text_tokenizer/
options.rs

1use crate::{SentenceBreaker, UnicodeSentenceBreaker};
2
3use std::collections::BTreeSet;
4
5pub trait IntoTokenizer: Sized {
6    type IntoTokens;
7    fn into_tokenizer<S: SentenceBreaker>(self, params: TokenizerParams<S>) -> Self::IntoTokens;
8}
9
10#[derive(Debug, Copy, Clone, PartialEq, Eq, PartialOrd, Ord)]
11pub enum TokenizerOptions {
12    NoComplexTokens,
13    StructTokens,
14    SplitDot,
15    SplitUnderscore,
16    SplitColon,
17    SplitComa,
18    SplitSemiColon,
19    SplitNumberSign,
20    MergePunctuation,
21    MergeWhites,
22    WithSentences,
23}
24
25pub struct TokenizerParams<S: SentenceBreaker> {
26    pub(crate) options: BTreeSet<TokenizerOptions>,
27    pub(crate) sentence_breaker: S,
28}
29impl Default for TokenizerParams<()> {
30    fn default() -> TokenizerParams<()> {
31        TokenizerParams {
32            options: BTreeSet::new(),
33            sentence_breaker: (),
34        }
35    }
36}
37impl TokenizerParams<()> {
38    pub fn v1() -> TokenizerParams<()> {
39        TokenizerParams::default()
40            .add_option(TokenizerOptions::SplitDot)
41            .add_option(TokenizerOptions::SplitUnderscore)
42            .add_option(TokenizerOptions::SplitColon)
43            .add_option(TokenizerOptions::SplitSemiColon)
44            .add_option(TokenizerOptions::MergeWhites)
45            .add_option(TokenizerOptions::MergePunctuation)
46    }
47    pub fn v1_1() -> TokenizerParams<()> {
48        TokenizerParams::default()
49            .add_option(TokenizerOptions::SplitDot)
50            .add_option(TokenizerOptions::SplitUnderscore)
51            .add_option(TokenizerOptions::SplitColon)
52            .add_option(TokenizerOptions::SplitComa)
53            .add_option(TokenizerOptions::SplitSemiColon)
54            .add_option(TokenizerOptions::MergeWhites)
55            .add_option(TokenizerOptions::MergePunctuation)
56    }
57    pub fn basic() -> TokenizerParams<()> {
58        TokenizerParams::default()
59            .add_option(TokenizerOptions::NoComplexTokens)
60            .add_option(TokenizerOptions::MergeWhites)
61            .add_option(TokenizerOptions::MergePunctuation)
62    }
63    pub fn complex() -> TokenizerParams<()> {
64        TokenizerParams::default()
65            .add_option(TokenizerOptions::StructTokens)
66            .add_option(TokenizerOptions::MergeWhites)
67            .add_option(TokenizerOptions::MergePunctuation)
68    }
69}
70impl<S: SentenceBreaker> TokenizerParams<S> {
71    pub fn add_option(mut self, option: TokenizerOptions) -> TokenizerParams<S> {
72        self.options.insert(option);
73        self
74    }
75    pub fn with_default_sentences(mut self) -> TokenizerParams<UnicodeSentenceBreaker> {
76        self.options.insert(TokenizerOptions::WithSentences);
77        TokenizerParams {
78            options: self.options,
79            sentence_breaker: UnicodeSentenceBreaker,
80        }
81    }
82    pub fn with_sentence_breaker<U: SentenceBreaker>(mut self, sb: U) -> TokenizerParams<U> {
83        self.options.insert(TokenizerOptions::WithSentences);
84        TokenizerParams {
85            options: self.options,
86            sentence_breaker: sb,
87        }
88    }
89}
text_tokenizer/options.rs

text_tokenizer/
options.rs
