import gradio as gr import json from collections import Counter, defaultdict import os def analyze_spotify_data(files): # files: list of file objects # We'll parse each JSON file and aggregate the data all_records = [] for f in files: try: data = json.load(open(f)) if isinstance(data, list): all_records.extend(data) else: # If the JSON file isn't a list at top-level, skip or handle differently continue except: # If there's an error in loading JSON, skip that file continue # If no valid data found if not all_records: return "No valid data found in the uploaded files." # Aggregate listening stats artist_counter = Counter() track_counter = Counter() album_counter = Counter() # Note: album info not provided in the sample data, will only work if album is in the data # We also want to consider total listening time per artist/track/album artist_time = defaultdict(int) track_time = defaultdict(int) album_time = defaultdict(int) # Attempt to detect if albumName is present album_present = all("albumName" in record for record in all_records if isinstance(record, dict)) for record in all_records: if not isinstance(record, dict): continue artist = record.get("artistName", "Unknown Artist") track = record.get("trackName", "Unknown Track") ms_played = record.get("msPlayed", 0) # Album may not be present; handle gracefully album = record.get("albumName", "Unknown Album") if album_present else None artist_counter[artist] += 1 track_counter[track] += 1 artist_time[artist] += ms_played track_time[track] += ms_played if album_present and album is not None: album_counter[album] += 1 album_time[album] += ms_played # Determine top artists by number of tracks played (frequency) and also by time top_artists_by_count = artist_counter.most_common(10) top_artists_by_time = sorted(artist_time.items(), key=lambda x: x[1], reverse=True)[:10] # Determine top tracks by frequency and by time top_tracks_by_count = track_counter.most_common(10) top_tracks_by_time = sorted(track_time.items(), key=lambda x: x[1], reverse=True)[:10] # Determine top albums if available if album_present: top_albums_by_count = album_counter.most_common(10) top_albums_by_time = sorted(album_time.items(), key=lambda x: x[1], reverse=True)[:10] else: top_albums_by_count = [("No album data found", 0)] top_albums_by_time = [("No album data found", 0)] # Format the results into a readable output def format_list(title, data_list, time_data=False): result = f"**{title}**\n" if not time_data: for i, (name, count) in enumerate(data_list, 1): result += f"{i}. {name} ({count} plays)\n" else: for i, (name, ms) in enumerate(data_list, 1): hours = ms / (1000*60*60) result += f"{i}. {name} ({hours:.2f} hours)\n" result += "\n" return result output = "" output += format_list("Top Artists by Play Count", top_artists_by_count, time_data=False) output += format_list("Top Artists by Listening Time", top_artists_by_time, time_data=True) output += format_list("Top Tracks by Play Count", top_tracks_by_count, time_data=False) output += format_list("Top Tracks by Listening Time", top_tracks_by_time, time_data=True) output += format_list("Top Albums by Play Count", top_albums_by_count, time_data=False) output += format_list("Top Albums by Listening Time", top_albums_by_time, time_data=True) return output with gr.Blocks() as demo: gr.Markdown("# Spotify Listening Data Analyzer") gr.Markdown("Upload your Spotify JSON files (e.g., 'StreamingHistory0.json', 'StreamingHistory1.json', etc.) to get an overview of your top artists, albums, and tracks.") file_input = gr.File(file_count="multiple", type="filepath", label="Upload JSON files") analyze_button = gr.Button("Analyze") output_box = gr.Markdown() analyze_button.click(fn=analyze_spotify_data, inputs=file_input, outputs=output_box) if __name__ == "__main__": demo.launch()