canon 0.2.9 → 0.2.12

data/lib/canon/diff/diff_node_enricher.rb

@@ -1,9 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "diff_char_range"
-require_relative "text_decomposer"
-require_relative "source_locator"
-
 module Canon
   module Diff
     # Enriches DiffNodes with character position data (DiffCharRanges).

data/lib/canon/diff/diff_node_mapper.rb

@@ -1,6 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "diff_line"
 require "set"
 
 module Canon
@@ -159,8 +158,6 @@ module Canon
       # @param diff_lines [Array<DiffLine>] The diff lines to update
       # @param lcs_diffs [Array<Diff::LCS::Change>] The LCS changes
       def apply_block_formatting!(diff_lines, lcs_diffs)
-        require_relative "formatting_detector"
-
         blocks = group_change_blocks(lcs_diffs)
 
         blocks.each do |block|
@@ -265,7 +262,6 @@ module Canon
       # @param line2 [String] Second line
       # @return [Boolean] true if formatting-only difference
       def formatting_only_line?(line1, line2)
-        require_relative "formatting_detector"
         FormattingDetector.formatting_only?(line1, line2)
       end
 
@@ -323,7 +319,7 @@ module Canon
       # @param line2 [String] New line content (for +/!)
       # @return [Boolean] true if formatting-only
       def formatting?(node, line1, line2)
-        return true if node.respond_to?(:formatting?) && node.formatting?
+        return true if node.is_a?(Canon::Diff::DiffNode) && node.formatting?
         return false if node
         return true if comment_only_line?(line1) || comment_only_line?(line2)
 
@@ -394,14 +390,14 @@ module Canon
                            end
 
           nodes_to_check.any? do |node|
-            # Check if the node itself has the matching name
-            if node.respond_to?(:name) && node.name == line_element_name
-              true
-            # Check if the node's parent has the matching name (for TextNode diffs)
-            elsif node.respond_to?(:parent) && node.parent.respond_to?(:name) && node.parent.name == line_element_name # rubocop:disable Style/IfWithBooleanLiteralBranches
+            next false unless node
+
+            node_name = Canon::Comparison::NodeInspector.name(node)
+            if node_name == line_element_name
               true
             else
-              false
+              parent = Canon::Comparison::NodeInspector.parent(node)
+              parent && Canon::Comparison::NodeInspector.name(parent) == line_element_name
             end
           end
         end
@@ -453,6 +449,8 @@ module Canon
         comment_lines
       end
 
+      public :build_comment_lines
+
       # Find a comment DiffNode for a line that falls within a comment range.
       # Matches by checking if the DiffNode's source node has name "comment".
       #
@@ -463,7 +461,7 @@ module Canon
         @comment_diff_nodes&.find do |diff_node|
           nodes_to_check = [diff_node.node1, diff_node.node2].compact
           nodes_to_check.any? do |node|
-            node.respond_to?(:name) && node.name == "comment"
+            Canon::Comparison::NodeInspector.name(node) == "comment"
           end
         end
       end

data/lib/canon/diff/diff_report_builder.rb

@@ -1,9 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "diff_report"
-require_relative "diff_block_builder"
-require_relative "diff_context_builder"
-
 module Canon
   module Diff
     # Builds a complete DiffReport from DiffLines

data/lib/canon/diff/formatting_detector.rb

@@ -99,7 +99,6 @@ module Canon
       def self.decode_xml_entities(text)
         return text unless text.include?("&")
 
-        require_relative "../tree_diff/core/xml_entity_decoder"
         Canon::TreeDiff::Core::XmlEntityDecoder.decode_xml_entities(text)
       end
 
@@ -297,8 +296,9 @@ module Canon
         [close_idx + 1, "<#{sorted}#{suffix}"]
       end
 
