detexify 0.4.0

The `detexify` crate is a port of the classifier that powers https://detexify.kirelabs.org/classify.html.
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
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115

use crate::stroke_sample::StrokeSample;
use itertools::Itertools;
use serde::Serialize;
use std::collections::HashMap;

pub(crate) trait Sample<T> {
    fn distance(a: T, b: T) -> f64;
}

#[derive(Debug)]
struct Hit<T> {
    sample_score: f64,
    sample: T,
}

#[derive(Debug, Serialize)]
pub struct Score {
    pub id: String,
    pub score: f64,
}

pub struct Classifier {
    samples: HashMap<String, Vec<StrokeSample>>,
}

impl Classifier {
    pub fn new() -> Classifier {
        Classifier {
            samples: HashMap::new(),
        }
    }

    pub fn from_snapshot<R: std::io::Read>(reader: R) -> serde_json::Result<Classifier> {
        let samples = serde_json::from_reader(reader);
        samples.map(|s| Classifier { samples: s })
    }

    pub fn classify(&self, unknown: StrokeSample) -> Option<Vec<Score>> {
        if unknown.is_empty() {
            return None;
        }

        Some(
            self.samples
                .iter()
                .map(|(id, samples)| {
                    let mean_dist = samples
                        .iter()
                        .cloned()
                        .map(|s| StrokeSample::distance(unknown.clone(), s))
                        .sorted_by(|x, y| x.partial_cmp(y).unwrap())
                        .take(2)
                        .fold(0.0, |acc, x| acc + x)
                        / 2.0;

                    (id, mean_dist)
                })
                .map(|(id, dist)| Score {
                    id: id.clone(),
                    score: dist,
                })
                .sorted_by(|x, y| x.score.partial_cmp(&y.score).unwrap())
                .collect(),
        )
    }
}

impl Default for Classifier {
    fn default() -> Self {
        let samples: HashMap<String, Vec<StrokeSample>> =
            serde_json::from_slice::<HashMap<String, Vec<StrokeSample>>>(include_bytes!(
                "../snapshot.json"
            ))
            .unwrap()
            .into_iter()
            .map(|(id_base64, strokes)| {
                let id = base64::decode(id_base64).unwrap();
                let id_base32 = base32::encode(base32::Alphabet::RFC4648 { padding: false }, &id);

                return (id_base32, strokes);
            })
            .collect();

        Classifier { samples }
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::stroke::Stroke;

    #[test]
    fn default_classifier() {
        Classifier::default();
    }
}

// fn insert_with_limit<T: Sample>(
//     limit: i64,
//     identifier: String,
//     sample: T,
//     map: HashMap<String, Vec>,
// ) -> HashMap<String, Vec> {
//     todo!()
// }

// fn train_classifier<T: Sample>(classifier: Classifier, identifier: String, sample: T) {
//     insert_with_limit(
//         classifier.sample_limit,
//         identifier,
//         sample,
//         classifier.samples,
//     );
// }