rbbt-text 1.3.10 → 1.3.11
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/lib/rbbt/bow/misc.rb +1 -1
- data/lib/rbbt/document/corpus/pubmed.rb +1 -1
- metadata +27 -27
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 587b7971fd0f8fec2602323ea480521445ca417c3a407d057cf95df1f3a36216
|
4
|
+
data.tar.gz: 4f66306f80e838a0a27299705d79e5856b38dd936005d2b18004539bbb192431
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 4aa191aa4e5cb5e3f7d4a49b30beb1eb7259a34074a7521a10b9951cd1cc7a097a06ba6a97d9f4f4e100b2058de3f94f5199cb069a030f93a3f69bf1ecec09ff
|
7
|
+
data.tar.gz: f3d5eb11d12f8a9d951d1073abd7e6cb5ace99bd075e7dc897f0aa715ae1552271019b4eb6849a172529d830f30bbc09ce40dd351fd81f9c06f338b075523e36
|
data/lib/rbbt/bow/misc.rb
CHANGED
@@ -37,7 +37,7 @@ $greek.each{|l,s| $inverse_greek[s] = l }
|
|
37
37
|
|
38
38
|
class String
|
39
39
|
CONSONANTS = []
|
40
|
-
if File.
|
40
|
+
if File.exist? File.join(Rbbt.datadir, 'wordlists/consonants')
|
41
41
|
Object::Open.read(File.join(Rbbt.datadir, 'wordlists/consonants')).each_line{|l| CONSONANTS << l.chomp}
|
42
42
|
end
|
43
43
|
|
@@ -20,7 +20,7 @@ module Document::Corpus
|
|
20
20
|
elsif type.to_sym == :title
|
21
21
|
Document.setup(article.title || "", PUBMED_NAMESPACE, pmid, type.to_sym, self)
|
22
22
|
elsif type.to_sym == :title_and_abstract
|
23
|
-
Document.setup((article.title || "") + "
|
23
|
+
Document.setup((article.title || "") + " " + (article.abstract || ""), PUBMED_NAMESPACE, pmid, type.to_sym, self)
|
24
24
|
else
|
25
25
|
raise "No FullText available for #{ pmid }" if article.full_text.nil?
|
26
26
|
Document.setup(article.full_text, PUBMED_NAMESPACE, pmid, :fulltext, self, :corpus => self)
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: rbbt-text
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.3.
|
4
|
+
version: 1.3.11
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Miguel Vazquez
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2023-
|
11
|
+
date: 2023-04-11 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: rbbt-util
|
@@ -175,45 +175,45 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
175
175
|
- !ruby/object:Gem::Version
|
176
176
|
version: '0'
|
177
177
|
requirements: []
|
178
|
-
rubygems_version: 3.
|
178
|
+
rubygems_version: 3.4.8
|
179
179
|
signing_key:
|
180
180
|
specification_version: 4
|
181
181
|
summary: Text mining tools for the Ruby Bioinformatics Toolkit (rbbt)
|
182
182
|
test_files:
|
183
|
-
- test/
|
184
|
-
- test/
|
185
|
-
- test/rbbt/
|
183
|
+
- test/rbbt/nlp/test_nlp.rb
|
184
|
+
- test/rbbt/nlp/open_nlp/test_sentence_splitter.rb
|
185
|
+
- test/rbbt/nlp/genia/test_sentence_splitter.rb
|
186
186
|
- test/rbbt/bow/test_bow.rb
|
187
187
|
- test/rbbt/bow/test_misc.rb
|
188
|
-
- test/rbbt/
|
189
|
-
- test/rbbt/
|
190
|
-
- test/rbbt/
|
191
|
-
- test/rbbt/segment/test_named_entity.rb
|
192
|
-
- test/rbbt/segment/test_corpus.rb
|
193
|
-
- test/rbbt/segment/test_range_index.rb
|
194
|
-
- test/rbbt/segment/test_annotation.rb
|
195
|
-
- test/rbbt/entity/test_document.rb
|
188
|
+
- test/rbbt/bow/test_dictionary.rb
|
189
|
+
- test/rbbt/test_document.rb
|
190
|
+
- test/rbbt/document/test_annotation.rb
|
196
191
|
- test/rbbt/document/corpus/test_pubmed.rb
|
197
192
|
- test/rbbt/document/test_corpus.rb
|
198
|
-
- test/rbbt/
|
199
|
-
- test/rbbt/test_document.rb
|
193
|
+
- test/rbbt/entity/test_document.rb
|
200
194
|
- test/rbbt/ner/test_patterns.rb
|
195
|
+
- test/rbbt/ner/test_NER.rb
|
196
|
+
- test/rbbt/ner/test_abner.rb
|
201
197
|
- test/rbbt/ner/rnorm/test_tokens.rb
|
198
|
+
- test/rbbt/ner/test_rnorm.rb
|
199
|
+
- test/rbbt/ner/test_regexpNER.rb
|
202
200
|
- test/rbbt/ner/test_ngram_prefix_dictionary.rb
|
203
|
-
- test/rbbt/ner/test_token_trieNER.rb
|
204
|
-
- test/rbbt/ner/test_finder.rb
|
205
201
|
- test/rbbt/ner/test_brat.rb
|
206
|
-
- test/rbbt/ner/test_regexpNER.rb
|
207
202
|
- test/rbbt/ner/test_g_norm_plus.rb
|
208
|
-
- test/rbbt/ner/test_rnorm.rb
|
209
|
-
- test/rbbt/ner/test_linnaeus.rb
|
210
203
|
- test/rbbt/ner/test_chemical_tagger.rb
|
211
|
-
- test/rbbt/ner/
|
212
|
-
- test/rbbt/ner/
|
204
|
+
- test/rbbt/ner/test_banner.rb
|
205
|
+
- test/rbbt/ner/test_token_trieNER.rb
|
206
|
+
- test/rbbt/ner/test_finder.rb
|
213
207
|
- test/rbbt/ner/test_rner.rb
|
208
|
+
- test/rbbt/ner/test_linnaeus.rb
|
214
209
|
- test/rbbt/ner/test_oscar4.rb
|
215
|
-
- test/rbbt/ner/test_banner.rb
|
216
210
|
- test/rbbt/test_segment.rb
|
217
|
-
- test/rbbt/
|
218
|
-
- test/rbbt/
|
219
|
-
- test/rbbt/
|
211
|
+
- test/rbbt/segment/test_transformed.rb
|
212
|
+
- test/rbbt/segment/test_overlaps.rb
|
213
|
+
- test/rbbt/segment/test_annotation.rb
|
214
|
+
- test/rbbt/segment/test_named_entity.rb
|
215
|
+
- test/rbbt/segment/test_encoding.rb
|
216
|
+
- test/rbbt/segment/test_range_index.rb
|
217
|
+
- test/rbbt/segment/test_corpus.rb
|
218
|
+
- test/test_spaCy.rb
|
219
|
+
- test/test_helper.rb
|