Skip to content

A tool for scanning existing projects in search of potential signs of plagiarism

Notifications You must be signed in to change notification settings

DanySK/code-plagiarism-detector

Repository files navigation

Code Plagiarism Detector

A tool for scanning existing projects in search of potential signs of plagiarism.

Getting started

Prerequisites

In order to work, the following environment variables must be set:

Usage

CLI

Currently, the detector can be used running the JAR file. The tool has several options thanks to which is possible to configure it. You can see them specifying the --help option:

Usage: cli [OPTIONS] COMMAND [ARGS]...

  code-plagiarism-detector is a command line tool for scanning existing
  projects in search of potential signs of plagiarism.

Tokenization options:
  --language VALUE          Sources code language. Default: Java.
  --min-tokens INT          The minimum token length which should be reported
                            as a duplicate. Default: 15.
  --filter-threshold FLOAT  The cutoff threshold used to filter comparison
                            pairs.This could affect the detection
                            effectiveness.

Plain file options:
  --o, --output-dir VALUE  The path of the directory where to store the
                           reports.

Options:
  --technique-type [tokenization]  The technique used to analyze and detect
                                   similarities. Default: tokenization.
  --exporter-type [plain-text]     Output type format. Default: plain-text.
  --min-duplication FLOAT          The percentage of duplicated code in a
                                   source file under which matches are not
                                   reported. Default: 0.3.
  --exclude TEXT                   Comma separated list of files to be
                                   excluded from the check.
  -h, --help                       Show this message and exit
  --verbose                        Prints debug logs

Commands:
  submission  Submission options.
  corpus      Corpus Options.

To specify how to retrieve projects two subcommands are provided:

  • submission: set of projects whose authenticity is to be verified;
  • corpus: set of projects with which compare each of submission project.
Usage: cli submission [OPTIONS]

  Submission options.

Options:
  --service TEXT  A (list of) triple, possibly separated by commas, containing
                  a supported hosting service (github|bitbucket), the owner of
                  the repo and an optional repository name to search,
                  formatted: like this: `service-name:owner[/repo-name]`.
  --url VALUE     The URL addresses of the repositories to be retrieved,
                  possibly separated by commas.
  -h, --help      Show this message and exit

For example (equivalent for corpus subcommand):

  • to search by multiple URL:
    submission --url https://github.com/unibo-oop-projects/Student-Project-OOP-21-Bragari-Mennuti-Violani-Volfgit,https://github.com/unibo-oop-projects/Student-Project-OOP21-Bianchi-Ciccioni-stubborn
  • to search according to the following criteria: all GitHub repos owned by unibo-oop-projects and the Bitbucket repos named oop owned by danysk:
    submission --service github:unibo-oop-projects,bitbucket:danysk/oop

About

A tool for scanning existing projects in search of potential signs of plagiarism

Resources

Stars

Watchers

Forks

Packages

 
 
 

Contributors 4

  •  
  •  
  •  
  •