Add detection for null/empty files

[atruskie]

Sometimes sensors produce allocated space but otherwise empty files. We should add a detector for those.

This problem seems to be distinct from OE004.

corrupt_to_investigate.zip

[Adityapandeya]

import os
import math
from collections import Counter

def file_is_empty(filepath):
    try:
        with open(filepath, 'rb') as file:
            data = file.read(1024)  # Read the first 1024 bytes
            if not data:
                return True  # File is empty or has no readable content
            if data.count(data[:1]) == len(data):
                return True  # All bytes in the sample are the same
    except IOError as e:
        print(f"Error reading {filepath}: {e}")
    return False

def calculate_entropy(filepath):
    try:
        with open(filepath, 'rb') as file:
            data = file.read()
            if not data:
                return 0
            entropy = 0
            for x in Counter(data).values():
                p_x = x / len(data)
                entropy -= p_x * math.log2(p_x)
            return entropy
    except IOError as e:
        print(f"Error reading {filepath}: {e}")
    return None

# Example usage
path = '/path/to/suspected/file'
print("Empty:", file_is_empty(path))
print("Entropy:", calculate_entropy(path))

[atruskie]

@Adityapandeya I'm not sure what the point to your comment was but without further context, it was not helpful.