JPlag - Detecting Software Plagiarism

JPlag finds pairwise similarities among a set of multiple programs. It can reliably detect software plagiarism and collusion in software development, even when obfuscated. All similarities are calculated locally, and no source code or plagiarism results are ever uploaded to the internet. JPlag supports a large number of programming and modeling languages.

• 📈 JPlag Demo

• 🏛️ JPlag on Helmholtz RSD

• 🤩 Give us Feedback in a short (<5 min) survey

Supported Languages

All supported languages and their supported versions are listed below.

Language	Version	CLI Argument Name	state	parser
Java	21	java	mature	JavaC
C	11	c	legacy	JavaCC
C++	14	cpp	beta	ANTLR 4
C#	6	csharp	mature	ANTLR 4
Python	3.6	python3	beta	ANTLR 4
JavaScript	ES6	javascript	beta	ANTLR 4
TypeScript	~5	typescript	beta	ANTLR 4
Go	1.17	golang	beta	ANTLR 4
Kotlin	1.3	kotlin	beta	ANTLR 4
R	3.5.0	rlang	beta	ANTLR 4
Rust	1.60.0	rust	beta	ANTLR 4
Swift	5.4	swift	beta	ANTLR 4
Scala	2.13.8	scala	beta	Scalameta
LLVM IR	15	llvmir	beta	ANTLR 4
Scheme	?	scheme	legacy	JavaCC
EMF Metamodel	2.25.0	emf	beta	EMF
EMF Model	2.25.0	emf-model	alpha	EMF
SCXML	1.0	scxml	alpha	XML
Text (naive)	-	text	legacy	CoreNLP

Download and Installation

You need Java SE 21 to run or build JPlag.

Downloading a release

• Download a released version.
• In case you depend on the legacy version of JPlag we refer to the legacy release v2.12.1 and the legacy branch.

Via Maven

JPlag is released on Maven Central, it can be included as follows:

<dependency>
  <groupId>de.jplag</groupId>
  <artifactId>jplag</artifactId>
  <version><!--desired version--></version>
</dependency>

Building from sources

1. Download or clone the code from this repository.
2. Run mvn clean package from the root of the repository to compile and build all submodules. Run mvn clean package assembly:single instead if you need the full jar which includes all dependencies. Run mvn -P with-report-viewer clean package assembly:single to build the full jar with the report viewer. In this case, you'll need Node.js installed.
3. You will find the generated JARs in the subdirectory cli/target.

Usage

JPlag can either be used via the CLI or directly via its Java API. For more information, see the usage information in the wiki. If you are using the CLI, you can display your results via jplag.github.io. No data will leave your computer!

CLI

Note that the legacy CLI is varying slightly. The language can either be set with the -l parameter or as a subcommand (jplag [jplag options] <language name> [language options]). A subcommand takes priority over the -l option. When using the subcommand, language-specific arguments can be set. A list of language-specific options can be obtained by requesting the help page of a subcommand (e.g. jplag java -h).

Parameter descriptions: 
      [root-dirs[,root-dirs...]...]
                        Root-directory with submissions to check for plagiarism.
      -bc, --bc, --base-code=<baseCode>
                        Path to the base code directory (common framework used in all submissions).
  -l, --language=<language>
                        Select the language of the submissions (default: java). See subcommands below.
  -M, --mode=<{RUN, VIEW, RUN_AND_VIEW}>
                        The mode of JPlag: either only run analysis, only open the viewer, or do both (default: null)
  -n, --shown-comparisons=<shownComparisons>
                        The maximum number of comparisons that will be shown in the generated report, if set to -1 all comparisons will be shown (default: 500)
      -new, --new=<newDirectories>[,<newDirectories>...]
                        Root-directories with submissions to check for plagiarism (same as root).
      --normalize       Activate the normalization of tokens. Supported for languages: Java, C++.
      -old, --old=<oldDirectories>[,<oldDirectories>...]
                        Root-directories with prior submissions to compare against.
  -r, --result-file=<resultFile>
                        Name of the file in which the comparison results will be stored (default: results). Missing .zip endings will be automatically added.
  -t, --min-tokens=<minTokenMatch>
                        Tunes the comparison sensitivity by adjusting the minimum token required to be counted as a matching section. A smaller value increases the sensitivity but might lead to more
                          false-positives.

Advanced
      --csv-export      Export pairwise similarity values as a CSV file.
  -d, --debug           Store on-parsable files in error folder.
  -m, --similarity-threshold=<similarityThreshold>
                        Comparison similarity threshold [0.0-1.0]: All comparisons above this threshold will be saved (default: 0.0).
  -p, --suffixes=<suffixes>[,<suffixes>...]
                        comma-separated list of all filename suffixes that are included.
  -P, --port=<port>     The port used for the internal report viewer (default: 1996).
  -s, --subdirectory=<subdirectory>
                        Look in directories <root-dir>/*/<dir> for programs.
  -x, --exclusion-file=<exclusionFileName>
                        All files named in this file will be ignored in the comparison (line-separated list).

Clustering
      --cluster-alg, --cluster-algorithm=<{AGGLOMERATIVE, SPECTRAL}>
                        Specifies the clustering algorithm (default: spectral).
      --cluster-metric=<{AVG, MIN, MAX, INTERSECTION}>
                        The similarity metric used for clustering (default: average similarity).
      --cluster-skip    Skips the cluster calculation.

Subsequence Match Merging
      --gap-size=<maximumGapSize>
                        Maximal gap between neighboring matches to be merged (between 1 and minTokenMatch, default: 6).
      --match-merging   Enables merging of neighboring matches to counteract obfuscation attempts.
      --neighbor-length=<minimumNeighborLength>
                        Minimal length of neighboring matches to be merged (between 1 and minTokenMatch, default: 2).

Subcommands (supported languages):
  c
  cpp
  csharp
  emf
  emf-model
  go
  java
  javascript
  kotlin
  llvmir
  python3
  rlang
  rust
  scala
  scheme
  scxml
  swift
  text
  typescript

Java API

The new API makes it easy to integrate JPlag's plagiarism detection into external Java projects:

Language language = new JavaLanguage();
Set<File> submissionDirectories = Set.of(new File("/path/to/rootDir"));
File baseCode = new File("/path/to/baseCode");
JPlagOptions options = new JPlagOptions(language, submissionDirectories, Set.of()).withBaseCodeSubmissionDirectory(baseCode);

try {
    JPlagResult result = JPlag.run(options);

    // Optional
    ReportObjectFactory reportObjectFactory = new ReportObjectFactory(new File("/path/to/output"));
    reportObjectFactory.createAndSaveReport(result);
} catch (ExitException e) {
    // error handling here
} catch (FileNotFoundException e) {
    // handle IO exception here
}

Contributing

We're happy to incorporate all improvements to JPlag into this codebase. Feel free to fork the project and send pull requests. Please consider our guidelines for contributions.

Contact

If you encounter bugs or other issues, please report them here. For other purposes, you can contact us at jplag@ipd.kit.edu . If you are doing research related to JPlag, we would love to know what you are doing. Feel free to contact us!

More information can be found in our Wiki!