-
Notifications
You must be signed in to change notification settings - Fork 0
/
analyzer.py
641 lines (542 loc) · 25.7 KB
/
analyzer.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
from bs4 import BeautifulSoup
import requests
import psycopg2
from psycopg2 import OperationalError
import os
import threading
import random
# import pandas as pd
connection = None
FUNDAMENTAL_GENRES = ['classical', 'country', 'edm', 'folk', 'hiphop', 'jazz', 'latin', 'rap', 'rock']
'''
Playlist analyzer class using multithreading to serve
multiple clients simultaneously.
'''
class PlaylistAnalyzer(threading.Thread):
def __init__(self, id):
self.progress = 0
self.id = id
self.result = ""
self.playlist = None
self.artists = {
"male":{},
"female":{},
"nonbinary":{},
"mixed_gender":{},
"undetermined":{}
}
self.recs = {
"60-100":None,
"30-60":None,
"0-30":None
}
self.num_recs = 0
self.genres = {}
super().__init__()
def run(self):
CLIENT_ID = os.getenv('CLIENT_ID')
CLIENT_SECRET = os.getenv('CLIENT_SECRET')
AUTH_URL = 'https://accounts.spotify.com/api/token'
# POST
auth_response = requests.post(AUTH_URL, {
'grant_type': 'client_credentials',
'client_id': CLIENT_ID,
'client_secret': CLIENT_SECRET,
})
# convert the response to JSON
auth_response_data = auth_response.json()
# save the access token
access_token = auth_response_data['access_token']
headers = {
'Authorization': 'Bearer {token}'.format(token=access_token)
}
total = 0
underrepresented = 0
# Request playlist information
BASE_URL = 'https://api.spotify.com/v1/playlists/{playlist_id}'
self.playlist = requests.get(BASE_URL.format(playlist_id=self.id), headers=headers).json()
tracks = []
if(self.playlist != None and 'tracks' in self.playlist):
tracks = self.playlist['tracks']
if('items' in tracks):
artist_result = 'UND'
all_artists={}
# Start looping through all songs
for track in tracks['items']:
track_artists = track['track']['artists']
for artist in track_artists:
total += 1
artist_id = artist['id']
if artist_id in all_artists:
all_artists[artist_id] += 1
else:
all_artists[artist_id] = 1
artist_ids = list(all_artists.keys())
progress_count = 0
while(len(artist_ids) > 0):
# Group artists into batches of 50
batch = artist_ids[0]
artist_ids.pop(0)
count = 1
while(count < 50 and len(artist_ids) > 0):
batch += "," + artist_ids[0]
artist_ids.pop(0)
count += 1
# Request information from a batch of artists
BASE_URL='https://api.spotify.com/v1/artists?ids={batch}'
artist_batch = requests.get(BASE_URL.format(batch=batch), headers=headers).json()
if(artist_batch != None):
artist_batch = artist_batch["artists"]
for artist_info in artist_batch:
genres = artist_info['genres']
# Keep a list of the top genres in the playlist
for genre in genres:
if(genre in self.genres):
self.genres[genre] += 1
elif(not genre in FUNDAMENTAL_GENRES):
self.genres[genre] = 1
# Analyze and classify artist
artist_result = "UND"
try:
artist_result = analyze(artist_info)
except Exception as e:
print(str(e))
artist_info['occurrences'] = all_artists[artist_info['id']]
if(artist_result == "F"):
underrepresented += artist_info['occurrences']
self.artists['female'][artist_info['id']] = artist_info
elif(artist_result == "X"):
underrepresented += artist_info['occurrences']
self.artists['nonbinary'][artist_info['id']] = artist_info
elif(artist_result == "M"):
self.artists['male'][artist_info['id']] = artist_info
elif(artist_result == "MIX"):
underrepresented += artist_info['occurrences']
self.artists['mixed_gender'][artist_info['id']] = artist_info
else:
self.artists['undetermined'][artist_info['id']] = artist_info
print("RESULT: " + artist_result)
progress_count += 1
# Update the progress of the playlist
self.progress = 0.95 * (progress_count / total)
self.result = str(round((underrepresented / total) * 100, 1))
# Start generating recommendations for the playlist
rec_possible = True
self.num_recs = 0
exclude = list(all_artists.keys())
# Create a sorted list of the top genres
genres_ordered = dict(sorted(self.genres.items(), key=lambda item: item[1]))
