langdetect-ruby 0.1.1
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 +7 -0
- data/CHANGELOG.md +9 -0
- data/LICENSE +21 -0
- data/README.md +33 -0
- data/Rakefile +11 -0
- data/langdetect-ruby.gemspec +33 -0
- data/lib/lingua_ruby/configuration.rb +15 -0
- data/lib/lingua_ruby/detector.rb +214 -0
- data/lib/lingua_ruby/ngram.rb +34 -0
- data/lib/lingua_ruby/profile.rb +31 -0
- data/lib/lingua_ruby/profile_loader.rb +39 -0
- data/lib/lingua_ruby/profiles/ar.json +302 -0
- data/lib/lingua_ruby/profiles/en.json +302 -0
- data/lib/lingua_ruby/profiles/id.json +402 -0
- data/lib/lingua_ruby/profiles/ja.json +302 -0
- data/lib/lingua_ruby/profiles/jv.json +302 -0
- data/lib/lingua_ruby/profiles/ko.json +302 -0
- data/lib/lingua_ruby/profiles/ms.json +402 -0
- data/lib/lingua_ruby/profiles/nl.json +302 -0
- data/lib/lingua_ruby/profiles/su.json +402 -0
- data/lib/lingua_ruby/profiles/zh.json +302 -0
- data/lib/lingua_ruby/result.rb +54 -0
- data/lib/lingua_ruby/version.rb +5 -0
- data/lib/lingua_ruby.rb +49 -0
- metadata +95 -0
|
@@ -0,0 +1,54 @@
|
|
|
1
|
+
# frozen_string_literal: true
|
|
2
|
+
|
|
3
|
+
module LinguaRuby
|
|
4
|
+
class Result
|
|
5
|
+
attr_reader :language, :confidence, :name
|
|
6
|
+
|
|
7
|
+
LANGUAGE_NAMES = {
|
|
8
|
+
id: "Indonesian",
|
|
9
|
+
en: "English",
|
|
10
|
+
ms: "Malay",
|
|
11
|
+
jv: "Javanese",
|
|
12
|
+
su: "Sundanese",
|
|
13
|
+
nl: "Dutch",
|
|
14
|
+
ar: "Arabic",
|
|
15
|
+
zh: "Chinese",
|
|
16
|
+
ja: "Japanese",
|
|
17
|
+
ko: "Korean",
|
|
18
|
+
fr: "French",
|
|
19
|
+
de: "German",
|
|
20
|
+
es: "Spanish",
|
|
21
|
+
pt: "Portuguese",
|
|
22
|
+
it: "Italian",
|
|
23
|
+
ru: "Russian",
|
|
24
|
+
tr: "Turkish",
|
|
25
|
+
pl: "Polish",
|
|
26
|
+
sv: "Swedish",
|
|
27
|
+
da: "Danish",
|
|
28
|
+
no: "Norwegian",
|
|
29
|
+
fi: "Finnish",
|
|
30
|
+
th: "Thai",
|
|
31
|
+
vi: "Vietnamese",
|
|
32
|
+
tl: "Tagalog",
|
|
33
|
+
hi: "Hindi"
|
|
34
|
+
}.freeze
|
|
35
|
+
|
|
36
|
+
def initialize(language:, confidence:)
|
|
37
|
+
@language = language.to_sym
|
|
38
|
+
@confidence = confidence
|
|
39
|
+
@name = LANGUAGE_NAMES[@language] || @language.to_s.upcase
|
|
40
|
+
end
|
|
41
|
+
|
|
42
|
+
def to_h
|
|
43
|
+
{ language: @language, confidence: @confidence, name: @name }
|
|
44
|
+
end
|
|
45
|
+
|
|
46
|
+
def to_s
|
|
47
|
+
"#{@name} (#{@language}): #{(@confidence * 100).round(1)}%"
|
|
48
|
+
end
|
|
49
|
+
|
|
50
|
+
def <=>(other)
|
|
51
|
+
other.confidence <=> @confidence
|
|
52
|
+
end
|
|
53
|
+
end
|
|
54
|
+
end
|
data/lib/lingua_ruby.rb
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
1
|
+
# frozen_string_literal: true
|
|
2
|
+
|
|
3
|
+
require_relative "lingua_ruby/version"
|
|
4
|
+
require_relative "lingua_ruby/configuration"
|
