We know that calls to code model might be expensive; therefore clients (such as,
for example, Class View and Code Completion) call model in a sepatate thread.
For now they use RequestProcessor.getDefault() for this.
Besides, model implementation uses its own threads (for example, for warmng up
model upon project opening).
It would be fine if model got full control over all these threads. This would
help to solve thread issues.
- modelapi: Added CsmModel.enqueue method
- modelimpl: added CodeModelRequestProcessor class that is responsible for
threading issues. It in turn uses RequestProcessor.
classview & completion: used CsmModel.enqueue() instead of
*** Issue 75930 has been marked as a duplicate of this issue. ***