import os
import mido
from tqdm import tqdm

folder_path = 'LAKH-MuseNet-CC-BY-NC-SA2'
output_file = 'names4.txt'
max_file_size_kb = 400

with open(output_file, 'w', encoding="utf-8") as f:
    midi_files = []
    for subdir, dirs, files in os.walk(folder_path):
        for file in files:
            if file.endswith('.mid'):
                # f.write(filename + '\n')
                file_path = os.path.join(subdir, file)
                midi_files.append(file_path)
    for file_path in tqdm(midi_files, desc='writing MIDI name/path list', unit='file'):
        try:
            file_size_kb = os.path.getsize(file_path) / 1024  # Get file size in KB
            if file_size_kb <= max_file_size_kb:
                f.write(file_path + '\n')
            else:
                print(f'File is too large: {file_path}')
        except Exception as e:
            print(f'Error processing file {file_path}: {e}')