morphy 0.0.2
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +18 -0
- data/Gemfile +4 -0
- data/LICENSE +22 -0
- data/README.md +2 -0
- data/lib/dictionary/dawg.dat +0 -0
- data/lib/dictionary/grammemes.txt +5008 -0
- data/lib/dictionary/paradigms.dat +0 -0
- data/lib/dictionary/prefixes.txt +509 -0
- data/lib/dictionary/suffixes.txt +4979 -0
- data/lib/morphy.rb +98 -0
- data/morphy.gemspec +19 -0
- metadata +76 -0
data/lib/morphy.rb
ADDED
@@ -0,0 +1,98 @@
|
|
1
|
+
require "dawg"
|
2
|
+
|
3
|
+
class Morphy
|
4
|
+
class Word
|
5
|
+
attr_accessor :para_id
|
6
|
+
|
7
|
+
def initialize(word,para_id,index)
|
8
|
+
@word = word
|
9
|
+
@para_id = para_id.to_i
|
10
|
+
@index = index.to_i
|
11
|
+
@prefix_id = Morphy.paradigms[@para_id][@index*3]
|
12
|
+
@suffix_id = Morphy.paradigms[@para_id][@index*3+1]
|
13
|
+
@grammeme_id = Morphy.paradigms[@para_id][@index*3+2]
|
14
|
+
|
15
|
+
end
|
16
|
+
|
17
|
+
def to_s
|
18
|
+
@word
|
19
|
+
end
|
20
|
+
|
21
|
+
def normal_form
|
22
|
+
prefix = Morphy.prefixes[Morphy.paradigms[@para_id][0]]
|
23
|
+
suffix = Morphy.suffixes[Morphy.paradigms[@para_id][1]]
|
24
|
+
|
25
|
+
"#{prefix}#{stem}#{suffix}"
|
26
|
+
end
|
27
|
+
def grammemes
|
28
|
+
Morphy.grammemes[@grammeme_id]
|
29
|
+
end
|
30
|
+
def stem
|
31
|
+
prefix = Morphy.prefixes[Morphy.paradigms[@para_id][0]]
|
32
|
+
suffix = Morphy.suffixes[Morphy.paradigms[@para_id][1]]
|
33
|
+
grammeme = Morphy.grammemes[Morphy.paradigms[@para_id][2]]
|
34
|
+
word = @word.dup
|
35
|
+
word.gsub!(Morphy.prefixes[@prefix_id],"")
|
36
|
+
word.gsub!(Morphy.suffixes[@suffix_id],"")
|
37
|
+
word
|
38
|
+
end
|
39
|
+
|
40
|
+
def same_paradigm?(other)
|
41
|
+
@para_id == other.para_id
|
42
|
+
end
|
43
|
+
|
44
|
+
def tag
|
45
|
+
Morphy.grammemes[@grammeme_id].join(",")
|
46
|
+
end
|
47
|
+
|
48
|
+
def lexemme
|
49
|
+
(0..(Morphy.paradigms[@para_id].length / 3)-1).map do |index|
|
50
|
+
prefix = Morphy.prefixes[Morphy.paradigms[@para_id][index*3]]
|
51
|
+
suffix = Morphy.suffixes[Morphy.paradigms[@para_id][index*3+1]]
|
52
|
+
grammeme = Morphy.grammemes[Morphy.paradigms[@para_id][index*3+2]]
|
53
|
+
Word.new(prefix+stem+suffix,@para_id,index)
|
54
|
+
end
|
55
|
+
end
|
56
|
+
|
57
|
+
def inflect(grammemes)
|
58
|
+
words = lexemme
|
59
|
+
words.each do |word|
|
60
|
+
return word if word.grammemes.last(grammemes.length) == grammemes
|
61
|
+
end
|
62
|
+
nil
|
63
|
+
end
|
64
|
+
end
|
65
|
+
def initialize
|
66
|
+
|
67
|
+
path = File.dirname(__FILE__)+"/dictionary/"
|
68
|
+
|
69
|
+
@dawg = Dawg.load("#{path}/dawg.dat") # why it's eating so much memory?
