implement prototype for using new javac-based JDT parser

[martinlippert]

The majority of the information around the incubated version of JDT that allows you to switch between javac and ejc-based parsing of Java source code can be found here: https://github.com/redhat-developer/vscode-java/releases/tag/v1.36.0

The idea here is to build a prototype of the spring boot language server that uses this incubating version of JDT with javac-based parsing enabled. Questions would be:

• what do we need to change in order to adopt this?
  • parsing source code (bulk)
  • parsing source code (single file)
  • caching the environment (to speed-up lookup when re-parsing a file)
• what are the performance implications of this? (running sample projects of up to 60k source files)
• what are the memory implications of this? (running sample projects of up to 60k source files)

[martinlippert]

Made some very early steps here, but didn't got very far: eclipse-jdtls/eclipse-jdt-core-incubator#1077

Since this is probably the first time someone tries to use this outside of a running Eclipse environment, I am not surprised that this doesn't work (yet) out of the box, but probably good feedback for the people working on this.

The core implementation is in https://github.com/eclipse-jdtls/eclipse-jdt-core-incubator, the dom-with-javac branch. I build that locally and then added org.eclipse.jdt.core as well as org.eclipse.jdt.core.javac to the project dependencies, I haven't found a maven repository with this anywhere yet.

[martinlippert]

Here is the branch with my stupid early steps: https://github.com/spring-projects/sts4/tree/dom-with-javac