top_genres = list(genres_ordered.keys())[-3:]
# Get a mainstream artist, an emerging artist, and an obscure artist
for popularity in self.recs:
rec = None
if(rec_possible):
# Pick one of the playlist's top three genres
genre_start_index = random.randint(0, 2)
genre_index = genre_start_index
# Try to retrieve a recommendation related to that artist with the popularity level we want
# If a recommendation cannot be generated, go to the next artist
# And loop until we've checked all artists in the playlist
while(rec == None and genre_index != (genre_start_index - 1) % len(top_genres)):
rec = generate_rec(exclude, top_genres[genre_index], popularity)
genre_index = (genre_index + 1) % len(top_genres)
# If we couldn't get a recommendation, repeat the same process but disregard the popularity level
if(rec == None):
genre_index = genre_start_index
while(rec == None and genre_index != (genre_start_index - 1) % len(top_genres)):
rec = generate_rec(exclude, top_genres[genre_index])
genre_index = (genre_index + 1) % len(top_genres)
# If a recommendation still isn't possible, return an empty list
if(rec == None):
rec = generate_rec(exclude)
# At this point, note that it isn't possible to generate a recommendation from this playlist
# so don't bother to go through that whole process for the remaining recommendations
rec_possible = False
# Track how many recommendations it was possible to generate
else:
self.num_recs += 1
exclude.append(rec[0]) # Excludes duplicate recommendations
else:
self.num_recs += 1
exclude.append(rec[0]) # Excludes duplicate recommendations
else:
rec = generate_rec(exclude)
self.recs[popularity] = rec
self.progress = 1
# Go back through and update recommendations tables for underrepresented artists in this playlist
underrepresented_artists = {**self.artists['female'], **self.artists['nonbinary']}
for artist in underrepresented_artists:
update_recs_table(underrepresented_artists[artist])
else:
self.result = "Playlist not found"
# Determine the gender of an artist
def analyze(artist_json):
# Query the database to check the gender of the artist
id = artist_json['id']
artist = artist_json['name']
result = analyze_from_database(id)
artist_image = ""
if(len(artist_json['images']) > 0):
artist_image = artist_json['images'][0]['url']
popularity = sort_artist(artist_json)
# If we didn't get a result, determine the artist's gender by analyzing their last.fm bio
if(result == "UND"):
result = analyze_from_chartmetric(artist)
if(result == "UND"):
result = analyze_via_crawl(id, artist)
check_if_exists = execute_read_query(f"SELECT * FROM artists WHERE spotify_id='{id}'")
query = ""
# If we hadn't analyzed this artist before, add a new row to store their result
if(check_if_exists == None):
escaped_name = escape_sql_string(artist)
query = f"INSERT INTO artists (spotify_id, name, picture, popularity, consensus) VALUES ('{id}', " + \
f"'{escaped_name}', '{artist_image}', '{popularity}', '{result}');"
# Otherwise, update the existing row
else:
query = f"UPDATE artists SET consensus='{result}', picture='{artist_image}', popularity='{popularity}' WHERE spotify_id='{id}'"
execute_query(query)
# If we did get a result, make sure the image and popularity are up-to-date
else:
query = f"UPDATE artists SET picture='{artist_image}', popularity='{popularity}' WHERE spotify_id='{id}'"
execute_query(query)
return result
# Update the table of recommendations by genre
def update_recs_table(rec_artist):
create_recs_table = f"""
CREATE TABLE IF NOT EXISTS recs (
artist_id TEXT NOT NULL,
popularity TEXT NOT NULL,
genre1 TEXT NOT NULL,
genre2 TEXT NOT NULL,
genre3 TEXT NOT NULL,
genre4 TEXT NOT NULL,
genre5 TEXT NOT NULL
);
"""
execute_query(create_recs_table)
# Find the table we want to update
artist_row = execute_read_query(f"SELECT * from recs WHERE artist_id='{rec_artist['id']}'")
# Get a list of the artist's genres outside of the fundamental genres
genres_list = rec_artist['genres']
for genre in genres_list:
genre = escape_sql_string(genre)
