lingua_rs 0.1.0 → 0.3.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
checksums.yaml CHANGED
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  SHA256:
3
- metadata.gz: 95237de3af0469a6394db02288b9230472c763dc35cfc08768fc7a0233de6d18
4
- data.tar.gz: fe793fbe6f8fa05d408b43579738b1e386db9fc9f93a590700f4726f617f7fa2
3
+ metadata.gz: e301297292f2de86fd62416f6f06765c7ccd373c81020e2ad365e0c0151a55bf
4
+ data.tar.gz: fe961910a3e44fa7df10580100b4aecef6f2552c08a7faba278a84eaf7ae89fc
5
5
  SHA512:
6
- metadata.gz: 9a0fe59c439a1b27c7f0f95b9a3eba1a9c4dabe441ab5a07c0864727d9579b307e0e0c345ff0285b9c11854c7188c3c0726480103ff0390006ef799c3ee5c1bd
7
- data.tar.gz: 82c83d17a21affafcc79cf0999658f33c225b5a7f9d532372f00dde0834b5b663cb0a572630042d5d3e725ae05b25189770993a66eebcf257c2b4eec6b5ca5e2
6
+ metadata.gz: 31e45f89765b5feb782a8b47f9574891c9ce826b5883f840b5f0d06c8715f26c9d9caa999a8e08349f475e307b2ccd28ecb303677ae8e51caa79920d94cc91fc
7
+ data.tar.gz: 6ace3072e3b7e8e044c55712879ea398dc7f009475cb058444070cd22834bb7f4fa6525ab4a66bab95417d57b1f57b9599cda02dd159be80603280c0646b6414
data/README.md CHANGED
@@ -1,5 +1,10 @@
1
1
  # LinguaRs
2
2
 
3
+ [![CI](https://github.com/kochka/lingua_rs/actions/workflows/ci.yml/badge.svg)](https://github.com/kochka/lingua_rs/actions/workflows/ci.yml)
4
+ [![Gem Version](https://badge.fury.io/rb/lingua_rs.svg)](https://badge.fury.io/rb/lingua_rs)
5
+ [![Downloads](https://img.shields.io/gem/dt/lingua_rs)](https://rubygems.org/gems/lingua_rs)
6
+ [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/licenses/MIT)
7
+
3
8
  A Ruby gem wrapping the [Lingua](https://github.com/pemistahl/lingua-rs) Rust library for language detection.
4
9
 
5
10
  ## Installation
@@ -104,7 +109,7 @@ detector.detect('Hello world') # => #<Lingua::Language English>
104
109
 
105
110
  ```ruby
106
111
  detector.confidence('Bonjour le monde', :fr) # => 0.8217
107
- Lingua.confidence('Bonjour le monde', 'fr') # => 0.3794
112
+ Lingua.confidence('Bonjour le monde', 'fr') # => 0.8217
108
113
 
109
114
  results = detector.confidence_values('Bonjour le monde')
110
115
  results.first.language # => #<Lingua::Language French>
@@ -113,13 +118,41 @@ results.first.to_s # => "French (0.82)"
113
118
  results.sum(&:confidence) # => 1.0
114
119
  ```
115
120
 
121
+ ### Mixed-language detection
122
+
123
+ `detect_multiple` identifies multiple languages within a single text and returns an array of `Lingua::Segment` objects. Available on both `Lingua::Detector` and as a module method on `Lingua`.
124
+
125
+ ```ruby
126
+ text = "Parlez-vous français? Ich spreche Französisch nur ein bisschen. A little bit is better than nothing."
127
+
128
+ segments = Lingua.detect_multiple(text, languages: %w[en fr de])
129
+ segments.each do |s|
130
+ puts "#{s.language} (#{s.start_index}..#{s.end_index}): #{s.text}"
131
+ end
132
+ # French (0..22): Parlez-vous français?
133
+ # German (23..64): Ich spreche Französisch nur ein bisschen.
134
+ # English (65..101): A little bit is better than nothing.
135
+
136
+ # With a persistent detector
137
+ detector = Lingua::Detector.new(languages: %w[en fr de])
138
+ detector.detect_multiple(text)
139
+ ```
140
+
116
141
  ### `Lingua::Language` methods
117
142
 