-      private_class_method :normalize_for_comparison, :blank?,
-                           :decode_xml_entities, :normalize_attribute_order,
+      public_class_method :normalize_for_comparison, :blank?
+
+      private_class_method :decode_xml_entities, :normalize_attribute_order,
                            :sort_tag_attributes, :tokenize_tag_content,
                            :process_tag, :process_processing_instruction,
                            :process_comment, :process_regular_tag

data/lib/canon/diff/node_serializer.rb

@@ -1,12 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../xml/data_model"
-require_relative "../xml/nodes/text_node"
-require_relative "../xml/nodes/comment_node"
-require_relative "../xml/nodes/element_node"
-require_relative "../xml/nodes/processing_instruction_node"
-require_relative "../xml/nodes/root_node"
-
 module Canon
   module Diff
     # Serializes nodes from different parsing libraries into canonical strings

data/lib/canon/diff/xml_serialization_formatter.rb

@@ -57,9 +57,6 @@ module Canon
       rescue StandardError
         nil
       end
-
-      private_class_method :blank?, :text_node?, :extract_text_content,
-                           :empty_text_content_serialization_diff?
     end
   end
 end

data/lib/canon/diff.rb

@@ -0,0 +1,39 @@
+# frozen_string_literal: true
+
+module Canon
+  # Diff representation and pipeline construction.
+  #
+  # This namespace holds the structural artifacts produced by the diff
+  # pipeline — DiffNode, DiffLine, DiffBlock, DiffContext, DiffReport —
+  # plus the builders that assemble them (DiffNodeMapper, DiffLineBuilder,
+  # DiffBlockBuilder, DiffContextBuilder, DiffReportBuilder) and the
+  # supporting services (PathBuilder, NodeSerializer, DiffClassifier,
+  # FormattingDetector, SourceLocator, TextDecomposer, DiffCharRange,
+  # DiffNodeEnricher, XmlSerializationFormatter).
+  #
+  # All children are autoloaded from this file — never `require_relative`
+  # them from sibling files. Reference the constant and let autoload
+  # resolve it on first use.
+  module Diff
+    autoload :DiffBlock, "canon/diff/diff_block"
+    autoload :DiffBlockBuilder, "canon/diff/diff_block_builder"
+    autoload :DiffCharRange, "canon/diff/diff_char_range"
+    autoload :DiffClassifier, "canon/diff/diff_classifier"
+    autoload :DiffContext, "canon/diff/diff_context"
+    autoload :DiffContextBuilder, "canon/diff/diff_context_builder"
+    autoload :DiffLine, "canon/diff/diff_line"
+    autoload :DiffLineBuilder, "canon/diff/diff_line_builder"
+    autoload :DiffNode, "canon/diff/diff_node"
+    autoload :DiffNodeEnricher, "canon/diff/diff_node_enricher"
+    autoload :DiffNodeMapper, "canon/diff/diff_node_mapper"
+    autoload :DiffReport, "canon/diff/diff_report"
+    autoload :DiffReportBuilder, "canon/diff/diff_report_builder"
+    autoload :FormattingDetector, "canon/diff/formatting_detector"
+    autoload :NodeSerializer, "canon/diff/node_serializer"
+    autoload :PathBuilder, "canon/diff/path_builder"
+    autoload :SourceLocator, "canon/diff/source_locator"
+    autoload :TextDecomposer, "canon/diff/text_decomposer"
+    autoload :XmlSerializationFormatter,
+             "canon/diff/xml_serialization_formatter"
+  end
+end

data/lib/canon/diff_formatter/by_line/base_formatter.rb

@@ -2,8 +2,6 @@
 
 require "diff/lcs" unless RUBY_ENGINE == "opal"
 require "diff/lcs/hunk" unless RUBY_ENGINE == "opal"
-require_relative "../debug_output"
-require_relative "../theme"
 