if genre in FUNDAMENTAL_GENRES:
genres_list.remove(genre)
while(len(genres_list) < 5):
genres_list.append('')
# If the recommendation wasn't there before, add a new row
if(artist_row == None):
row = {
"artist_id":rec_artist['id'],
"popularity":sort_artist(rec_artist),
"genre1": escape_sql_string(genres_list[0]),
"genre2": escape_sql_string(genres_list[1]),
"genre3": escape_sql_string(genres_list[2]),
"genre4": escape_sql_string(genres_list[3]),
"genre5": escape_sql_string(genres_list[4])
}
query = f"INSERT INTO recs (artist_id, popularity, genre1, genre2, genre3, genre4, genre5) VALUES " + \
f"('{row['artist_id']}', '{row['popularity']}', '{row['genre1']}', '{row['genre2']}', '{row['genre3']}', " + \
f"'{row['genre4']}', '{row['genre5']}');"
execute_query(query)
# Otherwise, update the recommendation's existing row
else:
new_pop = sort_artist(rec_artist)
query = f"UPDATE recs SET popularity='{new_pop}' " + \
f"WHERE artist_id='{rec_artist['id']}'"
execute_query(query)
# Categorize the artist's level of popularity
def sort_artist(artist):
if(artist['popularity'] > 60):
return '60-100'
elif(artist['popularity'] > 30):
return '30-60'
else:
return '0-30'
# Generate a recommendation based on an artist and a playlist
def generate_rec(exclude, genre=None, popularity=None):
create_recs_table = f"""
CREATE TABLE IF NOT EXISTS recs (
artist_id TEXT NOT NULL,
popularity TEXT NOT NULL,
genre1 TEXT NOT NULL,
genre2 TEXT NOT NULL,
genre3 TEXT NOT NULL,
genre4 TEXT NOT NULL,
genre5 TEXT NOT NULL
);
"""
execute_query(create_recs_table)
genre = escape_sql_string(genre)
recs_table = None
if(popularity == None and genre != None):
# Retrieves a list of one-element tuples (ID of recommendation)
recs_table = execute_read_multiple_query(f"SELECT artist_id FROM recs WHERE genre1='{genre}' OR genre2='{genre}' " + \
f" OR genre3='{genre}' OR genre4='{genre}' OR genre5='{genre}'")
elif(popularity != None and genre != None):
recs_table = execute_read_multiple_query(f"SELECT artist_id FROM recs WHERE genre1='{genre}' OR genre2='{genre}' " + \
f" OR genre3='{genre}' OR genre4='{genre}' OR genre5='{genre}' AND popularity='{popularity}'")
else:
recs_table = execute_read_query(f"SELECT artist_id FROM recs ORDER BY RAND() LIMIT 50")
if(recs_table == None or len(recs_table) == 0):
return None
rec_id = random.choice(recs_table)
rec_artist = execute_read_query(f"SELECT spotify_id, name, popularity, picture, consensus FROM artists WHERE spotify_id='{rec_id[0]}'")
# If the recommendation is already in the playlist, or the recommendation is incorrect (category M or UND), remove it and try again
while(len(recs_table) > 0 and rec_id[0] in exclude or (rec_artist[4] == 'M' or rec_artist[4] == 'UND')):
recs_table.remove(rec_id)
if (rec_artist[4] == 'M' or rec_artist[4] == "UND"):
execute_query(f"DELETE from recs WHERE artist_id='{rec_artist[0]}'")
if(len(recs_table) > 0):
rec_id = random.choice(recs_table)
rec_artist = execute_read_query(f"SELECT spotify_id, name, popularity, picture, consensus FROM artists WHERE spotify_id='{rec_id[0]}'")
if(len(recs_table) == 0):
return None
return rec_artist
# Determine an artist's gender by crawling their last.fm bio
def analyze_via_crawl(id, artist, individual=False):
# Get the HTML code of their last.fm page
artist = artist.replace("/", "%2F")
artist_page = "https://www.last.fm/music/" + artist + "/+wiki"
page = requests.get(artist_page)
soup = BeautifulSoup(page.content, "html.parser")
content = soup.find(class_="wiki-content")
no_data = soup.find(class_="no-data-message")
# If no data was found, return undetermined
if(content == None or no_data != None):
return "UND"
# Otherwise, grab the bio paragraph by paragraph (not to exceed 2500 characters)
content = content.find_all("p")
paragraph_count = 0
bio_short = ""
while len(bio_short) < 2500 and paragraph_count < len(content):
bio_short += content[paragraph_count].text.lower() + " "
paragraph_count += 1
content = content[:paragraph_count]
