First you will need an environment with the autocomplete
script
in which to launch the Argot Server. You may either install the
autocomplete
script yourself locally as described in Muses/autocomplete
or launch Argot Server in its provided Docker image. The muse itself
is a fork of the Galois Autocompleter project with changes such as
the addition of an LSP server frontend.
docker pull mnemo/argot-server
With Mnemosyne running and the Galois autocomplete muse loaded, you can now invoke the muse for code completions guided by machine learning. Open up a Python source file, and add the following code:
if __name__ ==
Place the cursor after the ==
and then invoke a code completion request. (E.g.,
in Emacs with M-x completion-at-point
or in VSCode by pressing Ctrl+Space
).
The LSP server will return a list of applicable completions, at the top of which
should be '__main__'
and "__main__"
.