kokugo_tagger 0.0.2 → 0.0.3
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/lib/kokugo_tagger/parser.rb +3 -3
- data/lib/kokugo_tagger/tagger.rb +38 -2
- data/lib/kokugo_tagger/version.rb +1 -1
- metadata +3 -3
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 38c7ab0d9feb7ce099075914af22947354187b40
|
4
|
+
data.tar.gz: b240e3e012dce85475716a16ed6b1ff738d65327
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 28c3d0f941e530245655e9704d675343fdabb9f94123bf8065aad4044492b9de0b50e2cead2cd99b1ceecceb8a6d89a2417a8d52566d39a8aed3250aec50d93f
|
7
|
+
data.tar.gz: d2a488f3efdfeaabcaf72d610052db4327ba8af401b9e974f2e191e0579efe730d80dc5d9086a9db1a3eeaf4a4a5f89ab1b6e5136ced2cdfa75a66bda4a25797
|
data/lib/kokugo_tagger/parser.rb
CHANGED
@@ -29,10 +29,10 @@ module CabochaParser
|
|
29
29
|
return excab
|
30
30
|
end
|
31
31
|
def parse_chunk(line)
|
32
|
-
null, id,
|
33
|
-
link,
|
32
|
+
null, id, dep, part, score = line.chomp.split("\s")
|
33
|
+
link, rel = dep[0..-2], dep[-1]
|
34
34
|
head, func = part.split('/')
|
35
|
-
chunk = {type: 'CHUNK', id: id, link: link,
|
35
|
+
chunk = {type: 'CHUNK', id: id, link: link, rel: rel, head: head, func: func, score: score}
|
36
36
|
return chunk
|
37
37
|
end
|
38
38
|
def parse_token(line)
|
data/lib/kokugo_tagger/tagger.rb
CHANGED
@@ -25,8 +25,18 @@ module KokugoTagger
|
|
25
25
|
cform data
|
26
26
|
end
|
27
27
|
def segment_s(data)
|
28
|
+
@segments ||= []
|
29
|
+
@segments << data
|
30
|
+
@last_item = data
|
31
|
+
end
|
32
|
+
def group_s(data)
|
33
|
+
@groups ||= []
|
34
|
+
@groups << data
|
35
|
+
@last_item = data
|
28
36
|
end
|
29
37
|
def attr(data)
|
38
|
+
@last_item[:attributes] ||= []
|
39
|
+
@last_item[:attributes] << data
|
30
40
|
end
|
31
41
|
def eos(data)
|
32
42
|
before_eos
|
@@ -35,7 +45,7 @@ module KokugoTagger
|
|
35
45
|
puts '#! ATTR bccwj-kok:pred "%s述語"' % chunk[:pos] if chunk[:pred]
|
36
46
|
puts '#! ATTR bccwj-kok:conj "%s"' % chunk[:conj] if chunk[:conj]
|
37
47
|
end
|
38
|
-
@chunks, @chunk, @lpos, @segments = nil
|
48
|
+
@chunks, @chunk, @lpos, @segments, @groups = nil
|
39
49
|
end
|
40
50
|
def pos(token)
|
41
51
|
case token[:pos]
|
@@ -61,10 +71,12 @@ module KokugoTagger
|
|
61
71
|
case token[:text]
|
62
72
|
when 'が'
|
63
73
|
@chunk.update conj:'主語'
|
74
|
+
when 'を', 'に'
|
75
|
+
@chunk.update conj:'補語'
|
64
76
|
when 'の', 'との', 'という', 'といった'
|
65
77
|
@chunk.update conj:'修飾(連体)'
|
66
78
|
else
|
67
|
-
@chunk.update conj:'
|
79
|
+
@chunk.update conj:'修飾(連用)'
|
68
80
|
end
|
69
81
|
when /^(助詞-副助詞|助詞-係助詞)/
|
70
82
|
@chunk.update conj:'修飾(連用)'
|
@@ -88,6 +100,30 @@ module KokugoTagger
|
|
88
100
|
end
|
89
101
|
end
|
90
102
|
def before_eos
|
103
|
+
# BCCWJ-DepPara
|
104
|
+
@chunks.each do |chunk|
|
105
|
+
chunk[:conj] = [chunk[:conj], '断片'].compact.join(':') if chunk[:rel] == 'F'
|
106
|
+
chunk[:conj] = [chunk[:conj], '文節内'].compact.join(':') if chunk[:rel] == 'B'
|
107
|
+
chunk[:conj] = '文末' if chunk[:rel] == 'Z'
|
108
|
+
end
|
109
|
+
# 並列・同格関係
|
110
|
+
@groups ||= []
|
111
|
+
@segments ||= []
|
112
|
+
@groups.each do |group|
|
113
|
+
next unless group[:name] =~ /^(Parallel|Apposition)$/
|
114
|
+
members = group[:member].map{|n| n.to_i}
|
115
|
+
members = @segments.values_at(*members)
|
116
|
+
chunk_ids = members.map do |segment|
|
117
|
+
_end = segment[:end].to_i
|
118
|
+
chunk = @chunks.find{|c| c[:start] < _end and c[:end] >= _end}
|
119
|
+
chunk[:id].to_i if chunk
|
120
|
+
end
|
121
|
+
chunk_ids = chunk_ids.compact.uniq.sort
|
122
|
+
if chunk_ids.size > 1
|
123
|
+
conj = {'Parallel' => '並立', 'Apposition' => '同格'}[group[:name]]
|
124
|
+
chunk_ids[0..-2].each{|cid| @chunks[cid][:conj] = conj}
|
125
|
+
end
|
126
|
+
end
|
91
127
|
# 属性を付与できなかった文節に対して、係り受けを利用して属性を補完
|
92
128
|
# 連用成分を受ける文節を述語とみなす
|
93
129
|
@chunks.each do |chunk|
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: kokugo_tagger
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.0.
|
4
|
+
version: 0.0.3
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Mizuho IMADA
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date:
|
11
|
+
date: 2015-06-12 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: bundler
|
@@ -77,7 +77,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
77
77
|
version: '0'
|
78
78
|
requirements: []
|
79
79
|
rubyforge_project:
|
80
|
-
rubygems_version: 2.
|
80
|
+
rubygems_version: 2.2.2
|
81
81
|
signing_key:
|
82
82
|
specification_version: 4
|
83
83
|
summary: Write a short summary. Required.
|