# Words that may indicate the gender of the artist
masc_indicators = ['he', 'him', 'his', 'himself', 'frontman', 'boy']
fem_indicators = ['she', 'her', 'hers', 'herself', 'female', 'female-fronted', 'frontwoman', 'girl']
nonbinary_indicators = ['they', 'them', 'theirs', 'their', 'themself', 'they/them', 'non-binary', 'nonbinary']
bio_words = bio_short.split()
masc_count = 0
fem_count = 0
neutral_count = 0
isGroup = False
# Bands tend to have a "factbox" containing a list of their members
# Find that factbox if it exists
factbox = soup.find(class_="factbox")
factbox_items = None
members_list = None
# If there is a factbox
if(factbox != None):
factbox_items = factbox.find_all(class_="factbox-item")
for item in factbox_items:
heading = item.find(class_="factbox-heading")
# And if that factbox contains a list of members
if(heading != None and heading.text == "Members"):
# Then this artist is a group
isGroup = True
members_list = item
# Count how many times the gendered "indicators" appear in the bio
for indicator in masc_indicators:
masc_count += bio_words.count(indicator)
for indicator in fem_indicators:
fem_count += bio_words.count(indicator)
for indicator in nonbinary_indicators:
neutral_count += bio_words.count(indicator)
# If there is no factbox but the bio contains the following terms, then artist may be a group
if("was formed" in bio_short or "formed in" in bio_short or "consists of" in bio_short or "consisting of" in bio_short):
isGroup = True
# If we want to "force" analyzing this artist as an individual, rather than a group,
# override what we may have determined. (I.e. if we are calling this recursively on the members
# of a group.)
if(individual):
isGroup = False
result = "UND"
# If the artist is a group,
if(isGroup):
# Analyze their tags to determine the vocalists' gender
tags_analysis = analyze_based_on_tags(artist)
# If this isn't succesful, recursively analyze the bios
# of the band's individual members (if a member list was given)
if(tags_analysis == "UND"):
if(members_list != None):
members = members_list.find_all("li")
member_searches = []
for member in members:
member_name = member.find("span").text
result = analyze_via_crawl(None, member_name, individual=True)
if(result != "UND"):
member_searches.append(result)
# Determine how many genders are represented amongst the group's members
unique = set(member_searches)
if(len(unique) > 0):
if(len(unique) > 2):
result = "MIX"
else:
result = unique.pop()
else:
result = "UND"
else:
result = tags_analysis
# If the artist is an individual, use their bio/pronouns to determine their gender
elif("frontman" in bio_short or "boy band" in bio_short or \
(masc_count > fem_count and masc_count > neutral_count)):
result = "M"
elif("frontwoman" in bio_short or "female-fronted" in bio_short or "girl group" in bio_short or "all-girl" in bio_short or \
(fem_count > masc_count and fem_count > neutral_count)):
result = "F"
elif(("nonbinary" in bio_short or "non-binary" in bio_short or "they/them" in bio_short) or \
(neutral_count > masc_count and neutral_count > fem_count)):
result = "X"
# Bio takes priority over tags (especially since tags are binary in gender)
# But if bio analysis wasn't successful, check the tags
else:
result = analyze_based_on_tags(artist)
# if(id != None):
return result
# Determine the artist's gender based on tags
def analyze_based_on_tags(artist):
# Get the HTML code of their last.fm page
artist_page = "https://www.last.fm/music/" + artist + "/+tags"
page = requests.get(artist_page)
soup = BeautifulSoup(page.content, "html.parser")
# Find their tags
tags = soup.find_all(class_="big-tags-item")
for tag in tags:
tag_name = tag.find(class_="link-block-target").text
# If the artist is tagged with their gender, return that result
if(tag_name.lower() == "female vocalists" or tag_name.lower() == "girl group"):
return "F"
elif(tag_name.lower() == "male vocalists" or tag_name.lower() == "boy band"):
return "M"
return "UND"
# Determine an artist's gender based on what we've recorded in our database
def analyze_from_database(id):
