#455 #456 #348 Update coffee-editor to latest versions and run servers as JAR products

[ndoschek]

#455 Update to Model Server V2 and latest GLSP version

• Migrate Model Server to latest version and to V2
• Migrate GLSP Server to latest version 1.1.0
• Reuse EMF server and modelserver glsp integration code and remove obsolete classes

#456 Consider running backend services as JAR products

• Run Model Server product as Java application instead of Eclipse product
  • Package Model Server as standalone JAR
• Run GLSP Server product as Java application instead of Eclipse product
  • Package GLSP server as standalone JAR

General

• Move example project 'SuperBrewer3000' to client/workspace and adapt all paths accordingly
• Rename frontend folder to client and adapt all paths

Client

• Update client code to Theia 1.27.0 an current versions of glsp, sprotty and emfcloud components
• Update to newer typescript version and adapt code accordingly
• Update theia plugins to newest compatible versions to ensure cpp and java debugging
• Reuse tsconfig, eslint and prettier configs by eclipse-glsp
• Align build scripts, devDependencies and package.json descriptions
• Fix custom file icons alignment
• Migrate run.sh to nodejs #348 Replace run.sh script with ts-node scripts and yarn scripts
• Create parent package.json that offers the main build scripts build:server and build:client as well as a convenience script that combines those two
• Move copying of backend products in the build of the repsonsible client package coffee-servers and replace with ts-node script
• Move all server backend registrations to coffee-servers package
• Fix ESLint warnings and leftovers
• Update Getting Started
  • Enhance styling and replace outdated Theia variables
  • Structure Java and CPP features in detail sections
  • Improve link styling, behaviour and add titles
• Add EclipseSource application branding
• Align icons in diagram editor and coffee tree editor
• Align CSS stylings (diagram editor, coffee tree editor, workflow analyser)
• Update coffee-tree-editor
  • Update tree node factory and assign actual types and ids to the tree nodes for easier navigation and identification of the tree nodes
  • Update commands to new ModelServerCommand structure
  • Introduce typescript version of the semantic model (coffee-model) and provide type checks for all coffee model elements
  • Update add/remove/edit tree elements either via custom commands (e.g. add/removing nodes and flows), otherwise use json patches for editing

Model Server

• Update coffee model and use ID attributes for EObjects
  • Rename RAM member type to ramType to avoid possible overlaps with json attributes $type and type
• Fix Coffee resource handling
  • Remove obsolete coffee codec
  • Use basic generated coffee resource implementation
  • Remove unused CoffeeModelServerClient
• Cleanup id attribute handling and naming
• Fix Coffee commands
  • Ensure UUIDs are used as id attributes
  • Reuse notation element commands from modelserver-glsp-integration

WorkflowAnalyzerServer

• Fix debugging of WorkflowAnalyzerServer
  • Clean up start up arguments
  • Start wf-analyzer server with arguments to be able to connect from client in debug mode (either start:debug oder launch config external servers)

Remark:

• Currently, there is an issue with the update after pure notation changes (see also API v2: Subscription option for multi-resource update notifications emfcloud-modelserver#236)
• Workaround:
  • The CoffeeMasterTreeWidget overrides the theia-tree-editor one to be able to select nodes via its unique id. If the tree-editor is updated this custom implementation should be removed again.

Resolves #455
Resolves #456
Resolves #348
Resolves #449
Resolves #180
Resolves #260
Resolves #246

[sgraband]

Thank you Nina! Changes work great!

I tested them, but did not do a full code review due to the size of the PR. However while shortly checking the changes i noticed nothing.

I also provided feedback on your inline comments with my commits and fixed the model comparison. Could you test the script once more?

Then i would ask @koegel to do a last testing round.

[ndoschek]

Thanks @sgraband! I also re-tested and updated some small bits in the README.
In general, the demo works for me, however if I run it via the docker container, I cannot get the cpp debugging to run, did it work for you in the container?

[sgraband]

I just retested and the debugging works for me! Thank you for the README changes 👍

[sgraband]

For further improvements i created these follow ups:
#460
#459
#458

(And i guess we need to update the jenkins build as it seems to stil reference the run.sh script)

[ndoschek]

I just retested and the debugging works for me! Thank you for the README changes +1

Thanks a lot @sgraband, then it is fine for me as well, perhaps I overlooked something in the setup.

For further improvements i created these follow ups: #460 #459 #458

Great thank you!

(And i guess we need to update the jenkins build as it seems to stil reference the run.sh script)

No, the Jenkinsfile is already updated, we have again the random vscode-ripgrep issues, although we added the necessary credentials, but it fails again randomly, I will re-trigger a little later, hopefully it will be successful at some point.

[eneufeld]

I will test the code now. But it looks good to me.
I left some comments.

[eneufeld]

All in all I suggest to create some follow up tickets and merge this big PR

[sgraband]

Thank you Nina! Since the follow up tickets were created we should merge this and get started with the follow ups 🙂