sudachi/dic/
connect.rs

1/*
2 *  Copyright (c) 2021-2026 Works Applications Co., Ltd.
3 *
4 *  Licensed under the Apache License, Version 2.0 (the "License");
5 *  you may not use this file except in compliance with the License.
6 *  You may obtain a copy of the License at
7 *
8 *      http://www.apache.org/licenses/LICENSE-2.0
9 *
10 *   Unless required by applicable law or agreed to in writing, software
11 *  distributed under the License is distributed on an "AS IS" BASIS,
12 *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 *  See the License for the specific language governing permissions and
14 *  limitations under the License.
15 */
16
17use nom::number::complete::le_i16;
18
19use crate::error::{SudachiError, SudachiResult};
20use crate::util::cow_array::CowArray;
21
22pub struct ConnectionMatrix<'a> {
23    data: CowArray<'a, i16>,
24    num_left: usize,
25    num_right: usize,
26}
27
28impl<'a> ConnectionMatrix<'a> {
29    pub fn from_bytes(buf: &'a [u8]) -> SudachiResult<ConnectionMatrix<'a>> {
30        let (rest, (num_left, num_right)) = nom::sequence::tuple((le_i16, le_i16))(buf)?;
31        Self::from_offset_size(rest, 0, num_left as usize, num_right as usize)
32    }
33
34    pub fn from_offset_size(
35        data: &'a [u8],
36        offset: usize,
37        num_left: usize,
38        num_right: usize,
39    ) -> SudachiResult<ConnectionMatrix<'a>> {
40        let size = num_left * num_right;
41        let end = offset + size * std::mem::size_of::<i16>();
42        if end > data.len() {
43            return Err(SudachiError::InvalidDictionaryGrammar.with_context("connection matrix"));
44        }
45
46        Ok(ConnectionMatrix {
47            data: CowArray::from_bytes(data, offset, size),
48            num_left,
49            num_right,
50        })
51    }
52
53    #[inline(always)]
54    fn index(&self, left: u16, right: u16) -> usize {
55        let uleft = left as usize;
56        let uright = right as usize;
57        debug_assert!(
58            uleft < self.num_left,
59            "left id {} is out of range (num_left={})",
60            uleft,
61            self.num_left
62        );
63        debug_assert!(
64            uright < self.num_right,
65            "right id {} is out of range (num_right={})",
66            uright,
67            self.num_right
68        );
69        let index = uright * self.num_left + uleft;
70        debug_assert!(index < self.data.len());
71        index
72    }
73
74    /// Gets the value of the connection matrix
75    ///
76    /// It is performance critical that this function
77    /// 1. Has no branches
78    /// 2. Is inlined to the caller
79    ///
80    /// This is UB if index is out of bounds, but that can't happen
81    /// except in the case if the binary dictionary was tampered with.
82    /// It is OK to make usage of tampered binary dictionaries UB.
83    #[inline(always)]
84    pub fn cost(&self, left: u16, right: u16) -> i16 {
85        let index = self.index(left, right);
86        *unsafe { self.data.get_unchecked(index) }
87    }
88
89    pub fn update(&mut self, left: u16, right: u16, value: i16) {
90        let index = self.index(left, right);
91        self.data.set(index, value);
92    }
93
94    /// Returns maximum number of left connection ID
95    pub fn num_left(&self) -> usize {
96        self.num_left
97    }
98
99    /// Returns maximum number of right connection ID
100    pub fn num_right(&self) -> usize {
101        self.num_right
102    }
103}
104
105impl ConnectionMatrix<'static> {
106    pub fn empty() -> Self {
107        ConnectionMatrix {
108            data: CowArray::from_owned(Vec::new()),
109            num_left: 0,
110            num_right: 0,
111        }
112    }
113}
sudachi/dic/connect.rs

sudachi/dic/
connect.rs
