The X object is a shortcut for writing simple, one-argument lambdas.
It’s roughly equivalent to this: lambda x: x
~X.some_attr.some_func(1, 2, 3).whatever # produces: lambda x: x.some_attr.some_func(1, 2, 3).whatever ~(X > 5) # produces: lambda x: x > 5
Any of the supported operations performed on an X object yield another one that remembers what was done to it, so they can be chained.
The ~ operator creates the actual function that can be called (without creating just another X instance). However, this can be omitted within pipe and pipe-utils context where X is converted to a function automatically:
users > where(X.is_staff) | foreach(X.first_name) | X[:10]
But you can make use of X even outside pipetools, you just have to remember to prefix it with ~
my_list = list(users) my_list.sort(key=~X.last_login.date())
These are added as I need them. It’s quite complete now, but if you’re still missing something you can add it yourself or create an issue on github.
X() X(some, agrs, some=kwargs)
X == something
not equal (__ne__):
X != something
attribute access (__getattr__):
X.attr getattr(X, 'something')
item access, slicing (__getitem__):
X['key'] X X[:10] X[::-1]
greater than (__gt__):
X > 3
greater or equal (__ge__):
X >= 3
less than (__lt__):
X < 3
less or equal (__le__):
X <= 3
X % 2
X * 3
X / 3
X + 3 X + " ...that's what she said!"
X - 3
X ** 2
is contained in (_in_)
Unfortunately, X in container can’t be done (because the magic method is called on the container) so there’s a special method for that: