class IMDBApi(object):
def __init__(self, title, year=0):
self.query = dict(title=title, year=int(year))
self.rt = RT(rotten)
def __call__(self):
matches = []
year = self.query['year']
results = self.rt.search(self.query['title'])
self.partial = copy.copy(results)
if year:
results = [x for x in results if x['year']==self.query['year']]
for r in results:
matches.append(self.rt.info(r['id']))
return matches
from rottentomatoes import RT
import time
rt = RT()
#movies = ['fight club','gravity','toy story 3','american hustle','skyfall','jack and jill','basic instinct 2','white out','lost souls','babylon','argo','bears','her','up','a beautiful mind','braveheart','the hurt locker','gambit','paranoia','getaway']
movies = [line.strip('\n') for line in open('movies.txt')]
f = open('reviews-date.txt', 'w')
count = 0
for u in movies:
movlst = rt.search(u, page_limit=1)
if movlst:
if movlst[0][u'id'] != '':
review = rt.info(movlst[0][u'id'], 'reviews')
review.viewkeys()
rlt = review[
u'reviews'] #review is a dict, value of each key is list, each list element is a dict
for a in rlt:
f.write(a[u'quote'] + '\t' + a[u'date'] + '\n')
f2.write(a[u'date'] + '\n')
count = count + 1
#print a[u'quote']
if len(rlt) == 0:
print u
else:
print count
else:
print u
else:
print u
time.sleep(5)
f.close()
from rottentomatoes import RT
import time
rt = RT()
# movies = ['fight club','gravity','toy story 3','american hustle','skyfall','jack and jill','basic instinct 2','white out','lost souls','babylon','argo','bears','her','up','a beautiful mind','braveheart','the hurt locker','gambit','paranoia','getaway']
movies = [line.strip("\n") for line in open("movies.txt")]
f = open("reviews-date.txt", "w")
count = 0
for u in movies:
movlst = rt.search(u, page_limit=1)
if movlst:
if movlst[0][u"id"] != "":
review = rt.info(movlst[0][u"id"], "reviews")
review.viewkeys()
rlt = review[u"reviews"] # review is a dict, value of each key is list, each list element is a dict
for a in rlt:
f.write(a[u"quote"] + "\t" + a[u"date"] + "\n")
f2.write(a[u"date"] + "\n")
count = count + 1
# print a[u'quote']
if len(rlt) == 0:
print u
else:
print count
else:
print u
else:
print u
time.sleep(5)
f.close()
class MovieInfo(object):
def __init__(self, movie, rotten_tomatoe_api_key, tmdb_api_key,
aws_access_key, aws_secret_key, affiliate_key, rt_id=None, tmdb_id=None):
self._movie = movie
# amazon
self._amazon_product_search = AmazonProductSearch(aws_access_key,
aws_secret_key, affiliate_key)
# rotten tomatoes
self._rt = RT(rotten_tomatoe_api_key)
if rt_id:
self._rt_data = self._rt.info(rt_id)
else:
self._rt_data = self._rt.search(movie)[0]
# tmdb
self._tmdb = tmdb
self._tmdb.configure(tmdb_api_key)
movie = self._tmdb.Movies(movie, limit=True,
expected_release_date=self._rt_data['release_dates']['theater']).get_best_match()
self._tmdb_data = self._tmdb.Movie(movie[1]['id'])
# youtube
self._yt_service = gdata.youtube.service.YouTubeService()
def get_amazon_purchase_links(self, top_cast, runtime):
products = self._amazon_product_search.item_search(self._movie,
top_cast, runtime)
return products
@property
def cast(self):
'''
Returns the names of the full cast for this movie
'''
full_cast = self._rt.info(self._rt_data['id'], 'cast')
names = [cast['name'] for cast in full_cast['cast']]
return names
@property
def critic_reviews(self):
'''
Returns a list of critic reviews for this movie. The list
is componsed of json document.
