not work on Macs with Python versions later than 3.6 and recommends avoiding Mac for this workflow. source
Parallel processing
The controller splits the detection task across multiple nodes, whose results must be aggregated. source
Parallel processing
The tool splits clone detection across multiple nodes, whose output must be aggregated. source
Platform limitation
The tutorial says the tokenizer requires Python 3.6 on Mac because later versions have multiprocessing incompatibilities. source
Processing
Source code must be tokenized before SourcererCC runs, and the repository provides a tokenizer for that step. source
Purpose
SourcererCC is a token-based code clone detector for very large code bases and Internet-scale project repositories. source
Purpose
SourcererCC is a token-based code clone detector for very large code bases and Internet-scale project repositories. source
Similarity threshold
The documented default threshold flags clones at 80% similarity, and the threshold can be changed. source
Similarity threshold
The documented default flags clones at 80% similarity, and the threshold can be changed. source
Support
The README invites users to contact the project for questions or technical help, including performance tuning for their hardware. source
Support
The tutorial invites users to contact the team for technical help, including hardware performance tuning. source
VM artifact
The project describes a virtual machine artifact with pre-programmed steps from raw source code to a database with clone mappings and says it is an easy way to get started. source
Web tool
The README describes DéjàVu as a supporting web tool for quick and simple clone analysis. source

Checked against the maker's pages on 9 Oct 2026.