simple_inline_text_annotation 2.0.0 → 2.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8d8d14f48a850c835dbf62083dcda9ea3f2588e7f2544cacf3125e54481fca37
-  data.tar.gz: 50ebe7c23e5d841697f455586c9ed031d7df4d0da8e625630872948506145b2b
+  metadata.gz: 3c4a6b74cd54c5a3ef2263ee7e591cc1f5ef2acab8f36b991fdd3a7f84fc887c
+  data.tar.gz: a436a2add0c95d4484f096a4c56ca1dc283f5e54178377ebeb50f53c4cbe9ddb
 SHA512:
-  metadata.gz: 99e205823a643812b22d835d9b9eae0041e320430b07175ab95a8b87cfb6287d37c749b18eac48beada081c39bb89a3aa16179cbb930e1ef4b426ce5b4ed0334
-  data.tar.gz: ccd90e14e4b0247428bc271e1663c995126f7361c60c59dfa6a2805fe3d255ef82b73f02ca19ed4beefe0486f3075697a96e389a7577b454d29bdbf19bbe6488
+  metadata.gz: a34913353fd069aab8ad9e19955fcfaf50e8856d8e49c91034eb271c5eb9e156f3cf03371f4b053a9f4fc85d5f3d4e5fc62afaefbc7d5da0b70e93f72acee825
+  data.tar.gz: 4eddd143142df13d5b4653805bac281e9b2043fd3ec9ab4ac4d2bf1c104eb43ffa2b2f2491435a31e6a0f8ff0b1145ba29ba246df94d6aee7f93aea28fe20ab2

data/.rubocop.yml

@@ -1,5 +1,5 @@
 AllCops:
-  TargetRubyVersion: 3.1
+  TargetRubyVersion: 3.4.4
   NewCops: enable
 
 Style/StringLiterals:

data/.ruby-version

@@ -0,0 +1 @@
+3.4.4

data/lib/simple_inline_text_annotation/denotation_validator.rb

@@ -15,7 +15,11 @@ class SimpleInlineTextAnnotation
     private
 
     def remove_duplicates_from(denotations)
-      denotations.uniq(&:span)
+      # Deduplicate on the FULL identity (span + obj + id), not just span —
+      # two denotations sharing a span but pointing at different objs are the
+      # multi-label case (e.g. `[eye][UBERON_0000019|UBERON_0000955]`) and
+      # Generator pipe-joins them. Only truly identical entries are dropped.
+      denotations.uniq { |d| [d.span, d.obj, d.id] }
     end
 
     def remove_non_integer_positions_from(denotations)
@@ -40,12 +44,19 @@ class SimpleInlineTextAnnotation
       result = []
 
       sorted_denotations.each do |denotation|
-        result << denotation unless result.any? { |outer| denotation.nested_within?(outer) }
+        # Preserve denotations that share an EXACT span with an already-accepted
+        # one (the multi-label case — Generator will pipe-join their labels).
+        # Only drop those STRICTLY nested inside a larger outer span.
+        result << denotation unless result.any? { |outer| strictly_nested?(denotation, outer) }
       end
 
       result
     end
 
+    def strictly_nested?(inner, outer)
+      inner.nested_within?(outer) && (inner.begin_pos != outer.begin_pos || inner.end_pos != outer.end_pos)
+    end
+
     def remove_boundary_crosses_from(denotations)
       denotations.reject do |denotation|
         denotations.any? { |existing| denotation.boundary_crossing?(existing) }

data/lib/simple_inline_text_annotation/generator.rb

@@ -34,27 +34,46 @@ class SimpleInlineTextAnnotation
     end
 
     def annotate_text(text, denotations, relations)
+      # Group denotations by span position so multiple denotations on the
+      # SAME span emit ONE annotation with a pipe-joined label
+      # (e.g. `[eye][UBERON_0000019|UBERON_0000955]`) — the SIAF spec's
+      # extension for multi-labelled spans. Individual URL resolution via
+      # `entity types` config still applies to each label independently, so
+      # the tail reference block lists every underlying URL.
+      grouped = denotations.group_by { |d| [d.begin_pos, d.end_pos] }
+
       # Annotate text from the end to ensure position calculation.
-      denotations.sort_by(&:begin_pos).reverse_each do |denotation|
-        text = annotate_text_with_denotation(text, denotation, relations)
+      grouped.sort_by { |(b, _e), _| b }.reverse_each do |(_b, _e), ds|
+        text = annotate_text_with_denotations(text, ds, relations)
       end
 
       text
     end
 
-    def annotate_text_with_denotation(text, denotation, relations)
-      begin_pos = denotation.begin_pos
-      end_pos = denotation.end_pos
-      annotation = if denotation.id && !denotation.id.empty?
-                     get_annotations(denotation, relations)
-                   else
-                     get_obj(denotation.obj)
-                   end
+    def annotate_text_with_denotations(text, denotations, relations)
+      begin_pos = denotations.first.begin_pos
+      end_pos   = denotations.first.end_pos
+      composite = compose_label(denotations, relations)
 
