Hi there,
The lambda-py project has been going thus far and we are starting new things to push it further.
One of the deficiency of current lambda-py is that it generates too much thing internally to interpret: one simple python assignment statement will generate thousands of lines of core language.
What we are going to do next is to find out which parts of our interpreter are actually blowing up the whole things. We are going to set some flags on each of desugaring functions. By switching on and off the flag, we are able to spot the change of generated code, so much so that we may find out which parts of desugaring blow things up, and hopefully this can generally improve the whole efficiency of interpretation.
Some flags starting with "dsg-" have been added to python-desugar.rkt. Any suggestions and comments are welcome!