118
- `Lingua::Language` objects support equality (`==`) and can be used as Hash keys.
143
+ `Lingua::Language` objects support equality (`==`) and can be used as Hash keys. You can look up a language by name, ISO 639-1 code, or ISO 639-3 code using `[]`:
144
+
145
+ ```ruby
146
+ Lingua::Language['French'] # => #<Lingua::Language French>
147
+ Lingua::Language[:fr] # => #<Lingua::Language French>
148
+ Lingua::Language['fra'] # => #<Lingua::Language French>
149
+ Lingua::Language['xxx'] # => nil
150
+ ```
119
151
 
120
152
  | Method | Return type | Example |
121
153
  |---|---|---|
122
- | `to_s` | `String` | `'French'` |
154
+ | `name` | `String` | `'French'` |
155
+ | `to_s` | `String` | `'French'` (alias for `name`) |
123
156
  | `to_sym` | `Symbol` | `:french` |
124
157
  | `to_iso` | `String` | `'fr'` (alias for `to_iso6391`) |
125
158
  | `to_iso6391` | `String` | `'fr'` |
@@ -128,6 +161,14 @@ results.sum(&:confidence) # => 1.0
128
161
  | `==` | `Boolean` | Compare two languages |
129
162
  | `hash` | `Integer` | Hash value (usable as Hash key) |
130
163
 
164
+ Class methods:
165
+
166
+ | Method | Return type | Description |
167
+ |---|---|---|
168
+ | `Lingua::Language.all` | `Array<Lingua::Language>` | All supported languages |
169
+ | `Lingua::Language.names` | `Array<String>` | All language names (e.g. `'French'`) |
170
+ | `Lingua::Language.iso_codes` | `Array<String>` | All ISO 639-1 codes (e.g. `'fr'`) |
171
+
131
172
  ### `Lingua::ConfidenceResult` methods
132
173
 
133
174
  Returned by `confidence_values`.
@@ -139,10 +180,28 @@ Returned by `confidence_values`.
139
180
  | `to_s` | `String` | `'French (0.82)'` |
140
181
  | `inspect` | `String` | `'#<Lingua::ConfidenceResult French (0.8217)>'` |
141
182
 
183
+ ### `Lingua::Segment` methods
184
+
185
+ Returned by `detect_multiple`.
186
+
187
+ | Method | Return type | Example |
188
+ |---|---|---|
189
+ | `language` | `Lingua::Language` | `#<Lingua::Language French>` |
190
+ | `start_index` | `Integer` | `0` |
191
+ | `end_index` | `Integer` | `22` |
192
+ | `word_count` | `Integer` | `3` |
193
+ | `text` | `String` | `'Parlez-vous français? '` |
194
+ | `to_s` | `String` | `'French (0-22): Parlez-vous français? '` |
195
+ | `inspect` | `String` | `'#<Lingua::Segment French (0-22) "Parlez-vous français? ">'` |
196
+
142
197
  ## Development
143
198
 
144
199
  After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake compile` to build the native extension and `rake test` to run the tests.
145
200
 
201
+ ## Acknowledgements
202
+
203
+ This gem is built on top of [Lingua](https://github.com/pemistahl/lingua-rs) by [Peter M. Stahl](https://github.com/pemistahl), a highly accurate natural language detection library written in Rust.
204
+
146
205
  ## License
147
206
 
148
207
  The gem is available as open source under the terms of the [MIT License](https://opensource.org/licenses/MIT).
@@ -1,7 +1,18 @@
1
1
  use lingua::Language;
2
+ use magnus::{Error, RModule, Ruby, method, prelude::*};
2
3
 