 module Canon
   class DiffFormatter
@@ -22,10 +20,8 @@ module Canon
         def self.for_format(format, **options)
           case format
           when :xml
-            require_relative "xml_formatter"
             XmlFormatter.new(**options)
           when :html, :html4, :html5
-            require_relative "html_formatter"
             # Determine HTML version from format
             version = case format
                       when :html5 then :html5
@@ -34,13 +30,10 @@ module Canon
                       end
             HtmlFormatter.new(html_version: version, **options)
           when :json
-            require_relative "json_formatter"
             JsonFormatter.new(**options)
           when :yaml
-            require_relative "yaml_formatter"
             YamlFormatter.new(**options)
           else
-            require_relative "simple_formatter"
             SimpleFormatter.new(**options)
           end
         end
@@ -279,7 +272,7 @@ module Canon
 
           differences.select do |diff|
             # Handle both DiffNode objects and legacy Hash format
-            is_normative = if diff.respond_to?(:normative?)
+            is_normative = if diff.is_a?(Canon::Diff::DiffNode)
                              diff.normative?
                            elsif diff.is_a?(Hash) && diff.key?(:normative)
                              diff[:normative]
@@ -394,8 +387,6 @@ module Canon
         # @param diffs [Array] LCS diff array
         # @return [Array<Canon::Diff::DiffBlock>] Array of diff blocks
         def identify_diff_blocks(diffs)
-          require_relative "../../diff/diff_block"
-
           blocks = []
           current_start = nil
           current_types = []
@@ -467,8 +458,6 @@ module Canon
         # @return [Array<Canon::Diff::DiffContext>] Array of diff contexts
         def expand_contexts_with_context_lines(contexts, context_lines,
                                                 total_lines)
-          require_relative "../../diff/diff_context"
-
           contexts.map do |context|
             first_block = context.first
             last_block = context.last
@@ -736,10 +725,8 @@ module Canon
             next unless match.status == :matched
 
             [match.elem1, match.elem2].compact.each do |elem|
-              next unless elem.respond_to?(:children)
-
               elem.children.each do |child|
-                children.add(child) if child.respond_to?(:name)
+                children.add(child) if Canon::Comparison::NodeInspector.element_node?(child)
               end
             end
           end
@@ -778,8 +765,8 @@ module Canon
           return true if elements_with_semantic_diffs.include?(element)
 
           # Check all descendants
-          if element.respond_to?(:children)
-            element.children.any? do |child|
+          if Canon::Comparison::NodeInspector.element_node?(element)
+            Canon::Comparison::NodeInspector.children(element).any? do |child|
               has_semantic_diff_in_subtree?(child, elements_with_semantic_diffs)
             end
           else

data/lib/canon/diff_formatter/by_line/html_formatter.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "base_formatter"
-require_relative "../legend"
 require "set"
 
 module Canon
@@ -53,11 +51,6 @@ module Canon
             return ""
           end
 
-          require_relative "../../html/data_model"
-          require_relative "../../xml/element_matcher"
-          require_relative "../../xml/line_range_mapper"
-          require_relative "../../pretty_printer/html"
-
           output = []
 
           begin
@@ -107,7 +100,6 @@ module Canon
             end
 
             output << ""
-            require_relative "simple_formatter"
             simple = SimpleFormatter.new(
               use_color: @use_color,
               context_lines: @context_lines,
@@ -122,9 +114,6 @@ module Canon
 
         # Format using new DiffReportBuilder pipeline
         def format_with_pipeline(doc1, doc2)
-          require_relative "../../diff/diff_node_mapper"
-          require_relative "../../diff/diff_report_builder"
-
           # Layer 2: Map DiffNodes to DiffLines
           diff_lines = Canon::Diff::DiffNodeMapper.map(@differences, doc1, doc2)
 