'''
reviews = self._rt.info(self._rt_data['id'], 'reviews')
return reviews['reviews']
@property
def critics_score(self):
'''
Returns the rotten tomatoe critic score for this movie
'''
return self._rt_data['ratings']['critics_score']
@property
def director(self):
'''
Returns a list of directors for this movie
'''
return self._tmdb_data.get_director()
@property
def genres(self):
'''
Returns the genres of this movie, supplied by tmdb
'''
genres = self._tmdb_data.get_genres()
genres = [genre['name'].lower() for genre in genres]
return genres
@property
def imdb_id(self):
'''
Returns a list of directors for this movie
'''
try:
return "tt" + self._rt_data['alternate_ids']['imdb']
except:
return self._tmdb_data.get_imdb_id()
@property
def poster(self):
'''
Returns the poster of the movie, in its original size
'''
return self._tmdb_data.get_poster()
@property
def runtime(self):
'''
Return the runtime of this movie in minues
'''
try:
return int(self._rt_data['runtime'])
except:
return int(self._tmdb_data.get_runtime())
@property
def release_date(self):
'''
Returns this movie's release date in {year}-{month}-{day} format
'''
try:
return parser.parse(self._rt_data['release_dates']['theater'])
except:
return parser.parse(self._tmdb_data.get_release_date())
@property
def similar_movies(self):
'''
Returns a list of imdb ids of movies that are similar to this one
'''
movies = self._rt.info(self._rt_data['id'], 'similar')['movies']
# if movie is none or len(movies) == 0:
# search google current movie title +
# parse DOM for
return movies
@property
def synopsis(self):
'''
Returns this movie's synopsis
'''
synopsis = self._rt_data['synopsis']
if len(synopsis) == 0:
synopsis = self._tmdb_data.get_overview()
return synopsis
@property
def title(self):
'''
Returns this movie's title
'''
return self._rt_data['title']
@property
def trailers(self, limit=3):
'''
This function returns a list of trailers for this movie.
We will use TMDB's data if it returns 3 or more trailers. If not,
we will query youtube with the search term: "{movie_name} trailer
{release_year}" to find trailers for this movie.
Returns a list of youtube ids of the trailers
'''
trailers = self._tmdb_data.get_trailers()['youtube']
if len(trailers) > limit:
return [trailer['source'] for trailer in trailers]
else:
release_year = str(self.release_date).split('-')[0]
query = gdata.youtube.service.YouTubeVideoQuery()
query.vq = "{title} trailer {release_year} ".format(
title=self._movie, release_year=release_year)
query.orderby = 'relevance'
feed = self._yt_service.YouTubeQuery(query)
entries = self._remove_long_youtube_videos(feed.entry[:3])
entries = self._remove_unrelated_videos(entries)
unique_entries = self._remove_duplicate_youtube_videos(entries)
return unique_entries
def _remove_duplicate_youtube_videos(self, entries, threshold=5):
'''
This method removes duplicate videos by measuring the runtime
of the youtube videos.
If two videos are within 5 seconds (the threshold) of each
other in runtime, we assume that one of the videos is a duplicate
of the other.
'''
limit = 3
videos = []
for entry in entries:
runtime = int(entry.media.duration.seconds)
similar = [runtime >= int(video["runtime"]) - threshold and
runtime <= int(video["runtime"]) + threshold for
video in videos]
if not any(similar):
video_id = self._extract_youtube_id(entry.media.player.url)
videos.append({"yt_id": video_id, "runtime": runtime})
yt_ids = [video['yt_id'] for video in videos]
return yt_ids
def _remove_long_youtube_videos(self, entries, max_seconds=600):
entries = filter(lambda entry:
int(entry.media.duration.seconds) < max_seconds,
entries)
return entries
def _remove_unrelated_videos(self, entries):
entries = filter(lambda entry:
fuzzywuzzy.fuzz.ratio(entry.media.title.text.decode('utf-8').lower(), self._movie.lower()) > 20,
entries)
return entries
def _extract_youtube_id(self, youtube_url):
video_id = youtube_url.split('v=')[1]
ampersand_position = video_id.find('&')
if(ampersand_position != -1):
video_id = video_id[0:ampersand_position]
return video_id
class MovieInfo(object):
def __init__(self,
movie,
rotten_tomatoe_api_key,
tmdb_api_key,
aws_access_key,
aws_secret_key,
affiliate_key,
rt_id=None,
tmdb_id=None):
self._movie = movie
# amazon
self._amazon_product_search = AmazonProductSearch(
aws_access_key, aws_secret_key, affiliate_key)
# rotten tomatoes
self._rt = RT(rotten_tomatoe_api_key)
if rt_id:
self._rt_data = self._rt.info(rt_id)
else:
self._rt_data = self._rt.search(movie)[0]
# tmdb
self._tmdb = tmdb
self._tmdb.configure(tmdb_api_key)
movie = self._tmdb.Movies(
movie,
limit=True,
expected_release_date=self._rt_data['release_dates']
['theater']).get_best_match()
self._tmdb_data = self._tmdb.Movie(movie[1]['id'])
# youtube
self._yt_service = gdata.youtube.service.YouTubeService()
def get_amazon_purchase_links(self, top_cast, runtime):
products = self._amazon_product_search.item_search(
self._movie, top_cast, runtime)
return products
@property
def cast(self):
'''
Returns the names of the full cast for this movie
'''
full_cast = self._rt.info(self._rt_data['id'], 'cast')
names = [cast['name'] for cast in full_cast['cast']]
return names
@property
def critic_reviews(self):
'''
Returns a list of critic reviews for this movie. The list
is componsed of json document.