3
4
  use crate::language::WrappedLanguage;
4
5
 
6
+ pub fn define(ruby: &Ruby, module: &RModule) -> Result<(), Error> {
7
+ let class = module.define_class("ConfidenceResult", ruby.class_object())?;
8
+ class.undef_default_alloc_func();
9
+ class.define_method("language", method!(ConfidenceResult::language, 0))?;
10
+ class.define_method("confidence", method!(ConfidenceResult::confidence, 0))?;
11
+ class.define_method("to_s", method!(ConfidenceResult::to_s, 0))?;
12
+ class.define_method("inspect", method!(ConfidenceResult::inspect, 0))?;
13
+ Ok(())
14
+ }
15
+
5
16
  #[magnus::wrap(class = "Lingua::ConfidenceResult")]
6
17
  pub struct ConfidenceResult {
7
18
  pub language: Language,
@@ -1,10 +1,21 @@
1
1
  use lingua::{LanguageDetector, LanguageDetectorBuilder};
2
- use magnus::{Error, RArray, RHash, Ruby};
2
+ use magnus::{Error, RArray, RHash, RModule, Ruby, function, method, prelude::*};
3
3
 
4
4
  use crate::confidence_result::ConfidenceResult;
5
+ use crate::segment::Segment;
5
6
  use crate::helpers::{fetch_option, parse_language, value_to_string};
6
7
  use crate::language::WrappedLanguage;
7
8
 
9
+ pub fn define(ruby: &Ruby, module: &RModule) -> Result<(), Error> {
10
+ let class = module.define_class("Detector", ruby.class_object())?;
11
+ class.define_singleton_method("new", function!(RubyDetector::new, -1))?;
12
+ class.define_method("detect", method!(RubyDetector::detect, 1))?;
13
+ class.define_method("confidence", method!(RubyDetector::confidence, 2))?;
14
+ class.define_method("confidence_values", method!(RubyDetector::confidence_values, 1))?;
15
+ class.define_method("detect_multiple", method!(RubyDetector::detect_multiple, 1))?;
16
+ Ok(())
17
+ }
18
+
8
19
  pub fn compute_confidence(
9
20
  detector: &LanguageDetector,
10
21
  subject: String,
@@ -66,6 +77,32 @@ impl RubyDetector {
66
77
  pub fn confidence_values(&self, subject: String) -> Result<RArray, Error> {
67
78
  compute_confidence_values(&self.detector, subject)
68
79
  }
80
+
81
+ pub fn detect_multiple(&self, subject: String) -> Result<RArray, Error> {
82
+ compute_detect_multiple(&self.detector, &subject)
83
+ }
84
+ }
85
+
86
+ pub fn compute_detect_multiple(
87
+ detector: &LanguageDetector,
88
+ subject: &str,
89
+ ) -> Result<RArray, Error> {
90
+ let ruby = Ruby::get().unwrap();
91
+ let results = detector.detect_multiple_languages_of(subject);
92
+ let array = ruby.ary_new_capa(results.len());
93
+ for r in results {
94
+ let text = subject[r.start_index()..r.end_index()].to_string();
95
+ let start_index = subject[..r.start_index()].chars().count();
96
+ let end_index = start_index + text.chars().count();
97
+ array.push(Segment {
98
+ language: r.language(),
99
+ start_index,
100
+ end_index,
101
+ word_count: r.word_count(),
102
+ text,
103
+ })?;
104
+ }
105
+ Ok(array)
69
106
  }
70
107
 
