[FIXED] ipython tab completion of method parameters showing builtins

Issue

IPython tab completion for method parameters is showing more information than I'm expecting to see:

snowch$ ipython3
Python 3.6.2 |Anaconda, Inc.| (default, Sep 21 2017, 18:29:43)
Type 'copyright', 'credits' or 'license' for more information
IPython 6.1.0 -- An enhanced Interactive Python. Type '?' for help.

In [1]: from my_package import MyClass

In [2]: cf = MyClass(<TAB>
                       abs()                                 api_key=
                       all()                                 api_key_filename=
                       any()                                 ArithmeticError                       >
                       api_endpoint=                         ascii()
                       ...

This also happens on instance methods:

In [3]: cf.auth(<TAB>
                 abs()                                 ascii()
                       ...

Q) Should I be seeing builtin methods like abs() when I tab complete on my instance methods and constructors?.

My IPCompleter configuration looks like this:

In [4]: %config IPCompleter
IPCompleter options
-----------------
IPCompleter.backslash_combining_completions=<Bool>
    Current: True
IPCompleter.debug=<Bool>
    Current: False
IPCompleter.greedy=<Bool>
    Current: False
IPCompleter.jedi_compute_type_timeout=<Int>
    Current: 400
IPCompleter.limit_to__all__=<Bool>
    Current: False
IPCompleter.merge_completions=<Bool>
    Current: True
IPCompleter.omit__names=<Enum>
    Current: 2
IPCompleter.use_jedi=<Bool>
    Current: True

Solution

Yes you should, because having function call in methods or constructor is perfectly valid. Actually you can write any expressions in any field when you call a callable.

MyProgressMeter(any(params))
range(abs(value_a - value_b))
obj.frobulate(len(set(items)))
print(len(my_list))

Are reasonable, so on MyProgressMeter(<tab>, obj.frobulate(<tab>, print(<tab> and range(<tab> you're shown everything that is in local/global namespace that make sense, as well as named parameters. Instance methods are no different of course.

Note: it is not well known but range is actually a constructor:

>>> type(range)
<class 'type'>

Answered By - Matt