autocomplete-python: Improve test coverage

The goal is to add as much tests as possible, since right now the test coverage of autocomplete-python is close to zero.

To start, you must have Atom installed. Clone the repository using git clone and use apm link autocomplete-python to link the git repository as installed package in Atom. Use apm test to run the actual tests (they will fail by default, since the APIs been changed and nobody has fixed them).

Tests are located in spec directory with one example, which tests the actual completions.

Things to test, each as a separated task instance:

  • Fix existing test to pass
  • Go to definition command
  • Complete argument command
  • Override method command
  • Show usages command
  • Rename command (possibly with separated files as well)
  • Extra paths for packages config – create a dummy package and make sure that that dummy library is included in completions when setting is set
  • Autocomplete function parameters config (none, all, required) – test each of them and make sure the output is correct
  • Regex to trigger autocompletions config – make sure that completions are NOT triggered if it does not match this config (e.g. in your test set this config to ([a-zA-Z0-9]*) and make sure that completions are not triggered by . (dots))
  • Use fuzzy matcher for completions – make sure that turning this off works as intended

Task tags

  • coffeescript
  • atom
  • javascript
  • testing
  • nodejs

Students who completed this task

Nalin Bhardwaj

Task type

  • code Code
  • done_all Quality Assurance