|
5
|
+
require_relative "lingua_ruby/ngram"
|
|
6
|
+
require_relative "lingua_ruby/result"
|
|
7
|
+
require_relative "lingua_ruby/profile"
|
|
8
|
+
require_relative "lingua_ruby/profile_loader"
|
|
9
|
+
require_relative "lingua_ruby/detector"
|
|
10
|
+
|
|
11
|
+
module LinguaRuby
|
|
12
|
+
class Error < StandardError; end
|
|
13
|
+
|
|
14
|
+
class << self
|
|
15
|
+
def configuration
|
|
16
|
+
@configuration ||= Configuration.new
|
|
17
|
+
end
|
|
18
|
+
|
|
19
|
+
def configure
|
|
20
|
+
yield(configuration)
|
|
21
|
+
end
|
|
22
|
+
|
|
23
|
+
def reset_configuration!
|
|
24
|
+
@configuration = Configuration.new
|
|
25
|
+
@default_detector = nil
|
|
26
|
+
end
|
|
27
|
+
|
|
28
|
+
def detect(text)
|
|
29
|
+
default_detector.detect(text)
|
|
30
|
+
end
|
|
31
|
+
|
|
32
|
+
def detect_all(text)
|
|
33
|
+
default_detector.detect_all(text)
|
|
34
|
+
end
|
|
35
|
+
|
|
36
|
+
def detect_batch(texts)
|
|
37
|
+
raise ArgumentError, "texts must be an Array" unless texts.is_a?(Array)
|
|
38
|
+
|
|
39
|
+
detector = default_detector
|
|
40
|
+
texts.map { |text| detector.detect(text) }
|
|
41
|
+
end
|
|
42
|
+
|
|
43
|
+
private
|
|
44
|
+
|
|
45
|
+
def default_detector
|
|
46
|
+
@default_detector ||= Detector.new
|
|
47
|
+
end
|
|
48
|
+
end
|
|
49
|
+
end
|
metadata
ADDED
|
@@ -0,0 +1,95 @@
|
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
|
2
|
+
name: langdetect-ruby
|
|
3
|
+
version: !ruby/object:Gem::Version
|
|
4
|
+
version: 0.1.1
|
|
5
|
+
platform: ruby
|
|
6
|
+
authors:
|
|
7
|
+
- Johannes Dwi Cahyo
|
|
8
|
+
bindir: bin
|
|
9
|
+
cert_chain: []
|
|
10
|
+
date: 1980-01-02 00:00:00.000000000 Z
|
|
11
|
+
dependencies:
|
|
12
|
+
- !ruby/object:Gem::Dependency
|
|
13
|
+
name: minitest
|
|
14
|
+
requirement: !ruby/object:Gem::Requirement
|
|
15
|
+
requirements:
|
|
16
|
+
- - "~>"
|
|
17
|
+
- !ruby/object:Gem::Version
|
|
18
|
+
version: '5.0'
|
|
19
|
+
type: :development
|
|
20
|
+
prerelease: false
|
|
21
|
+
version_requirements: !ruby/object:Gem::Requirement
|
|
22
|
+
requirements:
|
|
23
|
+
- - "~>"
|
|
24
|
+
- !ruby/object:Gem::Version
|
|
25
|
+
version: '5.0'
|
|
26
|
+
- !ruby/object:Gem::Dependency
|
|
27
|
+
name: rake
|
|
28
|
+
requirement: !ruby/object:Gem::Requirement
|
|
29
|
+
requirements:
|
|
30
|
+
- - "~>"
|
|
31
|
+
- !ruby/object:Gem::Version
|
|
32
|
+
version: '13.0'
|
|
33
|
+
type: :development
|
|
34
|
+
prerelease: false
|
|
35
|
+
version_requirements: !ruby/object:Gem::Requirement
|
|
36
|
+
requirements:
|
|
37
|
+
- - "~>"
|
|
38
|
+
- !ruby/object:Gem::Version
|
|
39
|
+
version: '13.0'
|
|
40
|
+
description: Pure Ruby language detection library using character n-gram frequency
|
|
41
|
+
profiles. Detects 50+ languages with high accuracy.
