webscraping_sentiments.py

"""
Reddit DailyProgrammer Challenge #190 - Webscraping Sentiments.
http://bit.ly/11sBRa0
"""

import requests
from bs4 import BeautifulSoup

comments_url = 'http://www.youtube.com/all_comments?v={video_id}'
happy = ['love','loved','like','liked','awesome','amazing','good','great','excellent']
sad = ['hate','hated','dislike','disliked','awful','terrible','bad','painful','worst'] 

def scrape_video_comments(video_id):
    """Given a Youtube video id return
       a list of the videos comments."""
    response = requests.get(comments_url.format(video_id=video_id))
    soup = BeautifulSoup(response.content)
    comments = soup.findAll('div', {'class': 'comment-text-content'})
    comments = [comment.text for comment in comments if comment not in ['', ' ']]
    return comments

def happy_or_sad_comment(comment):
    """Given a comment return 1 if there are more
    happy words, or -1 if more sad words, else return 0"""
    happy_words = sum(1 for word in comment.split(' ') if word in happy)
    sad_words = sum(1 for word in comment.split(' ') if word in sad)
    return dict(happy=happy_words, sad=sad_words)

def happy_or_sad(video_id):
    comments = scrape_video_comments(video_id)
    sad = 0
    happy = 0
    for comment in comments:
        happy_sad_count = happy_or_sad_comment(comment)
        sad += happy_sad_count.get('sad')
        happy += happy_sad_count.get('happy')

    output = """\0
    From a sample size of {no_comments}. This video is mostly {verdict},
    it contained {happy} happy keywords and {sad} sad 
    keywords.
    """.format(no_comments=len(comments),
               verdict='Happy' if happy > sad else 'Sad',
               happy=happy, sad=sad)
    print(output)

if __name__ == '__main__':
   import sys
   happy_or_sad(sys.argv[1])