Comparing unrelated Python types¶

I've been researching dunder methods in preparation for my upcoming talk at PyTexas. Somehow this lead me to playing around with code like the following:

>>> x = range(50)
>>> x < 20

So, what's the value of x < 20?

I'll kill the suspense for you, it is False.

I know, it didn't make sense to me either. I don't know what comparing integers to lists means semantically, but I would like to know why the answer is False. Python must have some reason for this being False, and I figured it would be interesting to look into.

I thought the answer could potentially be relevant for my upcoming talk. After all, the implementation of the < operator is handled by the less-than dunder method for the list type.

Assumptions¶

My first thought was the < operator was comparing with len() for the list, which would make x < 51 True. Result: wrong.
OK, then I figured it was using the list's id. This would sort of make sense. Result: wrong again.

>>> id(x)
170461900
>>> x < 170461901
False

To the source!¶

At this point the sheer mystery drove me to download the Python 2.7 source. I could have scoured the Internet for a quick solution, but I haven't spent much time digging in the Python 2.7 source so now seemed like a good time to start.

I started my search by looking up the implementation for '<' in the list object. Now the fun of tracing C function pointers and macros begins...

I'll save you the time and just provide a summary of what I found. The comparison operation ends in the default_3way_compare function. Finally, the following snippet of code decides that our list is always greater than (not less than) all integers:

/* different type: compare type names; numbers are smaller */
if (PyNumber_Check(v))
    vname = "";
else
    vname = v->ob_type->tp_name;
if (PyNumber_Check(w))
    wname = "";
else
    wname = w->ob_type->tp_name;
c = strcmp(vname, wname);
if (c < 0)
    return -1;
if (c > 0)
    return 1;

Surprised? Confused? I was both.

TL;DR¶

The end result is that comparison is based on the actual name of the types! So, in our case List is greater alphabetically than Integer!

Not exactly the rationale I was hoping for, but at least we have an answer. However, now that we know what is happening see if you can figure out the answer to a few more puzzles:

>>> [] < ()
>>> [] < {}

Spoiler alert

A List is less than a Tuple, remember 'L' comes before 'T' in the alphabet, and of course by the same logic a List is NOT less than a Dict.

Easter Egg¶

None is really small in comparison:

/* None is smaller than anything */
if (v == Py_None)
    return -1;
if (w == Py_None)
    return 1;

Python 3¶

This is a bit weird, but wait there's more! What does Python 3 do in this case?

>>> x = range(50)
>>> x < 20
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
TypeError: unorderable types: range() < int()
>>> [] < ()
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
TypeError: unorderable types: list() < tuple()
>>> [] < {}
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
TypeError: unorderable types: list() < dict()

Nice! This little quirk has been fixed. Now when you upgrade your awesome application to Python 3 comparing completely different types will complain loudly as it should. Then, you will be free to forget everything you learned in this post. :)