sentiment_analyzer

from spotipy.oauth2 import SpotifyClientCredentials
import spotipy
import json
import csv
import pylyrics3
from nltk.sentiment.vader import SentimentIntensityAnalyzer as SIA
import nltk
import spotipy.util as util


nltk.download('vader_lexicon')

client_credentials_manager = SpotifyClientCredentials(client_id='befa36b5832642fdbb2aa9062f167a7b', client_secret='6ae5e340ddb048839359f53a0bf9abb2')

sp = spotipy.Spotify(client_credentials_manager=client_credentials_manager)

sentiment_analyzer = SIA()

def get_features(track_id):
    features_results = sp.audio_features(track_id)
    json_features = json.dumps(features_results)
    features_data = json.loads(json_features)
    features_list = list(features_data[0].values())
    return features_list

playlist_ids = [
    "6itzpx6OEo76NWSpy0f3fc",
    "3itylhqRqKPS4nOgWmU7lU",
    "63UBtWxXAtUIR6djuNhGj9",
    "6tZo11hLgEslYdMP5wu4q8",
    "3XXl0JH2uLXpmO5EbH6XPb",
    "5iNZJ2AZghsNgbZgNUngyv",
]




# Audio features
feature_names = [
  "danceability",
  "energy",
  "key",
  "loudness",
  "mode",
  "speechiness",
  "acousticness",
  "instrumentalness",
  "liveness",
  "valence",
  "tempo",
  "type",
  "id",
  "uri",
  "track_href",
  "analysis_url",
  "duration_ms",
  "time_signature"
]

username = 'h93pdaba60md5qbuazllf0mft'



# Write data to CSV file
data_file = open('data.csv','w')

writer = csv.writer(data_file)

# Write header
writer.writerow(['track_id', 'playlist_id', 'date_added', 'track_name', 'first_artist'] + feature_names + ['lyrics', 'neg', 'neu', 'pos', 'compound'])

for playlist_id in playlist_ids:
    print('Querying playlist: ' + str(playlist_id))

    repeat_query = True
    offset_n = 0
    for i in range(2):
        # Query Spotify API
        if i > 0:
            print('Repeating query')
            offset_n += 50
        results = sp.user_playlist_tracks(username, playlist_id, offset=offset_n)
        json_results = json.dumps(results)
        data = json.loads(json_results)

# Write rows

        for track in data['items']:
                        track_id = track['track']['id']
                        date_added = track['added_at']
                        track_name = track['track']['name']
                        first_artist = track['track']['artists'][0]['name']

# Track features

                        features = get_features(track_id)

# Try to get lyrics, if available

                        lyrics = ''

                        try:
                                        lyrics = pylyrics3.get_song_lyrics(first_artist, track_name)

                        except:
                                        pass

# Sentiment Analysis

                        neg = None
                        neu = None
                        pos = None

                        compound = None

                        if lyrics:
                                        snt = sentiment_analyzer.polarity_scores(lyrics)
                                        neg = snt['neg']
                                        neu = snt['neu']
                                        pos = snt['pos']
                                        compound = snt['compound']
                        try:
                            writer.writerow([track_id, playlist_id, date_added, track_name, first_artist] + features + [lyrics] + [neg, neu, pos, compound])
                        except:
                            print("error in csv writting")
                    #if data['total'] < 50:
                    #           break           ????

    print('Done querying')

data_file.close()