Check video files with ffprobe if there is only one i-frame in the first 20 seconds of the file

i_frame_check.py

import os
import subprocess
import time
import sys
import select

def rename_defective_file(filepath):
    """Rename the defective file by appending '_(i_frame_defect)' before the file extension."""
    root, ext = os.path.splitext(filepath)
    new_name = f"{root}_(i_frame_defect){ext}"
    os.rename(filepath, new_name)
    return new_name

# Detailed introductory message
print("DESCRIPTION:")
print("------------")
print("This script will recursively search the current directory and its subdirectories for video files.")
print("Each video file will be checked using 'ffprobe' to determine if there's only one I-frame in the first 20 seconds.")
print("Files with only one I-frame in the first 20 seconds will be deemed 'defective'.")

# Provide options to the user
print("\nOPTIONS:")
print("1. Just output the full path of affected files to 'defect_files.txt'.")
print("2. Output the full path of affected files to 'defect_files.txt' and rename the defective files by appending '_(i_frame_defect)' to their filename.")

# Get user's choice
choice = input("\nChoose an option (1/2): ")

# Get a list of all video files in the current directory and subdirectories
video_extensions = ['.mp4', '.mkv', '.avi']
video_files = [os.path.join(root, file)
               for root, dirs, files in os.walk('.')
               for file in files if os.path.splitext(file)[1] in video_extensions]

defective_files = []

# Check each video file with ffprobe for I-frames in the first 20 seconds
for index, video_file in enumerate(video_files, start=1):
    print(f"Checking file {index}/{len(video_files)}: {video_file}")
    
    command = [
        'ffprobe', 
        '-select_streams', 'v', 
        '-show_frames', 
        '-read_intervals', '%+#20', 
        '-of', 'csv=p=0', 
        '-i', video_file
    ]
    
    result = subprocess.run(command, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    output = result.stdout.decode('utf-8')
    i_frames = sum(1 for line in output.splitlines() if line.split(',')[1] == 'I')
    
    if i_frames == 1:
        if choice == '2':
            video_file = rename_defective_file(video_file)
        defective_files.append(os.path.abspath(video_file))

# Write all defective file paths to defect_files.txt, each on a new line
with open('defect_files.txt', 'w') as f:
    for file in defective_files:
        f.write(file + '\n')

# Write all defective file paths to defect_files_one_line.txt in a single line
with open('defect_files_one_line.txt', 'w') as f:
    f.write(' '.join(defective_files))

print(f"\nFinished checking. Found {len(defective_files)} files with only one I-frame in the first 20 seconds.")
print(f"Check 'defect_files.txt' and 'defect_files_one_line.txt' for lists of defective files.")

ffmpeg6 with libvpl (onevpl) can produce files with a broken gop, when using qsv to encode to hevc because of bad default settings. Files have only a single i-frame. Even though the files seem intact, the have problems when seeking(long buffering, the bigger the file the worse). This script identifies those files and outputs the full filepath to a txt file for batch ops.

first option: filepaths to txt file
second option: same + rename the affected files by appending _(iframe_defect) between filename and extension