A few helpful utilities that I find useful.
Dictionary type that allows for dynamic allocation of nested key values without prior defintion. ::
>>> a = AutoVivification()
>>> a[1][2] = 3
This module provides a set of methods for handling multi-threading patterns more easily. ::
>>> import user_metrics.utils.multiprocessing_wrapper as mpw
>>> mpw.build_thread_pool(['one','two'],len,2,[])
[2,2]
Defines the recordtype method which returns a type definition object that allows for the creation of mutable objects with specified attributes. These are essentially identical to collections.namedtuple except that the attributes may be changed after the namedtuple objects are created. ::
>>> Point = recordtype('Point', 'x y', default=0)
>>> Point() # instantiate with defaults
Point(x=0, y=0)
Some miscellaneous methods exposed by this package.
The format_timestamp
method takes a timestamp representation (string
or datetime) and convert to a standard string representation.
The enum
method implements an enumeration::
>>> Numbers = enum('ZERO', 'ONE', 'TWO')
>>> Numbers.ZERO
0
>>> Numbers.ONE
1
Given a set of types, values, and names builds a named tuple. This
method expects three lists all of the same length and returns a
dynamically built namedtuple object. Currently, only list
,
str
, int
, and float
cast methods are accepted as members
of types
. The mothod is named build_namedtuple
.
Unpacks the attributes and values of a record type or named tuple into an ordered dictionary.