Skip to main content

AI-Driven Unit Test Generation for Java

Project description

JavaTestGenie

MIT License PyPI

Demo

Click on the image to watch the video Video

Installation

pip install java-test-genie

Usage

Before running genie, set your OpenAI API key with export OPENAI_API_KEY=<API-KEY-HERE>. You can obtain it at https://platform.openai.com/account/api-keys

Ensure that you are in a at the same level with src/ when running the tool.

You need to create a config file named .genie.json within your java repository. This file should be at the same level with src/ as well as where you will run genie. In this file, you can specify which directories & files to include and exclude. The example below includes the service/ directory but excludes service/interface. Similarly, you can specify file names.

{
  "include": [
    {
      "parent_dir": "src/main/java/com/ase/restservice",
      "dir_names": ["service"],
      "file_names": []
    }
  ],
  "exclude": [
    {
      "parent_dir": "src/main/java/com/ase/restservice",
      "dir_names": ["service/interface"],
      "file_names": []
    }
  ]
}

The file is case-sensitive. Ensure that you do not list files in the "exclude" section unless they already match a path in the "include" section. Excluding a path will delete all matches, even if a specific class is explicitly listed in the "include" section. For instance, if you exclude "model/" but include "model/Account.java", it will not work. Instead, simply include "model/Account.java" and leave the "exclude" section empty, as the "model/" directory is already skipped.

Finally, you can run it with:

genie

Example

Here is an example repo that you can use the config file above with.

git clone https://github.com/boraelci/kaiserschmarrn.git

Arguments

  1. --help: Displays the help message and lists all available arguments.

  2. --mode: Allows you to switch between two modes:

    • methods (default): Generates test methods for individual methods within the input class. This mode is suitable for large input classes.
    • classes: Generates a complete test class, including imports, package name, and setup. This mode provides more comprehensive functionality but is not suitable for large input classes. If the input class is too large, you may encounter an error like Error: this file is too large.
  3. --do_overwrite: Enables automatic overwriting of existing test files without asking for confirmation. By default, this is set to false to protect existing test files. To enable automatic overwriting, just include --do_overwrite without specifying a parameter.

  4. --batch-size: Specifies the number of methods to process at the same time. It defaults to 15, but you can adjust this value based on your requirements.

Limitations

Currently, only 1 class per Java file is supported. The tool may produce errors if you supply a path that corresponds to a file with more than 1 class in it.

Grammar

Parsing Java files with the Tree-sitter library requires grammar files. They were obtained in the following way for this tool.

For macOS

git clone https://github.com/tree-sitter/tree-sitter-java.git
cd tree-sitter-java
tree-sitter generate
gcc -shared -o libtree-sitter-java.dylib -Isrc src/parser.c

For Windows

Used the one provided as java-grammar.so

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

java-test-genie-1.0.tar.gz (2.5 MB view details)

Uploaded Source

File details

Details for the file java-test-genie-1.0.tar.gz.

File metadata

  • Download URL: java-test-genie-1.0.tar.gz
  • Upload date:
  • Size: 2.5 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.10.2

File hashes

Hashes for java-test-genie-1.0.tar.gz
Algorithm Hash digest
SHA256 d5acdc4594a643a3547d2da12a32b7fd26c8ca3fa3b6a23cd8ec8c408251c965
MD5 04ce0730933664614ae45d72427138fc
BLAKE2b-256 83314468734d1f106529e20157a1c6f13ee472a77711d0572b1775b24d0511fa

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page