71
108
  pub fn build_detector_from_options(
@@ -1,11 +1,71 @@
1
1
  use lingua::Language;
2
- use magnus::{Ruby, Symbol};
2
+ use magnus::{Error, RArray, RModule, Ruby, Symbol, function, method, prelude::*};
3
+
4
+ use crate::helpers::value_to_string;
5
+
6
+ pub fn define(ruby: &Ruby, module: &RModule) -> Result<(), Error> {
7
+ let class = module.define_class("Language", ruby.class_object())?;
8
+ class.undef_default_alloc_func();
9
+ class.define_method("name", method!(WrappedLanguage::name, 0))?;
10
+ class.define_method("to_s", method!(WrappedLanguage::name, 0))?;
11
+ class.define_method("to_iso6391", method!(WrappedLanguage::to_iso6391, 0))?;
12
+ class.define_method("to_iso", method!(WrappedLanguage::to_iso6391, 0))?;
13
+ class.define_method("to_iso6393", method!(WrappedLanguage::to_iso6393, 0))?;
14
+ class.define_method("to_sym", method!(WrappedLanguage::to_sym, 0))?;
15
+ class.define_method("inspect", method!(WrappedLanguage::inspect, 0))?;
16
+ class.define_method("==", method!(WrappedLanguage::eq, 1))?;
17
+ class.define_method("eql?", method!(WrappedLanguage::eq, 1))?;
18
+ class.define_method("hash", method!(WrappedLanguage::hash, 0))?;
19
+ class.define_singleton_method("[]", function!(WrappedLanguage::lookup, 1))?;
20
+ class.define_singleton_method("all", function!(WrappedLanguage::all, 0))?;
21
+ class.define_singleton_method("names", function!(WrappedLanguage::names, 0))?;
22
+ class.define_singleton_method("iso_codes", function!(WrappedLanguage::iso_codes, 0))?;
23
+ Ok(())
24
+ }
3
25
 
4
26
  #[magnus::wrap(class = "Lingua::Language")]
5
27
  pub struct WrappedLanguage(pub Language);
6
28
 
7
29
  impl WrappedLanguage {
8
- pub fn to_s(&self) -> String {
30
+ pub fn lookup(value: magnus::Value) -> Result<Option<WrappedLanguage>, Error> {
31
+ let input = value_to_string(value)?;
32
+ Ok(crate::helpers::parse_language(&input).map(WrappedLanguage))
33
+ }
34
+
35
+ pub fn all() -> Result<RArray, Error> {
36
+ let ruby = Ruby::get().unwrap();
37
+ let mut langs: Vec<Language> = Language::all().into_iter().collect();
38
+ langs.sort_by(|a, b| a.to_string().cmp(&b.to_string()));
39
+ let array = ruby.ary_new_capa(langs.len());
40
+ for lang in langs {
41
+ array.push(WrappedLanguage(lang))?;
42
+ }
43
+ Ok(array)
44
+ }
45
+
46
+ pub fn names() -> Result<RArray, Error> {
47
+ let ruby = Ruby::get().unwrap();
48
+ let mut langs: Vec<Language> = Language::all().into_iter().collect();
49
+ langs.sort_by(|a, b| a.to_string().cmp(&b.to_string()));
50
+ let array = ruby.ary_new_capa(langs.len());
51
+ for lang in langs {
52
+ array.push(lang.to_string())?;
53
+ }
54
+ Ok(array)
55
+ }
56
+
57
+ pub fn iso_codes() -> Result<RArray, Error> {
58
+ let ruby = Ruby::get().unwrap();
59
+ let mut langs: Vec<Language> = Language::all().into_iter().collect();
60
+ langs.sort_by(|a, b| a.to_string().cmp(&b.to_string()));
61
+ let array = ruby.ary_new_capa(langs.len());
62
+ for lang in langs {
63
+ array.push(lang.iso_code_639_1().to_string())?;
64
+ }
65
+ Ok(array)
66
+ }
67
+
68
+ pub fn name(&self) -> String {
9
69
  self.0.to_string()
10
70
  }
11
71
 
@@ -2,11 +2,11 @@ mod confidence_result;
2
2
  mod detector;
3
3
  mod helpers;
4
4
  mod language;
5
+ mod segment;
5
6
 
6
- use magnus::{Error, RArray, RHash, Ruby, function, method, prelude::*};
7
+ use magnus::{Error, RArray, RHash, Ruby, function, prelude::*};
7
8
 
