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
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
|
use std::fmt;
use std::fs;
use crate::config::*;
const fn build_lookup_table() -> [usize; 128] {
let mut result = [0; 128];
let mut i = 0;
while i < NUM_KEYS {
result[KEY_CHARS[i] as usize] = i;
i += 1;
}
return result;
}
const CHAR_TO_INDEX: [usize; 128] = build_lookup_table();
const NUM_BIGRAMS: usize = NUM_KEYS * NUM_KEYS;
pub struct Corpus {
bigram_percs: [f32; NUM_BIGRAMS],
skipgram_percs: [f32; NUM_BIGRAMS],
character_count: [u32; 128],
total_count: u32,
total_bigrams: u32,
}
fn pair_to_index(x: u8, y: u8) -> usize {
let xi = CHAR_TO_INDEX[x as usize];
let yi = CHAR_TO_INDEX[y as usize];
xi + NUM_KEYS * yi
}
impl Corpus {
pub fn get_character_count(&self, c: u8) -> u32 {
self.character_count[c as usize]
}
pub fn get_bigram_perc(&self, x: u8, y: u8) -> f32 {
self.bigram_percs[pair_to_index(x, y)]
}
pub fn get_skipgram_perc(&self, x: u8, y: u8) -> f32 {
self.skipgram_percs[pair_to_index(x, y)]
}
pub fn get_total_count(&self) -> u32 {
self.total_count
}
pub fn get_total_bigrams(&self) -> u32 {
self.total_bigrams
}
pub fn load(path: &str) -> Result<Corpus, std::io::Error> {
let contents = fs::read_to_string(path)?;
let mut bigram_count = [0; NUM_BIGRAMS];
let mut skipgram_count = [0; NUM_BIGRAMS];
let mut character_count = [0; 128];
let mut total_count = 0;
let mut total_bigrams = 0;
let mut total_trigrams = 0;
let mut last_char: Option<u8> = None;
let mut last_last_char: Option<u8> = None;
for c in contents.chars() {
let current_char = canonicalise(c);
if let Some(c) = current_char {
if let Some(lc) = last_char {
if let Some(llc) = last_last_char {
if llc != c {
skipgram_count[pair_to_index(llc, c)] += 1;
skipgram_count[pair_to_index(c, llc)] += 1;
}
total_trigrams += 1;
}
if lc != c {
bigram_count[pair_to_index(c, lc)] += 1;
bigram_count[pair_to_index(lc, c)] += 1;
}
total_bigrams += 1;
}
character_count[c as usize] += 1;
total_count += 1;
}
last_last_char = last_char;
last_char = current_char;
}
let mut bigram_percs = [0.0; NUM_BIGRAMS];
let mut skipgram_percs = [0.0; NUM_BIGRAMS];
for i in 0..NUM_BIGRAMS {
bigram_percs[i] = bigram_count[i] as f32 / total_bigrams as f32;
skipgram_percs[i] = skipgram_count[i] as f32 / total_trigrams as f32;
}
return Ok(Corpus {
skipgram_percs,
bigram_percs,
character_count,
total_count,
total_bigrams,
});
}
}
fn dump_bigrams(corpus: &Corpus) -> Vec<(String, f32)> {
let mut result: Vec<(String, f32)> = Vec::new();
for (i, &x) in KEY_CHARS.iter().enumerate() {
for &y in &KEY_CHARS[i..] {
let mut pair = String::from(x);
pair.push(y);
result.push((pair, corpus.bigram_percs[pair_to_index(x as u8, y as u8)]));
}
}
result.sort_by(|(_, c1), (_, c2)| {
if c2 < c1 {
std::cmp::Ordering::Less
} else {
std::cmp::Ordering::Greater
}
});
return result;
}
impl fmt::Display for Corpus {
fn fmt(&self, formatter: &mut fmt::Formatter) -> fmt::Result {
formatter.write_str(&format!(
"Counted {} characters, top 5 bigrams {:?}",
self.total_count,
&dump_bigrams(self)[0..5]
))
}
}
|
