rbbt-text 1.2.0 → 1.3.4

data/test/rbbt/segment/test_overlaps.rb

@@ -0,0 +1,69 @@
+require File.join(File.expand_path(File.dirname(__FILE__)), '../..', 'test_helper.rb')
+require 'rbbt/segment'
+require 'rbbt/segment/overlaps'
+
+class TestOverlaps < Test::Unit::TestCase
+  def setup
+    @text = <<-EOF
+This is a first sentence. More recently, PPAR activators were shown to inhibit the activation of inflammatory response genes (such as IL-2, IL-6, IL-8, TNF alpha and metalloproteases) by negatively interfering with the NF-kappa B, STAT and AP-1 signalling pathways in cells of the vascular wall.
+    EOF
+
+    @entities = ["PPAR", "IL-2", "IL-6", "IL-8", "TNF", "TNF alpha", "NF-kappa B", "AP-1", "STAT"].collect do |literal|
+      Segment.setup(literal, :offset => @text.index(literal))
+    end
+    
+    @sentences = @text.partition(".").values_at(0, 2).collect do |sentence|
+      Segment.setup sentence, :offset => @text.index(sentence)
+    end
+  end
+
+  def test_make_relative
+    sentence = @sentences[1]
+
+    @entities.each do |e|
+      assert_equal e, @text[e.range]
+    end
+
+    sentence.make_relative @entities do
+      @entities.each do |e|
+        assert_equal e, sentence[e.range]
+      end
+
+      @entities.each do |e|
+        assert_not_equal e, @text[e.range]
+      end
+    end
+
+    @entities.each do |e|
+      assert_equal e, @text[e.range]
+    end
+  end
+
+  def test_range_in
+    sentence = @sentences[1]
+
+    @entities.each do |e|
+      assert_equal e.range_in(sentence).begin, sentence.index(e)
+      assert_equal e.range.begin - sentence.offset, sentence.index(e)
+    end
+  end
+
+  def test_includes
+    @entities.each do |e|
+      assert ! @sentences[0].include?(e)
+      assert @sentences[1].include?(e)
+      assert ! e.include?(@sentences[0])
+      assert ! e.include?(@sentences[1])
+    end
+  end
+
+  def test_overlaps?
+    @entities.each do |e|
+      assert ! @sentences[0].overlaps?(e)
+      assert @sentences[1].overlaps?(e)
+      assert ! e.overlaps?(@sentences[0])
+      assert e.overlaps?(@sentences[1])
+    end
+  end
+
+end

data/test/rbbt/segment/test_range_index.rb

@@ -0,0 +1,42 @@
+require File.join(File.expand_path(File.dirname(__FILE__)), '../..', 'test_helper.rb')
+require 'rbbt/document'
+require 'rbbt/document/corpus'
+require 'rbbt/segment'
+require 'rbbt/segment/range_index'
+
+class TestRangeIndex < Test::Unit::TestCase
+  def test_segment_index
+    text = "This sentence mentions the TP53 gene and the CDK5R1 protein"
+    Document.setup(text, "TEST", "test_doc1", nil)
+
+    corpus = Document::Corpus.setup({})
+
+    corpus.add_document(text)
+
+    gene1 = "TP53"
+    gene1.extend Segment
+    gene1.offset = text.index gene1
+    gene1.docid = text.docid
+
+    gene2 = "CDK5R1"
+    gene2.extend Segment
+    gene2.offset = text.index gene2
+    gene2.docid = text.docid
+
+    gene3 = "TP53 gene"
+    gene3.extend Segment
+    gene3.offset = text.index gene1
+    gene3.docid = text.docid
+
+    index = Segment::RangeIndex.index([gene1, gene2, gene3], corpus)
+    assert_equal "CDK5R1", index[gene2.offset + 1].segment.first
+
+    TmpFile.with_file do |fwt|
+      index = Segment::RangeIndex.index([gene1, gene2, gene3], corpus, fwt)
+      assert_equal %w(CDK5R1), index[gene2.offset + 1].segment
+      index = Segment::RangeIndex.index([gene1, gene2, gene3], corpus, fwt)
+      assert_equal %w(CDK5R1), index[gene2.offset + 1].segment
+    end
+  end
+end
+