8
- use confidence_result::ConfidenceResult;
9
- use detector::{RubyDetector, build_detector_from_options, compute_confidence, compute_confidence_values};
9
+ use detector::{build_detector_from_options, compute_confidence, compute_confidence_values, compute_detect_multiple};
10
10
  use language::WrappedLanguage;
11
11
 
12
12
  fn detect(ruby: &Ruby, arguments: RArray) -> Result<Option<WrappedLanguage>, Error> {
@@ -32,42 +32,28 @@ fn confidence_values(ruby: &Ruby, arguments: RArray) -> Result<RArray, Error> {
32
32
  compute_confidence_values(&detector, subject)
33
33
  }
34
34
 
35
+ fn detect_multiple(ruby: &Ruby, arguments: RArray) -> Result<RArray, Error> {
36
+ let subject = arguments
37
+ .shift::<String>()
38
+ .map_err(|_| Error::new(ruby.exception_arg_error(), "expected a string as first argument"))?;
39
+ let options = arguments.shift::<RHash>().ok();
40
+ let detector = build_detector_from_options(ruby, options.as_ref())?;
41
+ compute_detect_multiple(&detector, &subject)
42
+ }
43
+
35
44
  #[magnus::init]
36
45
  fn init(ruby: &Ruby) -> Result<(), Error> {
37
46
  let module = ruby.define_module("Lingua")?;
38
47
 
39
- // Lingua::Language
40
- let language_class = module.define_class("Language", ruby.class_object())?;
41
- language_class.undef_default_alloc_func();
42
- language_class.define_method("to_s", method!(WrappedLanguage::to_s, 0))?;
43
- language_class.define_method("to_iso6391", method!(WrappedLanguage::to_iso6391, 0))?;
44
- language_class.define_method("to_iso", method!(WrappedLanguage::to_iso6391, 0))?;
45
- language_class.define_method("to_iso6393", method!(WrappedLanguage::to_iso6393, 0))?;
46
- language_class.define_method("to_sym", method!(WrappedLanguage::to_sym, 0))?;
47
- language_class.define_method("inspect", method!(WrappedLanguage::inspect, 0))?;
48
- language_class.define_method("==", method!(WrappedLanguage::eq, 1))?;
49
- language_class.define_method("eql?", method!(WrappedLanguage::eq, 1))?;
50
- language_class.define_method("hash", method!(WrappedLanguage::hash, 0))?;
51
-
52
- // Lingua::ConfidenceResult
53
- let confidence_class = module.define_class("ConfidenceResult", ruby.class_object())?;
54
- confidence_class.undef_default_alloc_func();
55
- confidence_class.define_method("language", method!(ConfidenceResult::language, 0))?;
56
- confidence_class.define_method("confidence", method!(ConfidenceResult::confidence, 0))?;
57
- confidence_class.define_method("to_s", method!(ConfidenceResult::to_s, 0))?;
58
- confidence_class.define_method("inspect", method!(ConfidenceResult::inspect, 0))?;
59
-
60
- // Lingua::Detector
61
- let detector_class = module.define_class("Detector", ruby.class_object())?;
62
- detector_class.define_singleton_method("new", function!(RubyDetector::new, -1))?;
63
- detector_class.define_method("detect", method!(RubyDetector::detect, 1))?;
64
- detector_class.define_method("confidence", method!(RubyDetector::confidence, 2))?;
65
- detector_class.define_method("confidence_values", method!(RubyDetector::confidence_values, 1))?;
48
+ language::define(ruby, &module)?;
49
+ confidence_result::define(ruby, &module)?;
50
+ segment::define(ruby, &module)?;
51
+ detector::define(ruby, &module)?;
66
52
 
67
- // Functional API (module methods)
68
53
  module.define_singleton_method("detect", function!(detect, -2))?;
69
54
  module.define_singleton_method("confidence", function!(confidence, 2))?;
70
55
  module.define_singleton_method("confidence_values", function!(confidence_values, -2))?;
56
+ module.define_singleton_method("detect_multiple", function!(detect_multiple, -2))?;
71
57
 