|
70
|
+
@@suffixes ||= File.open("#{path}/suffixes.txt", 'r').read.split("\n")
|
71
|
+
@@prefixes ||= File.open("#{path}/prefixes.txt", 'r').read.split("\n")
|
72
|
+
@@grammemes ||= File.open("#{path}/grammemes.txt", 'r').read.split("\n").map{|g| g.split(",")}
|
73
|
+
@@paradigms ||= Marshal.load(File.read("#{path}/paradigms.dat"))
|
74
|
+
end
|
75
|
+
def self.paradigms
|
76
|
+
@@paradigms
|
77
|
+
end
|
78
|
+
def self.prefixes
|
79
|
+
@@prefixes
|
80
|
+
end
|
81
|
+
def self.suffixes
|
82
|
+
@@suffixes
|
83
|
+
end
|
84
|
+
def self.grammemes
|
85
|
+
@@grammemes
|
86
|
+
end
|
87
|
+
def find_similar(word)
|
88
|
+
results = @dawg.find_similar(word)
|
89
|
+
results = results.map do |result|
|
90
|
+
word,para_id,index = result.split(" ")
|
91
|
+
Word.new(word,para_id,index)
|
92
|
+
end
|
93
|
+
results
|
94
|
+
end
|
95
|
+
def to_s
|
96
|
+
"Morphy"
|
97
|
+
end
|
98
|
+
end
|
data/morphy.gemspec
ADDED
@@ -0,0 +1,19 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
|
3
|
+
Gem::Specification.new do |spec|
|
4
|
+
spec.name = "morphy"
|
5
|
+
spec.version = "0.0.2"
|
6
|
+
spec.date = '2015-05-29'
|
7
|
+
spec.authors = ["Maksatbek Mansurov"]
|
8
|
+
spec.email = ["maksat.mansurov@gmail.com"]
|
9
|
+
spec.description = %q{Morphological analyzer (POS tagger + inflection engine) for Russian language in ruby. Inspired by pymorphy2}
|
10
|
+
spec.summary = %q{Morphological analyzer for Russian language in ruby. Inspired by pymorphy2}
|
11
|
+
spec.homepage = "https://github.com/baltavay/morphy"
|
12
|
+
spec.license = "MIT"
|
13
|
+
|
14
|
+
spec.files = `git ls-files`.split($/)
|
15
|
+
spec.executables = spec.files.grep(%r{^bin/}) { |f| File.basename(f) }
|
16
|
+
spec.test_files = spec.files.grep(%r{^(test|spec|features)/})
|
17
|
+
spec.require_paths = ["lib"]
|
18
|
+
spec.add_development_dependency 'dawg', '~> 0.0', '>= 0.0.2'
|
19
|
+
end
|
metadata
ADDED
@@ -0,0 +1,76 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: morphy
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.2
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Maksatbek Mansurov
|
8
|
+
autorequire:
|
9
|
+
bindir: bin
|
10
|
+
cert_chain: []
|
11
|
+
date: 2015-05-29 00:00:00.000000000 Z
|
12
|
+
dependencies:
|
13
|
+
- !ruby/object:Gem::Dependency
|
14
|
+
name: dawg
|
15
|
+
requirement: !ruby/object:Gem::Requirement
|
16
|
+
requirements:
|
17
|
+
- - "~>"
|
18
|
+
- !ruby/object:Gem::Version
|
19
|
+
version: '0.0'
|
20
|
+
- - ">="
|
21
|
+
- !ruby/object:Gem::Version
|
22
|
+
version: 0.0.2
|
23
|
+
type: :development
|
24
|
+
prerelease: false
|
25
|
+
version_requirements: !ruby/object:Gem::Requirement
|
26
|
+
requirements:
|
27
|
+
- - "~>"
|
28
|
+
- !ruby/object:Gem::Version
|
29
|
+
version: '0.0'
|
30
|
+
- - ">="
|
31
|
+
- !ruby/object:Gem::Version
|
32
|
+
version: 0.0.2
|
33
|
+
description: Morphological analyzer (POS tagger + inflection engine) for Russian language
|
34
|
+
in ruby. Inspired by pymorphy2
|
35
|
+
email:
|
36
|
+
- maksat.mansurov@gmail.com
|
37
|
+
executables: []
|
38
|
+
extensions: []
|
39
|
+
extra_rdoc_files: []
|
40
|
+
files:
|
41
|
+
- ".gitignore"
|
42
|
+
- Gemfile
|
43
|
+
- LICENSE
|
44
|
+
- README.md
|
45
|
+
- lib/dictionary/dawg.dat
|
46
|
+
- lib/dictionary/grammemes.txt
|
47
|
+
- lib/dictionary/paradigms.dat
|
48
|
+
- lib/dictionary/prefixes.txt
|
49
|
+
- lib/dictionary/suffixes.txt
|
50
|
+
- lib/morphy.rb
|
51
|
+
- morphy.gemspec
|
52
|
+
homepage: https://github.com/baltavay/morphy
|
53
|
+
licenses:
|
54
|
+
- MIT
|
55
|
+
metadata: {}
|
56
|
+
post_install_message:
|
57
|
+
rdoc_options: []
|
58
|
+
require_paths:
|
59
|
+
- lib
|
60
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
61
|
+
requirements:
|
62
|
+
- - ">="
|
63
|
+
- !ruby/object:Gem::Version
|
64
|
+
version: '0'
|
65
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
66
|
+
requirements:
|
67
|
+
- - ">="
|
68
|
+
- !ruby/object:Gem::Version
|
69
|
+
version: '0'
|
70
|
+
requirements: []
|
71
|
+
rubyforge_project:
|
72
|
+
rubygems_version: 2.2.2
|
73
|
+
signing_key:
|
74
|
+
specification_version: 4
|
75
|
+
summary: Morphological analyzer for Russian language in ruby. Inspired by pymorphy2
|
76
|
+
test_files: []
|