data/test/rbbt/{text/segment → segment}/test_transformed.rb

@@ -1,10 +1,21 @@
-require File.join(File.expand_path(File.dirname(__FILE__)), '../../..', 'test_helper.rb')
-require 'rbbt/text/segment/transformed'
-require 'rbbt/text/segment/named_entity'
+require File.join(File.expand_path(File.dirname(__FILE__)), '../..', 'test_helper.rb')
+require 'rbbt/segment/transformed'
+require 'rbbt/segment/named_entity'
 require 'rexml/document'
 
-class TestClass < Test::Unit::TestCase
-  def test_sort
+class TestTransformed < Test::Unit::TestCase
+
+  def setup
+    @text = <<-EOF
+More recently, PPAR activators were shown to inhibit the activation of inflammatory response genes (such as IL-2, IL-6, IL-8, TNF alpha and metalloproteases) by negatively interfering with the NF-kappa B, STAT and AP-1 signalling pathways in cells of the vascular wall.
+    EOF
+
+    @entities = ["PPAR", "IL-2", "IL-6", "IL-8", "TNF alpha", "NF-kappa B", "AP-1", "STAT"].collect do |literal|
+      NamedEntity.setup(literal, :offset => @text.index(literal))
+    end
+  end
+
+  def test_transform
     text = <<-EOF
 More recently, PPAR activators were shown to inhibit the activation of inflammatory response genes (such as IL-2, IL-6, IL-8, TNF alpha and metalloproteases) by negatively interfering with the NF-kappa B, STAT and AP-1 signalling pathways in cells of the vascular wall.
     EOF
@@ -13,52 +24,25 @@ More recently, PPAR activators were shown to inhibit the activation of inflammat
       NamedEntity.setup(literal, :offset => text.index(literal))
     end
 
-    Transformed.with_transform(text, entities, Proc.new{|e| "[" + e.upcase + "]" }) do 
-      assert text.include? "such as [IL-2]"
-    end
+    Transformed.transform(text, entities, Proc.new{|e| "[" + e.upcase + "]" }) 
+    assert text.include? "such as [IL-2]"
   end
 
-  def ___test_transform
-    a = "This sentence mentions the TP53 gene and the CDK5 protein"
-    original = a.dup
-
-    gene1 = "TP53"
-    gene1.extend Segment
-    gene1.offset = a.index gene1
-
-    gene2 = "CDK5"
-    gene2.extend Segment
-    gene2.offset = a.index gene2
-
-    assert_equal gene1, a[gene1.range]
-    assert_equal gene2, a[gene2.range]
-    
-    c = a.dup
-
-    c[gene2.range] = "GN"
-    assert_equal c, Transformed.transform(a,[gene2], "GN")
-    c[gene1.range] = "GN"
-    assert_equal c, Transformed.transform(a,[gene1], "GN")
-
-    iii a.transformation_offset_differences
-    raise
-    assert_equal gene2.offset, a.transformation_offset_differences.first.first.first
-    assert_equal gene1.offset, a.transformation_offset_differences.last.first.first
-
-
-    gene3 = "GN gene"
-    gene3.extend Segment
-    gene3.offset = a.index gene3
-
-    assert_equal gene3, a[gene3.range]
+  def test_with_transform
+    text = <<-EOF
+More recently, PPAR activators were shown to inhibit the activation of inflammatory response genes (such as IL-2, IL-6, IL-8, TNF alpha and metalloproteases) by negatively interfering with the NF-kappa B, STAT and AP-1 signalling pathways in cells of the vascular wall.
+    EOF
 
-    a.restore([gene3])
-    assert_equal original, a
-    assert_equal "TP53 gene", a[gene3.range]
+    entities = ["PPAR", "IL-2", "IL-6", "IL-8", "TNF alpha", "NF-kappa B", "AP-1", "STAT"].reverse.collect do |literal|
+      NamedEntity.setup(literal, :offset => text.index(literal))
+    end
 
