cinekey 0.4.5

Fast parser for movie and TV release names
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
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141

#![forbid(unsafe_code)]

/// The delimiter immediately before a token.
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
pub enum Separator {
    Start,
    Dot,
    Space,
    Dash,
    Underscore,
    Plus,
    Bracket,
    Other,
}

/// A token borrows the original input; offsets are UTF-8 byte offsets.
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
pub struct Token<'a> {
    pub text: &'a str,
    pub start: usize,
    pub end: usize,
    pub before: Separator,
    pub bracketed: bool,
}

#[derive(Clone, Copy, Debug, Eq, PartialEq)]
pub(crate) struct ReleaseToken<'a> {
    pub text: &'a str,
    pub start: usize,
    pub end: usize,
    pub before: Separator,
    pub bracketed: bool,
    pub word: Option<crate::vocabulary::Word>,
}

impl ReleaseToken<'_> {
    pub(crate) fn is_technical(self) -> bool {
        self.word.is_some_and(crate::vocabulary::Word::boundary)
    }
}

impl<'a> From<Token<'a>> for ReleaseToken<'a> {
    fn from(token: Token<'a>) -> Self {
        Self {
            text: token.text,
            start: token.start,
            end: token.end,
            before: token.before,
            bracketed: token.bracketed,
            word: crate::vocabulary::classify(token.text),
        }
    }
}

const _: () = assert!(size_of::<ReleaseToken<'_>>() == size_of::<Token<'_>>());

pub(crate) fn release_tokens(input: &str) -> Vec<ReleaseToken<'_>> {
    tokenize_with(input, ReleaseToken::from)
}

/// Split a release name into Unicode-aware word tokens that borrow the input.
/// Punctuation remains represented by each token's preceding separator and span.
pub fn tokenize(input: &str) -> Vec<Token<'_>> {
    tokenize_with(input, |token| token)
}

fn tokenize_with<'a, T>(input: &'a str, mut make: impl FnMut(Token<'a>) -> T) -> Vec<T> {
    let mut tokens = Vec::with_capacity((input.len() / 6).min(32));
    let mut token_start = None;
    let mut before = Separator::Start;
    let mut bracket_depth = 0usize;

    for (offset, ch) in input.char_indices() {
        if ch.is_alphanumeric() {
            token_start.get_or_insert(offset);
            continue;
        }

        if let Some(start) = token_start.take() {
            tokens.push(make(Token {
                text: &input[start..offset],
                start,
                end: offset,
                before,
                bracketed: bracket_depth > 0,
            }));
        }

        before = match ch {
            '.' => Separator::Dot,
            ' ' | '\t' | '\n' | '\r' => Separator::Space,
            '-' | '‐' | '‑' | '–' | '—' => Separator::Dash,
            '_' => Separator::Underscore,
            '+' => Separator::Plus,
            '[' | '(' | '{' => {
                bracket_depth = bracket_depth.saturating_add(1);
                Separator::Bracket
            }
            ']' | ')' | '}' => {
                bracket_depth = bracket_depth.saturating_sub(1);
                Separator::Bracket
            }
            _ => Separator::Other,
        };
    }

    if let Some(start) = token_start {
        tokens.push(make(Token {
            text: &input[start..],
            start,
            end: input.len(),
            before,
            bracketed: bracket_depth > 0,
        }));
    }
    tokens
}

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

    #[test]
    fn preserves_unicode_byte_spans_and_delimiters() {
        let input = "[字幕组] Made.in.Abyss.S02E03.mkv";
        let tokens = tokenize(input);
        assert_eq!(tokens[0].text, "字幕组");
        assert_eq!(&input[tokens[0].start..tokens[0].end], "字幕组");
        assert!(tokens[0].bracketed);
        assert_eq!(tokens[1].text, "Made");
        assert_eq!(tokens[1].before, Separator::Space);
        assert_eq!(tokens[3].before, Separator::Dot);
    }

    #[test]
    fn tokenizes_cjk_and_ascii_markers() {
        let tokens = tokenize("沉默的荣耀.2025.S01E02");
        assert_eq!(tokens[0].text, "沉默的荣耀");
        assert_eq!(tokens[1].text, "2025");
    }
}