# execute_query("DROP TABLE artists")
create_artists_table = """
CREATE TABLE IF NOT EXISTS artists (
spotify_id VARCHAR(128) NOT NULL PRIMARY KEY,
name TEXT NOT NULL,
picture TEXT DEFAULT '',
popularity TEXT,
votes_m INTEGER DEFAULT 0,
votes_f INTEGER DEFAULT 0,
votes_x INTEGER DEFAULT 0,
votes_mix INTEGER DEFAULT 0,
consensus TEXT DEFAULT 'UND',
locked INTEGER DEFAULT 0
)
"""
execute_query(create_artists_table)
result = execute_read_query(f"SELECT consensus FROM artists WHERE spotify_id='{id}'")
if(result == None):
return "UND"
else:
return result[0]
def analyze_from_chartmetric(name):
escaped_name = escape_sql_string(name)
result = execute_read_query(f"SELECT result FROM chartmetric WHERE name='{escaped_name}'")
if(result == None):
return "UND"
else:
execute_query(f"DELETE from chartmetric WHERE name='{escaped_name}'")
return result[0]
def cast_vote(artist_id, category):
category_to_update = "votes_" + category.lower()
query = f"UPDATE artists SET {category_to_update} = {category_to_update} + 1 WHERE spotify_id='{artist_id}'"
execute_query(query)
# Update artist category after fact-checking
def update_result(artists_list, updates_json):
for category in artists_list:
artists_popped = []
for id in artists_list[category]:
if id in updates_json:
if 'category' in updates_json[id]:
if(updates_json[id]['category'] == 'M'):
artists_list['male'][id] = artists_list[category][id]
cast_vote(id, 'M')
artists_popped.append(id)
elif(updates_json[id]['category'] == 'F'):
artists_list['female'][id] = artists_list[category][id]
cast_vote(id, 'F')
artists_popped.append(id)
elif(updates_json[id]['category'] == 'X'):
artists_list['nonbinary'][id] = artists_list[category][id]
cast_vote(id, 'X')
artists_popped.append(id)
elif(updates_json[id]['category'] == 'MIX'):
artists_list['mixed_gender'][id] = artists_list[category][id]
cast_vote(id, 'MIX')
artists_popped.append(id)
# Remove all artists from their old categories as displayed
for id in artists_popped:
artists_list[category].pop(id)
return artists_list
def get_unlocked(artists_list):
artists_unlocked = artists_list
for category in artists_list:
for id in category:
locked = execute_read_query(f"SELECT locked FROM artists WHERE spotify_id='{id}'")
if(locked == 1):
artists_unlocked[category].pop(id)
return artists_unlocked
def escape_sql_string(string):
if(string != None):
return string.replace('\0', '\\0').replace('\'', '\'\'').replace('\"', '\"\"').replace('\b', '\\b') \
.replace('\n', '\\n').replace('\r', '\\r').replace('\t', '\\t').replace('\Z', '\\Z').replace('\\', '\\\\') \
.replace('%', '\%').replace('_', '\_')
else:
return ""
# Execute SQL query
def execute_query(query):
global connection
if(connection == None):
# connection = psycopg2.connect(os.getenv('DATABASE_URL'), sslmode='require')
connection = psycopg2.connect(
host=os.getenv('DB_HOST'),
database="postgres",
user="postgres",
password=os.getenv('DB_PASSWORD'))
connection.autocommit = True
cursor = connection.cursor()
try:
cursor.execute(query)
except OperationalError as e:
print(f"The error '{e}' occurred")
# Read single row from SQL table
def execute_read_query(query):
global connection
if(connection == None):
connection = psycopg2.connect(
host=os.getenv('DB_HOST'),
database="postgres",
user="postgres",
password=os.getenv('DB_PASSWORD'))
connection.autocommit = True
cursor = connection.cursor()
result = None
try:
cursor.execute(query)
result = cursor.fetchone()
return result
except OperationalError as e:
print(f"The error '{e}' occurred")
# Read multiple rows from SQL table
def execute_read_multiple_query(query):
global connection
if(connection == None):
connection = psycopg2.connect(
host=os.getenv('DB_HOST'),
database="postgres",
user="postgres",
password=os.getenv('DB_PASSWORD'))
connection.autocommit = True
cursor = connection.cursor()
result = None
try:
cursor.execute(query)
result = cursor.fetchall()
return result
except OperationalError as e:
print(f"The error '{e}' occurred")