+    Transformed.with_transform(text, entities, Proc.new{|e| "[" + e.upcase + "]" }) do 
+      assert text.include? "such as [IL-2]"
+    end
   end
 
-  def test_with_transform
+  def test_with_transform_2
     a = "This sentence mentions the TP53 gene and the CDK5R1 protein"
     original = a.dup
 
@@ -117,18 +101,47 @@ More recently, PPAR activators were shown to inhibit the activation of inflammat
     assert_equal "CDK5R1 protein", exp2
   end
 
+  def test_with_transform_sentences
+    a = "This first sentence mentions Bread. This sentence mentions the TP53 gene and the CDK5R1 protein"
+    original = a.dup
+
+    gene1 = "TP53"
+    gene1.extend NamedEntity
+    gene1.offset = a.index gene1
+
+    gene2 = "CDK5R1"
+    gene2.extend NamedEntity
+    gene2.offset = a.index gene2
+
+    bread = "Bread"
+    bread.extend NamedEntity
+    bread.offset = a.index bread
+
+    sentences = Segment.align(a, a.split(". "))
+
+    Transformed.with_transform(sentences[1], [gene1, gene2, bread], "GN") do
+      assert sentences[1].include?("GN gene and the GN protein")
+    end
+
+    Transformed.with_transform(sentences[0], [gene1, gene2, bread], "BR") do
+      assert sentences[0].include?("first sentence mentions BR")
+    end
+
+
+  end
+
   def test_html
     a = "This sentence mentions the TP53 gene and the CDK5R1 protein"
 
     gene1 = "TP53"
     gene1.extend NamedEntity
     gene1.offset = a.index gene1
-    gene1.type = "Gene"
+    gene1.entity_type = "Gene"
 
     gene2 = "CDK5R1"
     gene2.extend NamedEntity
     gene2.offset = a.index gene2
-    gene2.type = "Protein"
+    gene2.entity_type = "Protein"
 
     Transformed.with_transform(a, [gene1,gene2], Proc.new{|e| e.html}) do 
       assert_equal "This sentence mentions the <span class='Entity' attr-entity-type='Gene'>TP53</span> gene and the <span class='Entity' attr-entity-type='Protein'>CDK5R1</span> protein", a
@@ -143,13 +156,13 @@ More recently, PPAR activators were shown to inhibit the activation of inflammat
     gene1.extend NamedEntity
     gene1.offset = a.index gene1 
     gene1.offset += 10
-    gene1.type = "Gene"
+    gene1.entity_type = "Gene"
 
     gene2 = "CDK5R1"
     gene2.extend NamedEntity
     gene2.offset = a.index gene2
     gene2.offset += 10
-    gene2.type = "Protein"
+    gene2.entity_type = "Protein"
 
     Transformed.with_transform(a, [gene1,gene2], Proc.new{|e| e.html}) do 
       assert_equal "This sentence mentions the <span class='Entity' attr-entity-type='Gene'>TP53</span> gene and the <span class='Entity' attr-entity-type='Protein'>CDK5R1</span> protein", a
@@ -162,12 +175,12 @@ More recently, PPAR activators were shown to inhibit the activation of inflammat
     gene1 = "TP53"
     gene1.extend NamedEntity
     gene1.offset = a.index gene1
-    gene1.type = "Gene"
+    gene1.entity_type = "Gene"
 
     gene2 = "TP53 gene"
     gene2.extend NamedEntity
     gene2.offset = a.index gene2
-    gene2.type = "Expanded Gene"
+    gene2.entity_type = "Expanded Gene"
 
     assert_equal [gene1], Segment.overlaps(Segment.sort([gene1,gene2]))
 
