osa1 github gitlab twitter cv rss

Half-baked idea: Abstract interpretation for code completion

October 16, 2013 - Tagged as: en.

I had this idea about implementing code completion on dynamic languages. Statically inferring information in dynamic languages is hard and never accurate enough. In my opinion, we should take a step towards dynamically checking the code by having some kind of abstract interpretation of the language that allows us infer required information and then running programs in this interpretation.

I didn’t figure out the details yet. I just wanted to write here as a TODO.