'''
reviews = self._rt.info(self._rt_data['id'], 'reviews')
return reviews['reviews']
@property
def critics_score(self):
'''
Returns the rotten tomatoe critic score for this movie
'''
return self._rt_data['ratings']['critics_score']
@property
def director(self):
'''
Returns a list of directors for this movie
'''
return self._tmdb_data.get_director()
@property
def genres(self):
'''
Returns the genres of this movie, supplied by tmdb
'''
genres = self._tmdb_data.get_genres()
genres = [genre['name'].lower() for genre in genres]
return genres
@property
def imdb_id(self):
'''
Returns a list of directors for this movie
'''
try:
return "tt" + self._rt_data['alternate_ids']['imdb']
except:
return self._tmdb_data.get_imdb_id()
@property
def poster(self):
'''
Returns the poster of the movie, in its original size
'''
return self._tmdb_data.get_poster()
@property
def runtime(self):
'''
Return the runtime of this movie in minues
'''
try:
return int(self._rt_data['runtime'])
except:
return int(self._tmdb_data.get_runtime())
@property
def release_date(self):
'''
Returns this movie's release date in {year}-{month}-{day} format
'''
try:
return parser.parse(self._rt_data['release_dates']['theater'])
except:
return parser.parse(self._tmdb_data.get_release_date())
@property
def similar_movies(self):
'''
Returns a list of imdb ids of movies that are similar to this one
'''
movies = self._rt.info(self._rt_data['id'], 'similar')['movies']
# if movie is none or len(movies) == 0:
# search google current movie title +
# parse DOM for
return movies
@property
def synopsis(self):
'''
Returns this movie's synopsis
'''
synopsis = self._rt_data['synopsis']
if len(synopsis) == 0:
synopsis = self._tmdb_data.get_overview()
return synopsis
@property
def title(self):
'''
Returns this movie's title
'''
return self._rt_data['title']
@property
def trailers(self, limit=3):
'''
This function returns a list of trailers for this movie.
We will use TMDB's data if it returns 3 or more trailers. If not,
we will query youtube with the search term: "{movie_name} trailer
{release_year}" to find trailers for this movie.
Returns a list of youtube ids of the trailers
'''
trailers = self._tmdb_data.get_trailers()['youtube']
if len(trailers) > limit:
return [trailer['source'] for trailer in trailers]
else:
release_year = str(self.release_date).split('-')[0]
query = gdata.youtube.service.YouTubeVideoQuery()
query.vq = "{title} trailer {release_year} ".format(
title=self._movie, release_year=release_year)
query.orderby = 'relevance'
feed = self._yt_service.YouTubeQuery(query)
entries = self._remove_long_youtube_videos(feed.entry[:3])
entries = self._remove_unrelated_videos(entries)
unique_entries = self._remove_duplicate_youtube_videos(entries)
return unique_entries
def _remove_duplicate_youtube_videos(self, entries, threshold=5):
'''
This method removes duplicate videos by measuring the runtime
of the youtube videos.
If two videos are within 5 seconds (the threshold) of each
other in runtime, we assume that one of the videos is a duplicate
of the other.
'''
limit = 3
videos = []
for entry in entries:
runtime = int(entry.media.duration.seconds)
similar = [
runtime >= int(video["runtime"]) - threshold
and runtime <= int(video["runtime"]) + threshold
for video in videos
]
if not any(similar):
video_id = self._extract_youtube_id(entry.media.player.url)
videos.append({"yt_id": video_id, "runtime": runtime})
yt_ids = [video['yt_id'] for video in videos]
return yt_ids
def _remove_long_youtube_videos(self, entries, max_seconds=600):
entries = filter(
lambda entry: int(entry.media.duration.seconds) < max_seconds,
entries)
return entries
def _remove_unrelated_videos(self, entries):
entries = filter(
lambda entry: fuzzywuzzy.fuzz.ratio(
entry.media.title.text.decode('utf-8').lower(),
self._movie.lower()) > 20, entries)
return entries
def _extract_youtube_id(self, youtube_url):
video_id = youtube_url.split('v=')[1]
ampersand_position = video_id.find('&')
if (ampersand_position != -1):
video_id = video_id[0:ampersand_position]
return video_id