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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246
//! Japanese text preprocessor for Text-to-Speech application (OpenJTalk rewrite in rust language).
//!
//! ## Example
//!
//! ```rust
//! # use std::error::Error;
//! # use std::path::PathBuf;
//! use jpreprocess::*;
//!
//! # fn main() -> Result<(), Box<dyn Error>> {
//! # let path = PathBuf::from("tests/min-dict");
//! let config = JPreprocessConfig {
//! dictionary: SystemDictionaryConfig::File(path),
//! user_dictionary: None,
//! };
//! let jpreprocess = JPreprocess::from_config(config)?;
//!
//! let jpcommon_label = jpreprocess
//! .extract_fullcontext("日本語文を解析し、音声合成エンジンに渡せる形式に変換します.")?;
//! assert_eq!(
//! jpcommon_label[2],
//! concat!(
//! "sil^n-i+h=o",
//! "/A:-3+1+7",
//! "/B:xx-xx_xx",
//! "/C:02_xx+xx",
//! "/D:02+xx_xx",
//! "/E:xx_xx!xx_xx-xx",
//! "/F:7_4#0_xx@1_3|1_12",
//! "/G:4_4%0_xx_1",
//! "/H:xx_xx",
//! "/I:3-12@1+2&1-8|1+41",
//! "/J:5_29",
//! "/K:2+8-41"
//! )
//! );
//! #
//! # Ok(())
//! # }
//! ```
mod dictionary;
mod normalize_text;
pub use dictionary::*;
pub use normalize_text::normalize_text_for_naist_jdic;
pub use jpreprocess_core::error;
pub use jpreprocess_njd::NJD;
use jpreprocess_core::*;
use jpreprocess_dictionary::{fetcher::WordDictionaryConfig, DictionaryFetcher, DictionaryStore};
use lindera_core::dictionary::{Dictionary, UserDictionary};
use lindera_dictionary::{load_user_dictionary, UserDictionaryConfig};
use lindera_tokenizer::tokenizer::Tokenizer;
pub struct JPreprocessConfig {
pub dictionary: SystemDictionaryConfig,
pub user_dictionary: Option<UserDictionaryConfig>,
}
pub struct JPreprocess {
tokenizer: Tokenizer,
dictionary_fetcher: Box<dyn DictionaryFetcher>,
}
impl JPreprocess {
/// Loads the dictionary from JPreprocessConfig.
///
/// This supports importing files and built-in dictionary (needs feature).
/// If you need to import from data, please use [`with_dictionaries`] instead.
///
/// [`with_dictionaries`]: #method.with_dictionaries
///
/// ## Example 1: Load from file
///
/// ```rust
/// # use std::error::Error;
/// # use std::path::PathBuf;
/// use jpreprocess::*;
///
/// # fn main() -> Result<(), Box<dyn Error>> {
/// # let path = PathBuf::from("tests/min-dict");
/// let config = JPreprocessConfig {
/// dictionary: SystemDictionaryConfig::File(path),
/// user_dictionary: None,
/// };
/// let jpreprocess = JPreprocess::from_config(config)?;
/// #
/// # Ok(())
/// # }
/// ```
///
/// ## Example 2: Load bundled dictionary (This requires a feature to be enabled)
///
/// ```rust
/// # use std::error::Error;
/// use jpreprocess::{*, kind::*};
///
/// # #[cfg(feature = "naist-jdic")]
/// # fn main() -> Result<(), Box<dyn Error>> {
/// let config = JPreprocessConfig {
/// dictionary: SystemDictionaryConfig::Bundled(JPreprocessDictionaryKind::NaistJdic),
/// user_dictionary: None,
/// };
/// let jpreprocess = JPreprocess::from_config(config)?;
/// #
/// # Ok(())
/// # }
/// # #[cfg(not(feature = "naist-jdic"))]
/// # fn main() {}
/// ```
pub fn from_config(config: JPreprocessConfig) -> JPreprocessResult<Self> {
let dictionary = config.dictionary.load()?;
let user_dictionary = match config.user_dictionary {
Some(user_dict_conf) => Some(load_user_dictionary(user_dict_conf)?),
None => None,
};
Ok(Self::with_dictionaries(dictionary, user_dictionary))
}
/// Creates JPreprocess with provided dictionary data.
pub fn with_dictionaries(
dictionary: Dictionary,
user_dictionary: Option<UserDictionary>,
) -> Self {
let dictionary_fetcher = WordDictionaryConfig {
system: dictionary.serlializer_hint(),
user: user_dictionary
.as_ref()
.map(DictionaryStore::serlializer_hint),
};
Self::with_dictionary_fetcher(dictionary, user_dictionary, Box::new(dictionary_fetcher))
}
/// Creates JPreprocess with provided dictionary fetcher.
