One of my bad habits as a Python programmer is that I am both lazy and impatient. I will go out of my way to do something overly clever just because it saves me some boring boiler-plate code, and because Python is such a compact language already my threshold for this irritation is very low.

One of the things that this leads me to is writing conditional variable initializations like this:

var = None
if condition:
    var = something()

I've recently come to the conclusion that this is a bad idiom that I should avoid. On the 'good' side, I've saved a line and the annoyance of writing the boiler-plate if/else (at the trivial cost of an extra assignment). On the bad side, what I've found recently is that that the 'var = None' assignment can easily look a little bit too much like it could have been a temporary addition done for debugging or code testing. In fact I almost removed such an initialization today during a cleanup pass of stripping my debugging and testing code from a program before committing the changes to the repository. Actually putting the else: in makes it clear to me that both initializations are supposed to be there.

(Of course there's all the general style and 'clever: -5 points' issues involved with writing code like this. But, well, I'm lazy and impatient and so I sometimes do this stuff anyways.)

I suppose I should be thankful that I don't really like Python's conditional expressions, and also that I'm still writing code for Python 2.4.6, which doesn't have them so I don't have a choice about liking or not liking them. Writing the above as:

var = something() if condition else None

is just far too clever, partly because I find that the order makes it hard to read.