72
58
  Ok(())
73
59
  }
@@ -0,0 +1,59 @@
1
+ use lingua::Language;
2
+ use magnus::{Error, RModule, Ruby, method, prelude::*};
3
+
4
+ use crate::language::WrappedLanguage;
5
+
6
+ pub fn define(ruby: &Ruby, module: &RModule) -> Result<(), Error> {
7
+ let class = module.define_class("Segment", ruby.class_object())?;
8
+ class.undef_default_alloc_func();
9
+ class.define_method("language", method!(Segment::language, 0))?;
10
+ class.define_method("start_index", method!(Segment::start_index, 0))?;
11
+ class.define_method("end_index", method!(Segment::end_index, 0))?;
12
+ class.define_method("word_count", method!(Segment::word_count, 0))?;
13
+ class.define_method("text", method!(Segment::text, 0))?;
14
+ class.define_method("to_s", method!(Segment::to_s, 0))?;
15
+ class.define_method("inspect", method!(Segment::inspect, 0))?;
16
+ Ok(())
17
+ }
18
+
19
+ #[magnus::wrap(class = "Lingua::Segment")]
20
+ pub struct Segment {
21
+ pub language: Language,
22
+ pub start_index: usize,
23
+ pub end_index: usize,
24
+ pub word_count: usize,
25
+ pub text: String,
26
+ }
27
+
28
+ impl Segment {
29
+ pub fn language(&self) -> WrappedLanguage {
30
+ WrappedLanguage(self.language)
31
+ }
32
+
33
+ pub fn start_index(&self) -> usize {
34
+ self.start_index
35
+ }
36
+
37
+ pub fn end_index(&self) -> usize {
38
+ self.end_index
39
+ }
40
+
41
+ pub fn word_count(&self) -> usize {
42
+ self.word_count
43
+ }
44
+
45
+ pub fn text(&self) -> String {
46
+ self.text.clone()
47
+ }
48
+
49
+ pub fn to_s(&self) -> String {
50
+ format!("{} ({}-{}): {}", self.language, self.start_index, self.end_index, self.text)
51
+ }
52
+
53
+ pub fn inspect(&self) -> String {
54
+ format!(
55
+ "#<Lingua::Segment {} ({}-{}) \"{}\">",
56
+ self.language, self.start_index, self.end_index, self.text
57
+ )
58
+ }
59
+ }
@@ -1,5 +1,5 @@
1
1
  # frozen_string_literal: true
2
2
 
3
3
  module Lingua
4
- VERSION = '0.1.0'
4
+ VERSION = '0.3.0'
5
5
  end
metadata CHANGED
@@ -1,7 +1,7 @@
1
1
  --- !ruby/object:Gem::Specification
2
2
  name: lingua_rs
3
3
  version: !ruby/object:Gem::Version
4
- version: 0.1.0
4
+ version: 0.3.0
5
5
  platform: ruby
6
6
  authors:
7
7
  - Sébastien Vrillaud
@@ -59,6 +59,7 @@ files:
59
59
  - ext/lingua/src/helpers.rs
60
60
  - ext/lingua/src/language.rs
61
61
  - ext/lingua/src/lib.rs
62
+ - ext/lingua/src/segment.rs
62
63
  - lib/lingua.rb
63
64
  - lib/lingua/version.rb
64
65
  - sig/lingua.rbs
@@ -66,9 +67,10 @@ homepage: https://github.com/kochka/lingua_rs
66
67
  licenses:
67
68
  - MIT
68
69
  metadata:
69
- homepage_uri: https://github.com/kochka/lingua_rs
70
70
  source_code_uri: https://github.com/kochka/lingua_rs
71
71
  changelog_uri: https://github.com/kochka/lingua_rs/blob/main/CHANGELOG.md
72
+ documentation_uri: https://github.com/kochka/lingua_rs#readme
73
+ bug_tracker_uri: https://github.com/kochka/lingua_rs/issues
72
74
  rdoc_options: []
73
75
  require_paths:
74
76
  - lib