Why does Python 'for word in words:' iterate on individual characters instead of words?

Question

When I run the following code on a string words:

def word_feats(words):
    return dict([(word, True) for word in words])
print(word_feats("I love this sandwich."))

I get the output dict-comprehension in letters instead of words:

{'a': True, ' ': True, 'c': True, 'e': True, 'd': True, 'I': True, 'h': True, 'l': True, 'o': True, 'n': True, 'i': True, 's': True, 't': True, 'w': True, 'v': True, '.': True}

What am I doing wrong?

Answer 1

You have to split the words string:

def word_feats(words):
    return dict([(word, True) for word in words.split()])
print(word_feats("I love this sandwich."))

Example

>>> words = 'I love this sandwich.'
>>> words = words.split()
>>> words
['I', 'love', 'this', 'sandwich.']

You can also use other characters on which to split:

>>> s = '23/04/2014'
>>> s = s.split('/')
>>> s
['23', '04', '2014']

Your Code

def word_feats(words):
    return dict([(word, True) for word in words.split()])
print(word_feats("I love this sandwich."))

[OUTPUT]
{'I': True, 'love': True, 'this': True, 'sandwich.': True}

Answer 2

You need to explicitly split the string on whitespace:

def word_feats(words):
    return dict([(word, True) for word in words.split()])

This uses str.split() without arguments, splitting on arbitrary-width whitespace (including tabs and line separators). A string is a sequence of individual characters otherwise, and direct iteration will indeed just loop over each character.

Splitting into words, however, has to be an explicit operation you need to perform yourself, because different use-cases will have different needs on how to split a string into separate parts. Does punctuation count, for example? What about parenthesis or quoting, should words grouped by those not be split, perhaps? Etc.

If all you are doing is setting all values to True, it'll be much more efficient to use dict.fromkeys() instead:

def word_feats(words):
    return dict.fromkeys(words.split(), True)

Demo:

>>> def word_feats(words):
...     return dict.fromkeys(words.split(), True)
... 
>>> print(word_feats("I love this sandwich."))
{'I': True, 'this': True, 'love': True, 'sandwich.': True}