bleuscore 0.2.0

A fast bleu score calculator
Documentation 
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

use bleuscore::tokenizer::{Tokenizer, Tokenizer13a, TokenizerRegex};
use divan::black_box;

fn main() {
    // Run registered benchmarks.
    divan::main();
}

/// Benchmark Tokenizer13a with short text
#[divan::bench(sample_count = 1000, sample_size = 100)]
fn tokenizer_13a_short() {
    let tokenizer = Tokenizer13a::new();
    let line = "Hello, &quot;World!<skipped>";

    black_box(tokenizer.tokenize(black_box(line)));
}

/// Benchmark Tokenizer13a with medium text
#[divan::bench(sample_count = 1000, sample_size = 100)]
fn tokenizer_13a_medium() {
    let tokenizer = Tokenizer13a::new();
    let line = "/usr/sbin/sendmail - 0 errors, 12 warnings";

    black_box(tokenizer.tokenize(black_box(line)));
}

/// Benchmark Tokenizer13a with long text
#[divan::bench(sample_count = 500, sample_size = 100)]
fn tokenizer_13a_long() {
    let tokenizer = Tokenizer13a::new();
    let line = "The quick brown fox jumps over the lazy dog. This is a sample sentence for testing tokenization with longer text sequences.";

    black_box(tokenizer.tokenize(black_box(line)));
}

/// Benchmark Tokenizer13a with various text lengths
#[divan::bench(args = [10, 50, 100, 200, 500], sample_count = 500, sample_size = 50)]
fn tokenizer_13a_varying_length(word_count: usize) -> Vec<String> {
    let tokenizer = Tokenizer13a::new();
    let line = "word ".repeat(word_count);

    black_box(tokenizer.tokenize(black_box(&line)))
}

/// Benchmark TokenizerRegex with short text
#[divan::bench(sample_count = 1000, sample_size = 100)]
fn tokenizer_regex_short() {
    let tokenizer = TokenizerRegex::new();
    let line = "Hello, World!";

    black_box(tokenizer.tokenize(black_box(line)));
}

/// Benchmark TokenizerRegex with medium text
#[divan::bench(sample_count = 1000, sample_size = 100)]
fn tokenizer_regex_medium() {
    let tokenizer = TokenizerRegex::new();
    let line = "/usr/sbin/sendmail - 0 errors, 12 warnings";

    black_box(tokenizer.tokenize(black_box(line)));
}

/// Benchmark TokenizerRegex with long text
#[divan::bench(sample_count = 500, sample_size = 100)]
fn tokenizer_regex_long() {
    let tokenizer = TokenizerRegex::new();
    let line = "The quick brown fox jumps over the lazy dog. This is a sample sentence for testing tokenization with longer text sequences.";

    black_box(tokenizer.tokenize(black_box(line)));
}

/// Compare both tokenizers
#[divan::bench(types = [Tokenizer13a, TokenizerRegex], sample_count = 1000, sample_size = 100)]
fn tokenizer_comparison<T>()
where
    T: Tokenizer + Default,
{
    let tokenizer = T::default();
    let line = "Hello, World! This is a test.";

    black_box(tokenizer.tokenize(black_box(line)));
}

/// Benchmark tokenizer with special characters
#[divan::bench(sample_count = 1000, sample_size = 100)]
fn tokenizer_special_chars() {
    let tokenizer = Tokenizer13a::new();
    let line = "Hello, &quot;World&amp;Rust&lt;Code&gt;!<skipped>";

    black_box(tokenizer.tokenize(black_box(line)));
}