@@ -379,5 +392,46 @@ This is another sentence. Among the nonstructural proteins, the leader protein (
       end
     end
   end
+
+  def ___test_transform
+    a = "This sentence mentions the TP53 gene and the CDK5 protein"
+    original = a.dup
+
+    gene1 = "TP53"
+    gene1.extend Segment
+    gene1.offset = a.index gene1
+
+    gene2 = "CDK5"
+    gene2.extend Segment
+    gene2.offset = a.index gene2
+
+    assert_equal gene1, a[gene1.range]
+    assert_equal gene2, a[gene2.range]
+    
+    c = a.dup
+
+    c[gene2.range] = "GN"
+    assert_equal c, Transformed.transform(a,[gene2], "GN")
+    c[gene1.range] = "GN"
+    assert_equal c, Transformed.transform(a,[gene1], "GN")
+
+    iii a.transformation_offset_differences
+    raise
+    assert_equal gene2.offset, a.transformation_offset_differences.first.first.first
+    assert_equal gene1.offset, a.transformation_offset_differences.last.first.first
+
+
+    gene3 = "GN gene"
+    gene3.extend Segment
+    gene3.offset = a.index gene3
+
+    assert_equal gene3, a[gene3.range]
+
+    a.restore([gene3])
+    assert_equal original, a
+    assert_equal "TP53 gene", a[gene3.range]
+
+  end
+
 end
 

data/test/rbbt/test_document.rb

@@ -0,0 +1,14 @@
+require File.join(File.expand_path(File.dirname(__FILE__)), '..', 'test_helper.rb')
+require 'rbbt/document'
+
+class TestDocument < Test::Unit::TestCase
+
+  def test_docid
+    text = "This is a document"
+    Document.setup(text, "TEST", "test_doc1", nil)
+
+    assert_equal ["TEST", "test_doc1", nil, Misc.digest(text)] * ":", text.docid
+  end
+
+end
+