|
|
42
|
+
email:
|
|
43
|
+
- johannes@example.com
|
|
44
|
+
executables: []
|
|
45
|
+
extensions: []
|
|
46
|
+
extra_rdoc_files: []
|
|
47
|
+
files:
|
|
48
|
+
- CHANGELOG.md
|
|
49
|
+
- LICENSE
|
|
50
|
+
- README.md
|
|
51
|
+
- Rakefile
|
|
52
|
+
- langdetect-ruby.gemspec
|
|
53
|
+
- lib/lingua_ruby.rb
|
|
54
|
+
- lib/lingua_ruby/configuration.rb
|
|
55
|
+
- lib/lingua_ruby/detector.rb
|
|
56
|
+
- lib/lingua_ruby/ngram.rb
|
|
57
|
+
- lib/lingua_ruby/profile.rb
|
|
58
|
+
- lib/lingua_ruby/profile_loader.rb
|
|
59
|
+
- lib/lingua_ruby/profiles/ar.json
|
|
60
|
+
- lib/lingua_ruby/profiles/en.json
|
|
61
|
+
- lib/lingua_ruby/profiles/id.json
|
|
62
|
+
- lib/lingua_ruby/profiles/ja.json
|
|
63
|
+
- lib/lingua_ruby/profiles/jv.json
|
|
64
|
+
- lib/lingua_ruby/profiles/ko.json
|
|
65
|
+
- lib/lingua_ruby/profiles/ms.json
|
|
66
|
+
- lib/lingua_ruby/profiles/nl.json
|
|
67
|
+
- lib/lingua_ruby/profiles/su.json
|
|
68
|
+
- lib/lingua_ruby/profiles/zh.json
|
|
69
|
+
- lib/lingua_ruby/result.rb
|
|
70
|
+
- lib/lingua_ruby/version.rb
|
|
71
|
+
homepage: https://github.com/johannesdwicahyo/lingua-ruby
|
|
72
|
+
licenses:
|
|
73
|
+
- MIT
|
|
74
|
+
metadata:
|
|
75
|
+
homepage_uri: https://github.com/johannesdwicahyo/lingua-ruby
|
|
76
|
+
source_code_uri: https://github.com/johannesdwicahyo/lingua-ruby
|
|
77
|
+
changelog_uri: https://github.com/johannesdwicahyo/lingua-ruby/blob/main/CHANGELOG.md
|
|
78
|
+
rdoc_options: []
|
|
79
|
+
require_paths:
|
|
80
|
+
- lib
|
|
81
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
|
82
|
+
requirements:
|
|
83
|
+
- - ">="
|
|
84
|
+
- !ruby/object:Gem::Version
|
|
85
|
+
version: 3.0.0
|
|
86
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
87
|
+
requirements:
|
|
88
|
+
- - ">="
|
|
89
|
+
- !ruby/object:Gem::Version
|
|
90
|
+
version: '0'
|
|
91
|
+
requirements: []
|
|
92
|
+
rubygems_version: 3.6.9
|
|
93
|
+
specification_version: 4
|
|
94
|
+
summary: Language detection for Ruby using n-gram profiles
|
|
95
|
+
test_files: []
|