crate/cratedb-examples - Githubissues

# CrateDB Examples ✨ _A collection of clear and concise examples how to work with [CrateDB]._ ✨ 🔗 Quick links: [Application](./application) • [Dataframe](./by-dataframe) • [Language](./by-language) • [Testing](./testing) • [Topic](./topic) 📖 More information: [Drivers and Integrations](https://cratedb.com/docs/clients/) • [Integration Tutorials](https://community.cratedb.com/t/overview-of-cratedb-integration-tutorials/1015) • [Reference Documentation](https://cratedb.com/docs/crate/reference/)

👨‍💻 Usage

You can explore the content by browsing folders within the repository. Main sections can be explored by using the quick links in the header area.
If you are looking for something specific, please use GitHub search, for example, searching for "jdbc".
You can use the code snippets for educational and knowledge base purposes, or as blueprints within your own projects.
The repository is also used to support QA processes. Each example is designed to be invoked as an integration test case, accompanied by a corresponding CI validation job.

🧐 What's inside

This section gives you an overview about what's inside the relevant folders.

by-dataframe contains example code snippets how to work with dataframe libraries like pandas, Polars, Dask, Spark, and friends.
by-language contains demo programs / technical investigations outlining how to get started quickly with CrateDB using different programming languages and frameworks.
application contains integration scenarios with full-fledged applications and software frameworks.
testing contains reference implementations about how to use different kinds of test layers for testing your applications with CrateDB.
topic mostly contains Jupyter Notebooks outlining different use cases around working with time-series data, and demonstrating machine learning technologies together with CrateDB.

✅ CI Status

Please visit the Build Status page to inspect the build status of relevant drivers, applications, and integrations for CrateDB, on one page.

🏕️ Testing

In the same way as on CI, you can invoke the example programs easily on your workstation, in order to quickly get started on behalf of working example code, or to verify connectivity within your computing environment.

Prerequisites

For invoking the software integration tests, you will need installations of Docker, Python, and Git on your workstation.

Before running the tests, make sure to supply an instance of CrateDB. In order to use and verify the most recent available code, let's select the OCI image crate/crate:nightly.

docker run --rm -it --pull=always \
    --name=cratedb --publish=4200:4200 --publish=5432:5432 \
    --env=CRATE_HEAP_SIZE=4g \
    crate/crate:nightly -Cdiscovery.type=single-node

Test Runner `ngr`

The repository uses a universal test runner to invoke test suites of different languages and environments, called ngr.

In order to run specific sets of test cases, you do not need to leave the top-level directory, or run any kind of environment setup procedures. If all goes well, just select one of the folders of interest, and invoke ngr test on it, like that:

ngr test by-language/java-jdbc
ngr test by-language/python-sqlalchemy
ngr test by-language/php-amphp
ngr test by-dataframe/dask
ngr test application/apache-superset
ngr test testing/testcontainers/java
ngr test topic/machine-learning/llm-langchain

[!NOTE] It is recommended to invoke ngr from within a Python virtualenv, in order to isolate its installation from the system Python. Installing ngr works like this:
git clone https://github.com/crate/cratedb-examples
cd cratedb-examples
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Test Matrix Support

Some examples optionally obtain parameters on invocation time.

One example is the test suite for Npgsql, which accepts the version number of the Npgsql driver release to be obtained from the environment at runtime, overriding any internally specified versions. Example:

ngr test by-language/csharp-npgsql --npgsql-version=6.0.9

[!TIP] This feature is handy if you are running a test matrix, which is responsible for driving the version numbers, instead of using the version numbers nailed within local specification files of any sort.

💁 Contributing

Interested in contributing to this project? Thanks so much for your interest!

As an open-source project, we are always looking for improvements in form of contributions, whether it be in the form of a new feature, improved infrastructure, or better documentation.

Your bug reports, feature requests, and patches are greatly appreciated.