An example of how to build the Theia-based tools including graphical editos, form-based editors, tree-based editors, textual DSLs, model analyisis, debugging and more. The coffee editor is part of the emf.cloud project. Please visit the emf.cloud home page for an overview of all features and an online live demonstration!
If you have questions, contact us on our discussions page and have a look at our communication and support options.
To run the coffee editor locally, we recommend to use a Docker image.
To build the docker container run the following command once:
docker build -t "coffee-editor:latest" .
or use the provided yarn script:
yarn build:docker
Now you can start the Theia Demo app using the following command (or change the host port to your preferences):
docker run -it -p 0.0.0.0:3000:3000 coffee-editor:latest
or use the provided yarn script:
yarn docker:run
Next, open a browser pointing to http://localhost:3000/#/coffee-editor/client/workspace/SuperBrewer3000 in your host.
You should see a Theia application with an example project being loaded.
Refer to the Getting Started widget in the right area of the application for more details on how to use the example project.
The docker image includes the Theia application and the following VSCode plugins (among others):
- Clangd for C/C++ editing support
- Debug for C/C++ debugging support
- Java LSP
- Java Debug
The runtime environment has installed (among others)
- CMake for building C/C++ projects
- GBD for debugging C/C++ projects
The coffee-editor consists of a frontend and a backend.
The frontend is located in the client/ folder and frontend specific documentation can be found in the client README
The backend is located in the backend/ folder and backend specific documentation can be found in the backend README
We are relying on a bunch of projects:
- https://github.com/eclipsesource/jsonforms
- https://github.com/eclipse-glsp/glsp
- https://github.com/eclipse-emfcloud/emfcloud-modelserver
- https://github.com/eclipse-emfcloud/emfcloud-modelserver-theia
- https://github.com/eclipse-emfcloud/theia-tree-editor
If you encounter issues please report them in the corresponding project. This project should not contain much code and should mostly consist of 'glue' code to combine the different components.
The following libraries/frameworks need to be installed on your system:
| Java | 17 |
| Maven | >=3.8.6 |
| Node | >=14.18.0 |
You need Java 17 to build the Coffee Editor.
Install nvm
curl -o- https://raw.githubusercontent.com/creationix/nvm/v0.33.5/install.sh | bash
nvm install 16
nvm use 16
npm install -g yarn
Please check the installation documentation for maven.
On Ubuntu you can use:
sudo apt-get install maven
sudo apt-get install g++-4.8 libsecret-1-dev xvfb libx11-dev libxkbfile-dev libxml2-utils
Please check the installation description here.
On Windows the most reliable way seems to be to install Python and set npm config set python "C:\Path\To\python.exe".
Clone the coffee-editor:
git clone https://github.com/eclipsesource/coffee-editor.git
cd coffee-editor
Build the coffee-editor:
yarn build
Run the built coffee-editor:
yarn start
Open http://localhost:3000 in the browser.
In Theia open the example workspace client/workspace/SuperBrewer3000 and double click a .coffee file. This opens it in a tree master detail editor.
Create a npm user and login to the npm registry, more on npm publishing.
npm login
Publish packages with lerna to update versions properly across local packages, more on publishing with lerna.
npx lerna publish
- Install Eclipse
- Import projects from
backend - Set target
org.eclipse.emfcloud.coffee.target.target - All backend applications are bundles as eclipse products and can be found in
org.eclipse.emfcloud.coffee.product
To debug the code generation you would need to attach to the process when it is started from the backend.
Alternatively you can run the codegen.product or codegen_cpp.product with the correct parameters.
In order to debug, start the workflowanalyzer.product product in debug mode. The root application is : org.eclipse.emfcloud.coffee.workflow.analyzer.application.application.
Please make sure to set the --WF_ANALYZER=5083 parameter to the browser app. In the backend you need to pass -port and -host with parameters as application arguments.
You can also simply use the predefined WorkflowAnalyzerServer.launch run config.
In order to debug, start the workflowdsl.product product in debug mode.
Please make sure to add --WF_LSP=5017 parameter to the browser app. In the backend you need to pass -startSocket as an application argument.
You can also use the predefined RunSocketServer-Headless.launch run config.
Use the CoffeeModelServer launch config to start the Model Server.
Use the WorkflowGLSPServer launch config to start the GLSP Server.
- Install VSCode
- Open folder
client
Use the Launch Browser Backend launch config inside VSCode.
Use the Launch Browser Frontend launch config inside VSCode. This will open Chrome and attach a debug listener.
This project's repository has branches prefixed with examples/ that serve to demonstrate common use cases of expansion or customization of
the capabilities of the frameworks on which the Coffee Editor is based. Each of these comprises a single commit that can be compared with its
base to see the changes from end to end that accomplish some developer use case. Often these examples will cross-cut the whole Coffee Editor
technology stack, but some will have a smaller focus. It all depends on the use case. The names of these branches are succinct but descriptive
of the use cases that they illustrate.
Because these examples are branches and not integrated into the main-line, they will be based on some past commit in the history of the project. It is strongly encouraged when contributing changes to the Coffee Editor that, where such changes would meaningfully impact a use case that is manifest in one of these example branches, that such example branch be rebased and updated as a part of that contribution. Thus it is important that users of these example branches not base their own work on them (in git terms) because they will be force-pushed from time to time, rewriting their history to rebase them on more recent baselines of the Coffee Editor.