TypeError: list indices must be integers, not unicode - django python

Question

comment_json = urllib2.urlopen("https://graph.facebook.com/comments/?ids=http://www.places4two.de/location/"+locname+"/"+str(lid)+"/")
comment_dict = json.loads(comment_json.read())

if I print comment_dict, i am getting dict in this form this:

{u'http://www.places4two.de/location/date-in-caroussel/2406/': 
{u'comments': 
   {u'paging': 
     {u'cursors': {u'after': u'MQ==', u'before': u'Mg=='}}, 
     u'data': [
        {u'from': {u'name': u'John Purlore', u'id': u'100005454794537'}, 
        u'like_count': 0, 
        u'can_remove': False, 
        u'created_time': u'2013-10-30T09:02:51+0000', 
        u'message': u'Essen ist richtig lecker\n', 
        u'id': u'573597026010502_13875759', 
        u'user_likes': False},
               ]
     }
    }
   }

and now I just want to check if data has some value or not:

if comment_dict['http://www.places4two.de/location/'+locname+'/'+str(lid)+'/']['comments']['data']:

but in this line, i am getting the error:

TypeError: list indices must be integers, not unicode

what am i doing wrong? locname is locationname and str(lid) is string version of location_id - so they are variables.

isnot it because of str()?

Solution 1

Use .get() instead of []

url = 'http://www.places4two.de/location/'+locname+'/'+str(lid)+'/'
if comment_dict.get(url).get('comments').get('data'):
     #my code.

To do this safely (Not run into NoneType has no attribute issues ), you could also do:

url = 'http://www.places4two.de/location/'+locname+'/'+str(lid)+'/'
if comment_dict.get(url, {}).get('comments', {}).get('data', None):
     #my code.

Solution 2

This is because data returns a list of objects, while you're likely accessing it like a dictionary.

Try this:

data = comment_dict['http://www.places4two.de/location/'+locname+'/'+str(lid)+'/']['comments']['data']
if len(data) > 0 and data[0].get("like_count"):
     # code for when like_count is greater than 0

Comments

[Steve Jessop]

I don't believe that you have shown the faulty case: ideone.com/JTsusD. More likely you've hit a case where one of comment_dict or comment_dict[url] or comment_dict[url]['comments'] is an empty list [] rather than a dictionary, if that's permitted by the facebook API you're using.

[oleg]

can You find out which of [] return this ?

[VooDooNOFX]

.get() returns None by default. The final .get('data', None) call could be shortened to .get('data').

[oleg]

In this case KeyError should be raised not TypeError

[doniyor]

so i dont need urllib2 even?

[karthikr]

You do need urllib2 - This is just the dictionary parsing logic.

[karthikr]

@VooDooNOFX Agreed, but I wanted to just show it explicitly so that the OP can choose to do .get('data', [])

[doniyor]

yeah, thanks, that was stupid question. i deleted it.