1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109

use std::io::Read;
use std::fs::File;

/// Contains metadata about the file that's being used in the Entropy calculation.
///
/// `byte_count` is a lookup table that contains the number of occurances of
/// a byte specified by the index, e.g. 0x00 is `byte_count[0]`.
///
/// `length` is the number of bytes in the file.
pub struct Entropy {
    pub byte_count: [u64; 256],
    pub length: u64,
}

impl Entropy {
    /// Gets metadata for the Entropy calculation from a File reference
    pub fn new(file: &File) -> Entropy {
        let mut byte_count = [0u64; 256];
        for byte in file.bytes() {
            byte_count[byte.unwrap() as usize] += 1
        }

        Entropy {
            byte_count: byte_count,
            length: file.metadata().unwrap().len(),
        }
    }

    /// Measures the Shannon entropy based on the frequency table and returns
    /// it as a float.
    ///
    /// The equation is defined as: H(X) = - \sum_{i=0}^{n} P(x_i) log_2 P(x_i)
    /// It can be described as the minimum number of bits (per symbol) to encode
    /// the input. Thus the output will be between 0 and 8.
    /// See https://en.wikipedia.org/wiki/Entropy_(information_theory) for
    /// more information.
    pub fn shannon_entropy(&self) -> f32 {
        let mut entropy = 0f32;
        for &count in self.byte_count.iter() {
            if count != 0 {
                let symbol_probability = count as f32 / self.length as f32;
                entropy += symbol_probability * symbol_probability.log2();
            }
        }
        -entropy
    }

    /// Measures the metric entropy based on the Shannon entropy of the
    /// generated frequency table and returns it as a float between 0 and 1.
    ///
    /// Metric entropy is derived by dividing the Shannon entropy by the length
    /// of the string being measured.
    /// It can be described as the uncertainty or randomness of a string, where
    /// 1 means information is uniformly distributed across the string.
    pub fn metric_entropy(&self) -> f32 {
        self.shannon_entropy() / 8f32
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use std::io::{Seek, SeekFrom, Write};
    use tempfile::tempfile;

    #[test]
    fn test_new() {
        // Create a temporary file and write five bytes to it
        let mut test_file = tempfile().unwrap();
        test_file.write(&[0x00, 0x00, 0x01, 0x01, 0x02]).unwrap();

        test_file.seek(SeekFrom::Start(0)).unwrap();
        let test_entropy = Entropy::new(&test_file);

        // Test that the frequency table was populated correctly
        assert_eq!(test_entropy.byte_count[0], 2);
        assert_eq!(test_entropy.byte_count[1], 2);
        assert_eq!(test_entropy.byte_count[2], 1);

        // Test the length
        assert_eq!(test_entropy.length, 5);
    }

    #[test]
    fn test_shannon_entropy() {
        // Create a temporary file and write five bytes to it
        let mut test_file = tempfile().unwrap();
        test_file.write(&[0x00, 0x00, 0x01, 0x01, 0x02]).unwrap();

        test_file.seek(SeekFrom::Start(0)).unwrap();
        let test_entropy = Entropy::new(&test_file);

        let shannon_entropy = test_entropy.shannon_entropy();
        assert_eq!(shannon_entropy, 1.5219281);
    }

    #[test]
    fn test_metric_entropy() {
        // Create a temporary file and write five bytes to it
        let mut test_file = tempfile().unwrap();
        test_file.write(&[0x00, 0x00, 0x01, 0x01, 0x02]).unwrap();

        test_file.seek(SeekFrom::Start(0)).unwrap();
        let test_entropy = Entropy::new(&test_file);

        let metric_entropy = test_entropy.metric_entropy();
        assert_eq!(metric_entropy, 0.19024101);
    }
}