minor changes

scrape.py

@@ -1,6 +1,3 @@
-# def get_lyrics_url_from_website():
-#     # https://www.disneyclips.com/lyrics/
-
 import asyncio
 import json
 from collections import defaultdict
@@ -10,6 +7,10 @@ from typing import List, Optional, Tuple, TypedDict
 import aiohttp
 from bs4 import BeautifulSoup
 
+"""
+This file scrapes disney songs + lyrics from "https://www.disneyclips.com/lyrics/"
+"""
+
 URL = "https://www.disneyclips.com/lyrics/"
 
 

scripts/create_emotions_summary.py

@@ -1,3 +1,8 @@
+"""
+This script takes all the songs we have and use the lyric to create a list of 8 emotions we then use to replace the lyric itself.
+This is needed to properly match user's emotions to the songs.
+"""
+
 from dotenv import load_dotenv
 
 load_dotenv()

scripts/create_one_sentence_summary.py

@@ -1,3 +1,7 @@
+"""
+This script takes all the songs we have and create a summary for each lyric
+"""
+
 from dotenv import load_dotenv
 
 load_dotenv()
@@ -21,7 +25,7 @@ llm = ChatOpenAI(temperature=0)
 chain = LLMChain(llm=llm, prompt=prompt)
 
 with open(
-    "/home/zuppif/Documents/Work/ActiveLoop/ai-shazam/data/lyrics_with_spotify_url.json",
+    "data/lyrics_with_spotify_url.json",
     "r",
 ) as f:
     data = json.load(f)
@@ -38,7 +42,7 @@ for movie, lyrics in data.items():
         }
 
 with open(
-    "/home/zuppif/Documents/Work/ActiveLoop/ai-shazam/data/lyrics_with_spotify_url_and_summary.json",
+    "data/lyrics_with_spotify_url_and_summary.json",
     "w",
 ) as f:
     json.dump(lyrics_summaries, f)

scripts/keep_only_lyrics_on_spotify.py

@@ -1,9 +1,11 @@
-"""This script will keep only the lyrics that are in the Spotify "Disney Hits" playlist
+"""
+This script will keep only the songs that are in the Spotify "Disney Hits" playlist
 """
 from dotenv import load_dotenv
 
 load_dotenv()
 import json
+from collections import defaultdict
 
 import spotipy
 from spotipy.oauth2 import SpotifyClientCredentials
@@ -17,10 +19,6 @@ items = results["playlists"]["items"]
 uri = "spotify:playlist:37i9dQZF1DX8C9xQcOrE6T"
 playlist = spotify.playlist(uri)
 
-# with open("spotify_disney_songs.json", "w") as f:
-#     json.dump(playlist,f)
-
-
 with open("data/lyrics.json", "r") as f:
     data = json.load(f)
 
@@ -36,7 +34,6 @@ for item in playlist["tracks"]["items"]:
     }
 
 # here we add only songs that are in the Disney spotify playlist
-from collections import defaultdict
 
 data_filtered = defaultdict(list)
 tot = 0