The workflow to parse and extract definiens must be copied to LaCASt.
I do not think it is necessary for LaCASt to support the flink workflow now. Since the only bonus is it can handle large datasets efficiently but that is currently out of scope. The main task (later) is more to support translations as a service.
If we want to run LaCASt on entire datasets someday, we can easily re-integrate the flink workflow again. Or add LaCASt to mathosphere (might be easier). The cross dependency can be solved since mathosphere (once LaCASt is available) only require common module rather than entire LaCASt.
AndreG-P
commented
4 years ago
The workflow to parse and extract definiens must be copied to LaCASt.
I do not think it is necessary for LaCASt to support the flink workflow now. Since the only bonus is it can handle large datasets efficiently but that is currently out of scope. The main task (later) is more to support translations as a service.
If we want to run LaCASt on entire datasets someday, we can easily re-integrate the flink workflow again. Or add LaCASt to mathosphere (might be easier). The cross dependency can be solved since mathosphere (once LaCASt is available) only require
commonmodule rather than entire LaCASt.