Breaking a list into strings, not characters

Question

I have a defaultdict that looks like this:

"Some text" : ["Some date", "Some date", "Some Date"]

I am trying to access each individual value of each key like so:

for processedTweet, date in tweetsDict.iteritems():
    print date
    for d in date:
        print d

This works fine in a normal Python script. It prints the entire list first then in the for loop it prints each individual date.

But when I send this as part of a Map/Reduce job to Hadoop, it breaks the list into individual characters, not Strings, ie:

Some date

becomes

S
o
m

etc. Why is this happening and how can I fix it?

Answer 1

The map/reduce job in hadoop is distributing your list values to the for-loop instead of passing the entire list. By default, when python iterates over a string object each iteration returns the next character in the string.