Problem in understanding Python list comprehensions

Question

What does the last line mean in the following code?

import pickle, urllib                                                                                                                                                     

  handle = urllib.urlopen("http://www.pythonchallenge.com/pc/def/banner.p")
  data = pickle.load(handle)
  handle.close()

  for elt in data:
         print "".join([e[1] * e[0] for e in elt])

My attempt to the problem:

• "".join... uses join -method to empty text
• e[1] * e[0] multiplies two subsequent values in the sequence, e
• I am not sure what is e
• I am not sure, what it means, when you have something before for -loop, like: e[1] * e[0] for e in elt

Answer 1

Maybe best explained with an example:

print "".join([e[1] * e[0] for e in elt])

is the short form of

x = []
for e in elt:
  x.append(e[1] * e[0])
print "".join(x)

List comprehensions are simply syntactic sugar for for loops, which make an expression out of a sequence of statements.

elt can be an arbitrary object, since you load it from pickles, and e likewise. The usage suggests that is it a sequence type, but it could just be anything that implements the sequence protocol.

Answer 2

Firstly, you need to put http:// in front of the URL, ie:

handle = urllib.urlopen("http://www.pythonchallenge.com/pc/def/banner.p")

An expression [e for e in a_list] is a list comprehension which generates a list of values.

With Python strings, the * operator is used to repeat a string. Try typing in the commands one by one into an interpreter then look at data:

>>> data[0]
[(' ', 95)]

This shows us each line of data is a tuple containing two elements.

Thus the expression e[1] * e[0] is effectively the string in e[0] repeated e[1] times.

Hence the program draws a banner.

Answer 3

[e[1] * e[0] for e in elt] is a list comprehension, which evaluates to a list itself by looping through another list, in this case elt. Each element in the new list is e[1]*e[0], where e is the corresponding element in elt.

Answer 4

The question itself has already been fully answered but I'd like to add that a list comprehension also supports filtering. Your original line

print "".join([e[1] * e[0] for e in elt])

could, as an example, become

print "".join([e[1] * e[0] for e in elt if len(e) == 2])

to only operate on items in elt that have two elements.

Answer 5

join() is a string method, that works on a separator in new string

>>> ':'.join(['ab', 'cd'])
>>> 'ab:cd'

and list comprehension is not necessary there, generator would suffice

Answer 6

Andy's is a great answer!

If you want to see every step of the loop (with line-breaks) try this out:

for elt in data:
    for e in elt:
        print "e[0] == %s, e[1] == %d, which gives us:  '%s'" % (e[0], e[1], ''.join(e[1] * e[0]))