provenant/license_detection/
token_multiset.rs1use std::collections::HashMap;
2use std::ops::Deref;
3
4use rkyv::Archive;
5
6use crate::license_detection::index::dictionary::{TokenDictionary, TokenId, TokenKind};
7
8#[derive(Clone, Debug, PartialEq, Eq, Default, Archive, rkyv::Serialize, rkyv::Deserialize)]
10pub struct TokenMultiset(HashMap<TokenId, usize>);
11
12impl TokenMultiset {
13 pub fn from_token_ids(token_ids: &[TokenId]) -> Self {
15 let mut counts = HashMap::new();
16
17 for &tid in token_ids {
18 *counts.entry(tid).or_insert(0) += 1;
19 }
20
21 Self(counts)
22 }
23
24 pub fn total_count(&self) -> usize {
26 self.0.values().sum()
27 }
28
29 pub fn high_subset(&self, dictionary: &TokenDictionary) -> Self {
31 self.0
32 .iter()
33 .filter(|(tid, _)| dictionary.token_kind(**tid) == TokenKind::Legalese)
34 .map(|(&tid, &count)| (tid, count))
35 .collect()
36 }
37
38 pub fn intersection(&self, other: &Self) -> Self {
40 let (smaller, larger) = if self.0.len() < other.0.len() {
41 (&self.0, &other.0)
42 } else {
43 (&other.0, &self.0)
44 };
45
46 smaller
47 .iter()
48 .filter_map(|(&tid, &count)| {
49 larger
50 .get(&tid)
51 .map(|&other_count| (tid, count.min(other_count)))
52 })
53 .collect()
54 }
55}
56
57impl Deref for TokenMultiset {
58 type Target = HashMap<TokenId, usize>;
59
60 fn deref(&self) -> &Self::Target {
61 &self.0
62 }
63}
64
65impl FromIterator<(TokenId, usize)> for TokenMultiset {
66 fn from_iter<T: IntoIterator<Item = (TokenId, usize)>>(iter: T) -> Self {
67 Self(iter.into_iter().collect())
68 }
69}
70
71#[cfg(test)]
72mod tests {
73 use super::*;
74 use crate::license_detection::index::dictionary::{TokenDictionary, tid};
75
76 #[test]
77 fn test_from_token_ids() {
78 let token_ids = vec![tid(1), tid(2), tid(3), tid(2), tid(4), tid(1), tid(1)];
79 let multiset = TokenMultiset::from_token_ids(&token_ids);
80
81 assert_eq!(multiset.get(&tid(1)), Some(&3));
82 assert_eq!(multiset.get(&tid(2)), Some(&2));
83 assert_eq!(multiset.get(&tid(3)), Some(&1));
84 assert_eq!(multiset.get(&tid(4)), Some(&1));
85 }
86
87 #[test]
88 fn test_total_count() {
89 let token_ids = vec![tid(1), tid(2), tid(3), tid(2), tid(1), tid(1)];
90 let multiset = TokenMultiset::from_token_ids(&token_ids);
91
92 assert_eq!(multiset.total_count(), 6);
93 }
94
95 #[test]
96 fn test_high_subset() {
97 let token_ids = vec![tid(1), tid(1), tid(2), tid(5), tid(10)];
98 let multiset = TokenMultiset::from_token_ids(&token_ids);
99 let dict = TokenDictionary::new_with_legalese(&[("one", 1), ("two", 2)]);
100
101 let high_multiset = multiset.high_subset(&dict);
102
103 assert_eq!(high_multiset.len(), 2);
104 assert_eq!(high_multiset.get(&tid(1)), Some(&2));
105 assert_eq!(high_multiset.get(&tid(2)), Some(&1));
106 assert!(!high_multiset.contains_key(&tid(5)));
107 assert!(!high_multiset.contains_key(&tid(10)));
108 }
109
110 #[test]
111 fn test_intersection() {
112 let left = TokenMultiset::from_token_ids(&[tid(1), tid(1), tid(2), tid(3)]);
113 let right = TokenMultiset::from_token_ids(&[tid(1), tid(2), tid(2), tid(4)]);
114
115 let intersection = left.intersection(&right);
116
117 assert_eq!(intersection.get(&tid(1)), Some(&1));
118 assert_eq!(intersection.get(&tid(2)), Some(&1));
119 assert!(!intersection.contains_key(&tid(3)));
120 assert!(!intersection.contains_key(&tid(4)));
121 assert_eq!(intersection.total_count(), 2);
122 }
123}