Create Dockerfile

convert_song_format_for_pagalworld.py

@@ -1,5 +1,5 @@
 import json
-
+import os # Import os for file existence checks
 
 def convert_song_format(original_song: dict, song_id: int) -> dict:
     """
@@ -12,11 +12,49 @@ def convert_song_format(original_song: dict, song_id: int) -> dict:
     Returns:
         dict: Converted song dictionary.
     """
+    # Extract title, handling cases where it might be a list or dict
+    title = ""
+    original_title = original_song.get("song") # Changed from "song " to "song"
+    if isinstance(original_title, str):
+        title = original_title.strip()
+    elif isinstance(original_title, dict) and "name" in original_title:
+        title = original_title["name"].strip()
+    elif isinstance(original_title, list) and original_title:
+        # If it's a list, take the first element's name if available
+        if isinstance(original_title[0], dict) and "name" in original_title[0]:
+            title = original_title[0]["name"].strip()
+        elif isinstance(original_title[0], str):
+            title = original_title[0].strip()
+
+    # Extract all artist names and join with a comma
+    artist_names = []
+    original_singer_data = original_song.get("singer")
+    if isinstance(original_singer_data, list):
+        for singer_entry in original_singer_data:
+            if isinstance(singer_entry, dict) and "name" in singer_entry:
+                artist_names.append(singer_entry["name"].strip())
+            elif isinstance(singer_entry, str):
+                artist_names.append(singer_entry.strip())
+    elif isinstance(original_singer_data, str): # Fallback if it's directly a string
+        artist_names.append(original_singer_data.strip())
+
+    artist_string = ", ".join(artist_names)
+
+
+    # Extract thumbnail URL
+    album_art_url = ""
+    original_thumbnail = original_song.get("thumbnail", "")
+    if isinstance(original_thumbnail, str):
+        album_art_url = original_thumbnail.strip()
+    elif isinstance(original_thumbnail, dict) and "url" in original_thumbnail:
+        album_art_url = original_thumbnail["url"].strip()
+
+
     return {
         "id": str(song_id),
-        "title": original_song.get("Song Name ", "").strip(),
-        "artist": original_song.get("Singer(s) ", "").strip(),
-        "albumArtUrl": original_song.get("Thumbnail", "").strip(),
+        "title": title,
+        "artist": artist_string, # Use the comma-separated string here
+        "albumArtUrl": album_art_url,
         "audioUrl": original_song.get("Play Online", "").strip()
     }
 
@@ -31,8 +69,15 @@ def read_json_file(file_path: str) -> list:
     Returns:
         dict or list: Parsed JSON data.
     """
-    with open(file_path, 'r', encoding='utf-8') as file:
-        return json.load(file)
+    try:
+        with open(file_path, 'r', encoding='utf-8') as file:
+            return json.load(file)
+    except FileNotFoundError:
+        print(f"Error: The file '{file_path}' was not found.")
+        return []
+    except json.JSONDecodeError:
+        print(f"Error: Could not decode JSON from '{file_path}'. Check file format.")
+        return []
 
 
 def write_json_file(data, file_path: str, indent: int = 2):
@@ -49,29 +94,36 @@ def write_json_file(data, file_path: str, indent: int = 2):
 
 
 def main():
-    bollywood_song_metadata_json = "bollywood_song_metadata.json"
+    bollywood_song_metadata_json = "combined_song_metadata.json"
     list_of_original_song = read_json_file(bollywood_song_metadata_json)
-    # print(list_of_original_song)
+
     songs_json = "songs.json"
 
-    list_of_song= []
+    list_of_song = []
 
+    try:
+        if os.path.exists(songs_json) and os.path.getsize(songs_json) > 0:
+            existing_songs = read_json_file(songs_json)
+            if isinstance(existing_songs, list):
+                list_of_song = existing_songs
+                print(f"Loaded {len(list_of_song)} existing songs from {songs_json}")
+    except Exception as e:
+        print(f"Could not load existing {songs_json} due to error: {e}. Starting with an empty list.")
 
-    # list_of_song = read_json_file(songs_json)
 
     i = len(list_of_song)
 
+    print(f"Initial len of song list: {len(list_of_song)}")
 
-    print(f"initial  len of song list {len(list_of_song)}")
     for original_song in list_of_original_song:
-        i = i + 1
         converted = convert_song_format(original_song, song_id=i)
         list_of_song.append(converted)
+        i = i + 1
 
-    print(f"final len of song list {len(list_of_song)}")
-    print(list_of_song)
+    print(f"Final len of song list: {len(list_of_song)}")
     write_json_file(list_of_song, songs_json)
+    print(f"Converted song data saved to {songs_json}")
 
 
 if __name__ == "__main__":
-    main()
+    main()