rbbt-text 1.5.2 → 1.6.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 80e662635b01507c60638565bb8f00f789b4e6423227b98cee0ac38d8dd4b344
-  data.tar.gz: 6b9a1be5fdea2bb56a770ca2fe6838528102dfe2d15f27abb5f0e5b1849e6574
+  metadata.gz: fe1271c362c3e5f912d1d2236587dfd3f5202c8bef0246b547f3d109d3a321b5
+  data.tar.gz: abbd7559a60012f910a7d5c528484bf10950109772c6fd81cffba85efc0541bd
 SHA512:
-  metadata.gz: 383affa64fa2b1e6d54817e343b3710d134c01d68867699490c604258324a6806bd829492873e819a5666780e0282ab1a375ae03147691aa64ef78705d7d0097
-  data.tar.gz: 67a52fca24335775faacf2a68cd101b24ff19d0fbff9560b4468c61546cee62aa02cdf8c5a1d3db39e5784856bd4df4573c9a00f901e8ca368cf169f8a88101d
+  metadata.gz: bf1b61c0464eb675861694e21c597f0770b37b241e72008225d49e826ddeb93877c50ceaa63b5a05daee1274d9ce3b433308d32ff71fe9455e14af1c55ac7eca
+  data.tar.gz: 532381c12fc4886b2c178252c1a5d2df6cd45921e1725a8dc1a45f4325044710822ab67fe431e361bb7b14159d92058deeb7257c52f0f86e94fd88f851669933

data/lib/rbbt/document/annotation.rb

@@ -3,8 +3,8 @@ require 'rbbt/segment/annotation'
 
 module Document
   def self.define(type, &block)
-    send :property, type do 
-      segments = self.instance_exec &block
+    send :property, type do |*args,**kwargs|
+      segments = self.instance_exec *args, **kwargs, &block
 
       Segment.align(self, segments) unless segments.empty? || 
           (Segment === segments && segments.offset) || 
@@ -53,6 +53,7 @@ module Document
 
   def self.define_multiple(type, &block)
     send :property, type => :multiple do |list|
+      list = self if Array === self
       doc_segments = self.instance_exec list, &block
 
       doc_segments = doc_segments.chunked_values_at(list) if Hash === doc_segments

data/lib/rbbt/document/corpus.rb

@@ -1,4 +1,5 @@
 require 'rbbt-util'
+require 'rbbt/tsv'
 
 module Document::Corpus
 
@@ -12,7 +13,9 @@ module Document::Corpus
 
   def add_document(document)
     docid = document.docid
-    return self[docid] if self.include?(docid)
+    self.read_and_close do
+      return self[docid] if self.include?(docid)
+    end
     self.write_and_close do
       self[docid] = document
     end
@@ -34,7 +37,7 @@ module Document::Corpus
   def [](*args)
     docid, *rest = args
 
-    res = self.with_read do
+    res = self.read_and_close do
       super(*args)
     end
     

data/lib/rbbt/document.rb

@@ -41,7 +41,7 @@ module Document
   extend Entity
   self.annotation :namespace, :code, :type, :corpus
 
-  property :docid do |corpus=nil|
+  def docid(corpus=nil)
     digest = Misc.digest(self)
     corpus = self.corpus if corpus.nil?
 

data/lib/rbbt/ner/pubtator.rb

@@ -1,4 +1,3 @@
-
 require 'rbbt/segment'
 require 'rbbt/segment/named_entity'
 require 'rbbt/segment/transformed'

data/lib/rbbt/ner/token_trieNER.rb

@@ -266,8 +266,8 @@ class TokenTrieNER < NER
     file = [] if file.nil?
     file = [file] unless Array === file
     persist_options = Misc.pull_keys options, :persist
-    @index = Persist.persist_tsv(file, options, persist_options) do |data|
-      data.serializer = :marshal if data.respond_to? :serializer and data.serializer == :type
+    @index = Persist.persist_tsv(file, nil, options, persist_options) do |data|
+      data.serializer = :marshal if data.respond_to?(:serializer) and (data.serializer == :type || data.serializer.nil?)
 
       @index = data
       file.each do |f| 
@@ -282,7 +282,7 @@ class TokenTrieNER < NER
     case
     when TokenTrieNER === new
       Log.debug "TokenTrieNER merging other TokenTrieNER"
-      TokenTrieNER.merge(@index, new.index)
+     TokenTrieNER.merge(@index, new.index)
     when TSV === new
       Log.debug "TokenTrieNER merging TSV"
       new.with_unnamed do

data/lib/rbbt/segment/annotation.rb

@@ -36,7 +36,7 @@ module SegmentAnnotation
   include Object::Segment
   self.annotation :type
 
-  property :segid do
+  def segid
     case self
     when SegID
       self

data/test/rbbt/ner/test_pubtator.rb

@@ -12,7 +12,7 @@ class TestPubtator < Test::Unit::TestCase
     end
   end
 
-  def _test_align
+  def test_align
     pmids = "19522013|20861254|38267746".split("|")
     alignments = {} 
     with_corpus do |corpus|
@@ -47,7 +47,7 @@ class TestPubtator < Test::Unit::TestCase
     end
   end
 
-  def _test_greek
+  def test_greek
 
     pmids = "20861254".split("|")
     alignments = {} 

data/test/rbbt/ner/test_rnorm.rb

@@ -45,7 +45,7 @@ S000000376	AAA	GENE1	DDD
   end
 
   def test_token_evaluate
-    iii @norm.token_evaluate("PDGFRA","PDGFRalpha")
+    assert @norm.token_evaluate("PDGFRA","PDGFRalpha") > 0
   end
 
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rbbt-text
 version: !ruby/object:Gem::Version
-  version: 1.5.2
+  version: 1.6.0
 platform: ruby
 authors:
 - Miguel Vazquez
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2024-02-28 00:00:00.000000000 Z
+date: 2024-05-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rbbt-util
@@ -179,7 +179,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.5.0.dev
+rubygems_version: 3.5.10
 signing_key: 
 specification_version: 4
 summary: Text mining tools for the Ruby Bioinformatics Toolkit (rbbt)