A hands-on workshop in diagnosing unfamiliar software, testing changes, and turning a user request into a useful feature. The application lets you import tabular data, inspect it, and clean it in a local browser.
Start with EXERCISE.md for the workshop tasks.
Install JDK 21–26, Apache Maven, Node.js 24 or newer, and npm 11.16.0 or newer.
The Java version limits are defined in pom.xml; frontend requirements are in main/webapp/package.json.
git clone https://github.com/datarootsio/data-cleaning-workshop.git
cd data-cleaning-workshop
./refine build
./refineOpen http://127.0.0.1:3333. On Windows, use refine.bat build and refine.bat.
Stop the application with Ctrl+C in its terminal.
Use ./refine -d /path/to/workshop-data to keep workshop projects in a separate data directory.
Use ./refine -p 3334 if the default port is occupied.
- Exercise: tasks and what to bring to the discussion.
- Contributing: branches, validation, and review.
- Agent guidance: repository layout and commands.
- UI tests: end-to-end test setup.
This repository is maintained for the dataroots workshop. Use its issue tracker for workshop questions, or ask your facilitator during the session.
This workshop is derived from OpenRefine, created by Metaweb Technologies, Inc. and originally conceived and written by David Huynh. It became Google Refine after Google's acquisition of Metaweb in 2010, then the community project OpenRefine in 2012. The application includes work by its many contributors; existing authorship and copyright notices are retained. This workshop is maintained independently of that project.
The code is distributed under the BSD-3-Clause license in LICENSE.txt.
Third-party notices remain in packaging/THIRD-PARTY.txt,
main/webapp/licenses, and the extension license directories.