-      annotated_text = "[#{text[begin_pos...end_pos]}][#{annotation}]"
+      annotated_text = "[#{text[begin_pos...end_pos]}][#{composite}]"
       text[0...begin_pos] + annotated_text + text[end_pos..]
     end
 
+    # Compose each denotation's label independently (honors get_obj URL →
+    # short-label lookup and per-denotation relation composition), then
+    # pipe-join with dedupe to preserve insertion order.
+    def compose_label(denotations, relations)
+      denotations.map { |d| label_for(d, relations) }.uniq.join("|")
+    end
+
+    def label_for(denotation, relations)
+      if denotation.id && !denotation.id.empty?
+        get_annotations(denotation, relations)
+      else
+        get_obj(denotation.obj)
+      end
+    end
+
     def labeled_entity_types
       return nil unless @config
 

data/lib/simple_inline_text_annotation/parser.rb

@@ -2,6 +2,7 @@
 
 require_relative "entity_type_collection"
 require_relative "denotation"
+require_relative "relation_without_denotation_error"
 
 class SimpleInlineTextAnnotation
   class Parser
@@ -21,12 +22,10 @@ class SimpleInlineTextAnnotation
 
       process_annotations(full_text)
 
-      SimpleInlineTextAnnotation.new(
-        full_text,
-        @denotations,
-        @relations,
-        @entity_type_collection
-      )
+      SimpleInlineTextAnnotation.new full_text,
+                                     @denotations,
+                                     @relations,
+                                     @entity_type_collection
     end
 
     private
@@ -58,6 +57,7 @@ class SimpleInlineTextAnnotation
       return match.end(0) unless process_annotation(match[2], begin_pos, end_pos)
 
       full_text[match.begin(0)...match.end(0)] = target_text
+
       end_pos
     end
 

data/lib/simple_inline_text_annotation/relation_without_denotation_error.rb

@@ -0,0 +1,5 @@
+# frozen_string_literal: true
+
+class SimpleInlineTextAnnotation
+  class RelationWithoutDenotationError < StandardError; end
+end

data/lib/simple_inline_text_annotation/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 class SimpleInlineTextAnnotation
-  VERSION = "2.0.0"
+  VERSION = "2.2.0"
 end

data/lib/simple_inline_text_annotation.rb

@@ -25,6 +25,7 @@ class SimpleInlineTextAnnotation
     @denotations = denotations
     @relations = relations
     @entity_type_collection = entity_type_collection
+    check_denotations_and_relations
   end
 
   def self.parse(source)
@@ -70,4 +71,19 @@ class SimpleInlineTextAnnotation
 
     { "entity types" => @entity_type_collection.to_config }
   end
+
+  def check_denotations_and_relations
+    @relations.each do |relation|
+      unless referable_to?(relation, @denotations)
+        raise SimpleInlineTextAnnotation::RelationWithoutDenotationError,
+              "Relation #{relation.inspect} refers to missing denotation."
+      end
+    end
+  end
+
+  def referable_to?(relation, denotations)
+    denotation_ids = denotations.map(&:id)
+
+    denotation_ids.include?(relation["subj"]) && denotation_ids.include?(relation["obj"])
+  end
 end

metadata

@@ -1,14 +1,13 @@
 --- !ruby/object:Gem::Specification
 name: simple_inline_text_annotation
 version: !ruby/object:Gem::Version
-  version: 2.0.0
+  version: 2.2.0
 platform: ruby
 authors:
 - xaiBUh29wX
-autorequire:
 bindir: exe
 cert_chain: []
-date: 2025-07-30 00:00:00.000000000 Z
+date: 1980-01-02 00:00:00.000000000 Z
 dependencies: []
 description: This gem provides inline text annotation functionality, extracted from
   PubAnnotation, with support for denotations, entity types, and nested spans.
@@ -20,6 +19,7 @@ extra_rdoc_files: []
 files:
 - ".rspec"
 - ".rubocop.yml"
+- ".ruby-version"
 - CHANGELOG.md
 - CODE_OF_CONDUCT.md
 - LICENSE.txt
@@ -33,6 +33,7 @@ files:
 - lib/simple_inline_text_annotation/generator_error.rb
 - lib/simple_inline_text_annotation/parser.rb
 - lib/simple_inline_text_annotation/relation_validator.rb
+- lib/simple_inline_text_annotation/relation_without_denotation_error.rb
 - lib/simple_inline_text_annotation/version.rb
 - sig/simple_inline_text_annotation.rbs
 homepage: https://github.com/pubannotation/simple_inline_annotation_format
@@ -42,7 +43,6 @@ metadata:
   homepage_uri: https://github.com/pubannotation/simple_inline_annotation_format
   changelog_uri: https://github.com/pubannotation/simple_inline_annotation_format/blob/master/ruby/CHANGELOG.md
   rubygems_uri: https://rubygems.org/gems/simple_inline_text_annotation
-post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -50,15 +50,14 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: 3.1.0
+      version: 3.4.4
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.4.20
-signing_key:
+rubygems_version: 3.6.7
 specification_version: 4
 summary: A Ruby gem for inline text annotation with denotations and entity types.
 test_files: []