@@ -350,15 +339,14 @@ module Canon
 
           # Second pass: skip children of elements with diffs
           elements_with_diffs.each do |elem|
-            if elem.respond_to?(:parent)
-              current = elem.parent
-              while current
-                if current.respond_to?(:name) && elements_with_diffs.include?(current)
-                  elements_to_skip.add(elem)
-                  break
-                end
-                current = current.respond_to?(:parent) ? current.parent : nil
+            current = Canon::Comparison::NodeInspector.parent(elem)
+            while current
+              if Canon::Comparison::NodeInspector.element_node?(current) &&
+                  elements_with_diffs.include?(current)
+                elements_to_skip.add(elem)
+                break
               end
+              current = Canon::Comparison::NodeInspector.parent(current)
             end
           end
 

data/lib/canon/diff_formatter/by_line/json_formatter.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "base_formatter"
-require_relative "../legend"
 require "strscan"
 
 module Canon
@@ -38,7 +36,6 @@ module Canon
             output << colorize(
               "Warning: JSON parsing failed (#{e.message}), using simple diff", :yellow
             )
-            require_relative "simple_formatter"
             simple = SimpleFormatter.new(
               use_color: @use_color,
               context_lines: @context_lines,

data/lib/canon/diff_formatter/by_line/simple_formatter.rb

@@ -1,8 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "base_formatter"
-require_relative "../legend"
-
 module Canon
   class DiffFormatter
     module ByLine

data/lib/canon/diff_formatter/by_line/xml_formatter.rb

@@ -1,8 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "base_formatter"
-require_relative "../legend"
-require_relative "../../tree_diff/core/xml_entity_decoder"
 require "set"
 require "strscan"
 
@@ -44,10 +41,6 @@ module Canon
             return ""
           end
 
-          require_relative "../../diff/diff_node_enricher"
-          require_relative "../../diff/diff_line_builder"
-          require_relative "../../diff/diff_report_builder"
-
           # Compute line number width BEFORE formatting
           compute_line_num_width(doc1, doc2)
 
@@ -201,10 +194,6 @@ module Canon
             return ""
           end
 
-          require_relative "../../xml/data_model"
-          require_relative "../../xml/element_matcher"
-          require_relative "../../xml/line_range_mapper"
-
           output = []
 
           begin
@@ -247,7 +236,6 @@ module Canon
             end
 
             output << ""
-            require_relative "simple_formatter"
             simple = SimpleFormatter.new(
               use_color: @use_color,
               context_lines: @context_lines,
@@ -622,15 +610,14 @@ module Canon
 
           # Second pass: skip children of elements with diffs
           elements_with_diffs.each do |elem|
-            if elem.respond_to?(:parent)
-              current = elem.parent
-              while current
-                if current.respond_to?(:name) && elements_with_diffs.include?(current)
-                  elements_to_skip.add(elem)
-                  break
-                end
-                current = current.respond_to?(:parent) ? current.parent : nil
+            current = Canon::Comparison::NodeInspector.parent(elem)
+            while current
+              if Canon::Comparison::NodeInspector.element_node?(current) &&
+                  elements_with_diffs.include?(current)
+                elements_to_skip.add(elem)
+                break
               end
+              current = Canon::Comparison::NodeInspector.parent(current)
             end
           end
 
@@ -721,8 +708,6 @@ module Canon
 
         # Identify contiguous diff blocks
         def identify_diff_blocks(diffs)
-          require_relative "../../diff/diff_block"
-
           blocks = []
           current_start = nil
           current_types = []
@@ -784,8 +769,6 @@ module Canon
         # Expand contexts with context lines
         def expand_contexts_with_context_lines(contexts, context_lines,
                                                 total_lines)
-          require_relative "../../diff/diff_context"
-
           contexts.map do |context|
             first_block = context.first
             last_block = context.last
@@ -803,8 +786,6 @@ module Canon
 
