Pangrams CodeEval challenge

Question

I took a challenge on CodeEval. Although the code seems to work for the examples taken from the site, I feel it is not really pretty and must be more complicated than it should be.

Description:

The sentence 'A quick brown fox jumps over the lazy dog' contains every single letter in the alphabet. Such sentences are called pangrams. You are to write a program, which takes a sentence, and returns all the letters it is missing (which prevent it from being a pangram). You should ignore the case of the letters in sentence, and your return should be all lower case letters, in alphabetical order. You should also ignore all non US-ASCII characters.In case the input sentence is already a pangram, print out the string NULL.

import sys
filepath = sys.argv[1]
f = open(filepath)
wholealphabet = ('a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r',
 's','t','u','v','w','x','y','z')
for line in f:
 sortedletters = list(set(line.lower()))
 i = 0
 while i != len(sortedletters):
 if wholealphabet.count(sortedletters[i]) != 0: 
 i = i + 1
 else:
 sortedletters.pop(i)
 missingletters = ""
 for letter in wholealphabet:
 if sortedletters.count(letter) == 0:
 missingletters +=letter
 if len(missingletters) == 0:
 print("NULL")
 else:
 print(missingletters)

Answer 1

One of Python's greatest strengths is its built-in capability to use sets directly. I don't feel you've used sets to their fullest extent here. I'd also like to point out the with statement, which you should probably use to handle file handles.

from __future__ import with_statement
import sys
from string import ascii_lowercase
filepath = sys.argv[1]
wholealphabet = frozenset(ascii_lowercase)
# Use with to handle file ... handles
with open(filepath) as f:
 for line in f: # assume a line is a sentence, not exactly per spec?
 # sortedletters = list(set(line.lower())) # guaranteed to be *unsorted*
 missingletters = wholealphabet.difference(line.lower())
 if missingletters:
 print ''.join(sorted(missingletters))
 else:
 print 'NULL'

That's really all you need. Unless you want to reconsider the definition of a sentence. :)

• Tommy: wow it really is much simpler that way, didn't know about the with. It seems to work like the using statement in c#. I will do a research on each of the method used as I do not know them. I indeed forgot about actually doing the sort, I made some attempt with the interpreter at first and then pretty much copy pasted in a .py file. What did you mean about reconsidering the definition of a sentence? Wish I could upvote but I do not have the required rank yet. Thank you.
• Tommy: Nevermind about the sentence definition I rereaded the answer and understood :)
• kojiro: FYI I moved the from future... to the top of the code, because (as I'm sure you've discovered) the code won't work otherwise.
• Tommy: Indeed I had to change it, I also discover that string has a ascii_lowercase method which I used instead of letters.lower()
• kojiro: Great catch. So edited!

Answer 2

import sys
filepath = sys.argv[1]
f = open(filepath)

I recommend not using one letter variable names (usually). They make the code hard to read.

wholealphabet = ('a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r',
 's','t','u','v','w','x','y','z')

I'd have made this a string

wholealphabet = "abcdefghijklmnopqrstuvwxyz"

Shorter and works pretty much the same

for line in f:
 sortedletters = list(set(line.lower()))

Yeah, that's not sorted.

 i = 0
 while i != len(sortedletters):
 if wholealphabet.count(sortedletters[i]) != 0: 

This is the same as sortedletters[i] in wholealphabet, which is clearer.

 i = i + 1
 else:
 sortedletters.pop(i)

Modifying a list while iterating over it is bound to be confusing. Its hard to see what you are doing here. The best way in python is usually to create a new list. Like this:

valid_letters = []
for letter in sortedletters:
 if letter in wholealphabet:
 valid_letters.append(letter)

See how much easier it is to see the result of that? In fact, you can even do it more compactly:

valid_letters = [letter for letter in sortedletters if letter in wholealphabet]
 missingletters = ""

Adding to a string can be expensive, I recommend using a list

 for letter in wholealphabet:
 if sortedletters.count(letter) == 0:
 missingletters +=letter

Again, you can simplify this using a list comphrehension

missingletters = [letter for letter in wholealphabet if letter not in sortedletters]
 if len(missingletters) == 0:
 print("NULL")
 else:
 print(missingletters)

As kojrio points out, if you use sets in python you can implement this very easily. His advice to use a with statement is also good.

• Tommy: I agree that the string is really easier to see that the list of char. I was wondering if there wasn't already a way to test if a char is a valid letter if I had been in c++ I would have casted to a int and would have compared with the ascii table. Indeed forgot the sort that was pretty bad, didn't have enough test case to spot it. I find list comprehension to be hard to read but maybe it's just because I am not used to see it like that. I'll read about it. Thank you for your help.
• Winston Ewert: @Tommy, list comprehensions can be hard to read until you get used to them.
• Tommy: yes, I guess it something you get used to just like lambda expression and then start loving.