Viraj's Weblog

Learned some new things about Python dictionaries. Let's say you wanted to create a dictionary with keys given in a list, and a default value for all of them, say None.

>>> somelist=['foo', 'bar', 'baz']
>>> {}.fromkeys(somelist)
{'baz': None, 'foo': None, 'bar': None}

Or some other value, like True:

>>> {}.fromkeys(somelist, True)
{'baz': True, 'foo': True, 'bar': True}

An uglier way to do this is with a list comprehension:

>>> dict([(x, True) for x in somelist])
{'baz': True, 'foo': True, 'bar': True}

Now you can easily see a way to uniq-ize a list:

>>> somelist=['foo', 'bar', 'foo']
>>> {}.fromkeys(somelist).keys()
['foo', 'bar']

But really, there is an even better way to uniq-ize a list in 2.4:

>>> list(set(somelist))
['foo', 'bar']

On another note, let's say I wanted to loop through the keys of a dict in sorted order in Python 2.4:

>>> somelist=['foo', 'bar', 'baz']
>>> somedict={}.fromkeys(somelist)
>>> for key in sorted(somedict):
...   print key
... 
bar
baz
foo

Or, did you know you can create a dict with x=y arguments?

>>> dict(x=1,y=2)
{'y': 2, 'x': 1}