///
/// Note: I'm not sure if this is useful for someone. If you need this, please create an issue.
fn with_dictionary_fetcher(
dictionary: Dictionary,
user_dictionary: Option<UserDictionary>,
dictionary_fetcher: Box<dyn DictionaryFetcher>,
) -> Self {
let tokenizer = Tokenizer::new(
dictionary,
user_dictionary,
lindera_core::mode::Mode::Normal,
);
Self {
tokenizer,
dictionary_fetcher,
}
}
/// Alias of [`with_dictionaries`].
///
/// Note: `new` before v0.2.0 has moved to `from_config`.
///
/// [`with_dictionaries`]: #method.with_dictionaries
#[deprecated(since = "0.5.0", note = "please use `with_dictionaries` instead")]
pub fn new(dictionary: Dictionary, user_dictionary: Option<UserDictionary>) -> Self {
Self::with_dictionaries(dictionary, user_dictionary)
}
/// Tokenize input text and return NJD.
///
/// Useful for customizing text processing.
///
/// ```rust
/// # use std::error::Error;
/// # use std::path::PathBuf;
/// use jpreprocess::*;
/// use jpreprocess_jpcommon::*;
///
/// # fn main() -> Result<(), Box<dyn Error>> {
/// # let path = PathBuf::from("tests/min-dict");
/// # let config = JPreprocessConfig {
/// # dictionary: SystemDictionaryConfig::File(path),
/// # user_dictionary: None,
/// # };
/// let jpreprocess = JPreprocess::from_config(config)?;
///
/// let mut njd = jpreprocess.text_to_njd("日本語文を解析し、音声合成エンジンに渡せる形式に変換します.")?;
/// njd.preprocess();
///
/// // jpcommon utterance
/// let utterance = Utterance::from(njd.nodes.as_slice());
///
/// // Vec<([phoneme string], [context labels])>
/// let phoneme_vec = utterance_to_phoneme_vec(&utterance);
///
/// assert_eq!(&phoneme_vec[2].0, "i");
/// assert!(phoneme_vec[2].1.starts_with("/A:-3+1+7"));
///
/// // fullcontext label
/// let fullcontext = overwrapping_phonemes(phoneme_vec);
///
/// assert!(fullcontext[2].starts_with("sil^n-i+h=o"));
/// #
/// # Ok(())
/// # }
/// ```
pub fn text_to_njd(&self, text: &str) -> JPreprocessResult<NJD> {
let normalized_input_text = normalize_text_for_naist_jdic(text);
let tokens = self.tokenizer.tokenize(normalized_input_text.as_str())?;
NJD::from_tokens(&tokens, &self.dictionary_fetcher)
}
/// Tokenize a text, preprocess, and return NJD converted to string.
///
/// The returned string does not match that of openjtalk.
/// JPreprocess drops orig string and some of the CForm information,
/// which is unnecessary to preprocessing.
///
/// If you need these infomation, please raise a feature request as an issue.
pub fn run_frontend(&self, text: &str) -> JPreprocessResult<Vec<String>> {
let mut njd = Self::text_to_njd(self, text)?;
njd.preprocess();
Ok(njd.into())
}
/// Generate jpcommon features from NJD features(returned by [`run_frontend`]).
///
/// [`run_frontend`]: #method.run_frontend
pub fn make_label(&self, njd_features: Vec<String>) -> Vec<String> {
let njd = NJD::from_strings(njd_features);
jpreprocess_jpcommon::njdnodes_to_features(&njd.nodes)
}
/// Generate jpcommon features from a text.
///
/// This is not guaranteed to be same as calling [`run_frontend`] and [`make_label`].
///
/// [`run_frontend`]: #method.run_frontend
/// [`make_label`]: #method.make_label
pub fn extract_fullcontext(&self, text: &str) -> JPreprocessResult<Vec<String>> {
let mut njd = Self::text_to_njd(self, text)?;
njd.preprocess();
Ok(jpreprocess_jpcommon::njdnodes_to_features(&njd.nodes))
}
}