         # Format a context
         def format_context(context, diffs, base_line1, base_line2)
-          require_relative "../../diff/formatting_detector"
-
           # Pre-compute block-level formatting for multi-line changes
           formatting_indices = detect_block_formatting(context, diffs)
 

data/lib/canon/diff_formatter/by_line/yaml_formatter.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "base_formatter"
-require_relative "../legend"
 require "strscan"
 
 module Canon
@@ -37,7 +35,6 @@ module Canon
             output << colorize(
               "Warning: YAML parsing failed (#{e.message}), using simple diff", :yellow
             )
-            require_relative "simple_formatter"
             simple = SimpleFormatter.new(
               use_color: @use_color,
               context_lines: @context_lines,

data/lib/canon/diff_formatter/by_object/base_formatter.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require "set"
-require_relative "../theme"
 
 module Canon
   class DiffFormatter
@@ -76,19 +75,16 @@ show_diffs: :all, theme: nil)
 show_diffs: :all, theme: nil)
           case format
           when :xml, :html
-            require_relative "xml_formatter"
             XmlFormatter.new(use_color: use_color,
                              visualization_map: visualization_map,
                              show_diffs: show_diffs,
                              theme: theme)
           when :json
-            require_relative "json_formatter"
             JsonFormatter.new(use_color: use_color,
                               visualization_map: visualization_map,
                               show_diffs: show_diffs,
                               theme: theme)
           when :yaml
-            require_relative "yaml_formatter"
             YamlFormatter.new(use_color: use_color,
                               visualization_map: visualization_map,
                               show_diffs: show_diffs,
@@ -221,23 +217,20 @@ show_diffs: :all, theme: nil)
           current = node
           visited = Set.new
 
-          while current.respond_to?(:name)
-            # Prevent infinite loops by tracking visited nodes
+          while current
             break if visited.include?(current.object_id)
 
             visited << current.object_id
 
-            parts.unshift(current.name) if current.name
+            break if Canon::Comparison::NodeInspector.document?(current) ||
+              Canon::Comparison::NodeInspector.document_fragment?(current)
+
+            name = Canon::Comparison::NodeInspector.name(current)
+            break if name.nil?
 
-            # Stop at document or fragment roots
-            break if current.is_a?(Nokogiri::XML::Document) ||
-              current.is_a?(Nokogiri::HTML4::Document) ||
-              current.is_a?(Nokogiri::HTML5::Document) ||
-              current.is_a?(Nokogiri::XML::DocumentFragment) ||
-              current.is_a?(Nokogiri::HTML4::DocumentFragment) ||
-              current.is_a?(Nokogiri::HTML5::DocumentFragment)
+            parts.unshift(name) unless name.empty?
 
-            current = current.parent if current.respond_to?(:parent)
+            current = Canon::Comparison::NodeInspector.parent(current)
           end
 
           parts.empty? ? diff_node.dimension.to_s : parts.join(".")
@@ -270,9 +263,12 @@ show_diffs: :all, theme: nil)
 
             if diffs && !diffs.empty?
               # Render all differences at this path
+              has_children = value.is_a?(Hash) &&
+                (value.keys - [:__diffs__]).any?
+
               diffs.each_with_index do |diff, diff_idx|
-                # Use proper connector for each diff
-                current_connector = if diff_idx == diffs.length - 1
+                is_last_diff = diff_idx == diffs.length - 1
+                current_connector = if is_last_diff && !has_children
                                       connector
                                     else
                                       is_last_item ? "├── " : "├── "
@@ -282,6 +278,13 @@ show_diffs: :all, theme: nil)
                 output << line
                 @line_count += line.count("\n") + 1
               end
+
+              # Recurse into child diffs at this path
+              if has_children
+                subtree = render_tree(value, prefix: prefix + continuation,
+                                             is_last: is_last_item)
+                output << subtree
+              end
             else
               # Render intermediate path
               line = colorize("#{prefix}#{connector}#{key}:",

data/lib/canon/diff_formatter/by_object/json_formatter.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "base_formatter"
-
 module Canon
   class DiffFormatter
     module ByObject