data/test/rbbt/test_segment.rb

@@ -0,0 +1,182 @@
+require File.join(File.expand_path(File.dirname(__FILE__)), '..', 'test_helper.rb')
+require 'rbbt/document'
+require 'rbbt/document/corpus'
+require 'rbbt/segment'
+
+class TestSegment < Test::Unit::TestCase
+  def test_segment
+    text = "This is a document"
+    Document.setup(text, "TEST", "test_doc1", nil)
+
+    segment = Segment.setup("is", :offset => text.index("is"), :docid => text.docid)
+
+    assert_equal text.docid + ":" + segment.offset.to_s + ".." + segment.eend.to_s,  segment.segid
+  end
+
+  def test_segid
+    text = "This is a document"
+    Document.setup(text, "TEST", "test_doc1", nil)
+
+    corpus = Document::Corpus.setup({})
+
+    corpus.add_document(text)
+
+    segment = Segment.setup("is", :offset => text.index("is"), :docid => text.docid)
+
+    segid = segment.segid(corpus)
+
+    segment = segid.segment
+    assert_equal "is", segment
+  end
+
+  def test_info
+    segment = "test"
+    segment.extend Segment
+    segment.offset = 10
+    assert segment.info.include? :offset
+  end
+
+  def test_sort
+    text = "This sentence mentions the TP53 gene and the CDK5R1 protein"
+    Document.setup(text, "TEST", "test_doc1", nil)
+
+    corpus = Document::Corpus.setup({})
+
+    corpus.add_document(text)
+
+    gene1 = "TP53"
+    gene1.extend Segment
+    gene1.offset = text.index gene1
+    gene1.docid = text.docid
+
+    gene2 = "CDK5R1"
+    gene2.extend Segment
+    gene2.offset = text.index gene2
+    gene2.docid = text.docid
+
+    assert_equal [gene1,gene2], Segment.sort([gene2,gene1])
+
+    assert_equal [gene1,gene2], Segment.sort([gene2.segid(corpus),gene1.segid(corpus)]).collect{|segid| segid.segment}
+  end
+
+  def test_clean_sort
+    text = "This sentence mentions the TP53 gene and the CDK5R1 protein"
+    Document.setup(text, "TEST", "test_doc1", nil)
+
+    corpus = Document::Corpus.setup({})
+
+    corpus.add_document(text)
+
+    gene1 = "TP53"
+    gene1.extend Segment
+    gene1.offset = text.index gene1
+    gene1.docid = text.docid
+
+    gene2 = "CDK5R1"
+    gene2.extend Segment
+    gene2.offset = text.index gene2
+    gene2.docid = text.docid
+
+    gene3 = "TP53 gene"
+    gene3.extend Segment
+    gene3.offset = text.index gene1
+    gene3.docid = text.docid
+
+    assert_equal [gene1,gene2], Segment.sort([gene2,gene1])
+
+    assert_equal [gene3,gene2], Segment.clean_sort([gene2,gene1,gene3])
+  end
+
+  def test_split
+    text = "This sentence mentions the TP53 gene and the CDK5R1 protein"
+    Document.setup(text, "TEST", "test_doc1", nil)
+
+    corpus = Document::Corpus.setup({})
+
+    corpus.add_document(text)
+
+    gene1 = "TP53"
+    gene1.extend Segment
+    gene1.offset = text.index gene1
+    gene1.docid = text.docid
+
+    gene2 = "CDK5R1"
+    gene2.extend Segment
+    gene2.offset = text.index gene2
+    gene2.docid = text.docid
+
+    gene3 = "TP53 gene"
+    gene3.extend Segment
+    gene3.offset = text.index gene1
+    gene3.docid = text.docid
+
+    assert_equal ["This sentence mentions the ", gene3, " and the ", gene2, " protein"], Segment.split(text, [gene2, gene1, gene3])
+
+    assert_equal ["This sentence mentions the ", gene3, " and the ", gene2, " protein"], Segment.split(text, [gene2, gene1, gene3].collect{|s| s.segid})
+  end
+
+
+  def test_align
+    text =<<-EOF
+Atypical teratoid/rhabdoid tumors (AT/RTs) are highly aggressive brain tumors of early childhood poorly responding to therapy.
+    EOF
+
+    parts = text.split(/\W/)
+    Segment.align(text, parts)
+
+    assert_equal "Atypical teratoid/".length, parts.select{|s| s == "rhabdoid"}.first.offset
+
+    Document.setup(text, "TEST", "test_doc1", nil)
+
+    parts = text.split(/\W/)
+    Segment.align(text, parts)
+
+    assert_equal parts.first.docid, text.docid
+  end
+
+  def test_segment_index
+    text = "This sentence mentions the TP53 gene and the CDK5R1 protein"
+    Document.setup(text, "TEST", "test_doc1", nil)
+
+    corpus = Document::Corpus.setup({})
+
+    corpus.add_document(text)
+
+    gene1 = "TP53"
+    gene1.extend Segment
+    gene1.offset = text.index gene1
+    gene1.docid = text.docid
+
+    gene2 = "CDK5R1"
+    gene2.extend Segment
+    gene2.offset = text.index gene2
+    gene2.docid = text.docid
+
+    gene3 = "TP53 gene"
+    gene3.extend Segment
+    gene3.offset = text.index gene1
+    gene3.docid = text.docid
+
+    index = Segment.index([gene1, gene2, gene3], corpus)
+    assert_equal "CDK5R1", index[gene2.offset + 1].segment.first
+
+    TmpFile.with_file do |fwt|
+      index = Segment.index([gene1, gene2, gene3], corpus, fwt)
+      assert_equal %w(CDK5R1), index[gene2.offset + 1].segment
+      index = Segment.index([gene1, gene2, gene3], corpus, fwt)
+      assert_equal %w(CDK5R1), index[gene2.offset + 1].segment
+    end
+
+    index = Segment.index([gene1, gene2, gene3].collect{|s| s.segid}, corpus)
+    assert_equal "CDK5R1", index[gene2.offset + 1].segment.first
+
+    TmpFile.with_file do |fwt|
+      index = Segment.index([gene1, gene2, gene3].collect{|s| s.segid}, corpus, fwt)
+      assert_equal %w(CDK5R1), index[gene2.offset + 1].segment
+      index = Segment.index([gene1, gene2, gene3].collect{|s| s.segid}, corpus, fwt)
+      assert_equal %w(CDK5R1), index[gene2.offset + 1].segment
+    end
+  end
+
+end
+