How do I debug PyDev's autocompletion server to figure out why it can't analyze a module?

Question

I'm using PyDev to edit Python and have a module that can't be analyzed statically or as a forced builtin. How do I debug the autocompletion server Python process that PyDev runs to figure out why it can't provide completions for this module, and hopefully tweak things to make it work?

Answer 1

Well, the completions in the forced builtins are mostly driven by doing a dir() in the module and then getting the tokens (so if you can do that from a shell, the code-completion engine should be able to do it too -- but note it'll only import and do a dir, not actually execute your code).

Also, this will only work for code which is not below your project (ie: code which is from an external library in the system interpreter)

For analyzing your own code (below a project), PyDev will only use the static analyzer... in which case it depends on the code, you can usually enhance it with type info on docstrings: http://pydev.org/manual_adv_type_hints.html or by doing some 'if False:' and writing some code which would only be gotten for the static analysis...