canon 0.3.58 → 0.3.60

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: c1b3bba3210fd3ca95a66a4a3e439f6b9e1a0f95b30ddf4d90f4069a2b5a6d8b
-  data.tar.gz: fadf6bdf8fb892fbb3998d53754f599b17fac80b099936d77a7dae0881e4f5c5
+  metadata.gz: f1cbc73f1d667885f1f989a9534d8123dbcc67ae7eda250cc24cba00e7ad24a7
+  data.tar.gz: b9a27d00c02f533c6679237e190eb7f6bc6532fe0f945957bdc5d521bb84217b
 SHA512:
-  metadata.gz: e551b02b98c802c3d2593c1f684c89d271b223892fb32dbd20f186fe6cb13e06d18bd079839298325eb1d8a798b962876af9c95177763268ae291fc503828552
-  data.tar.gz: 8485c2e7e3049386ff977fca9c4162b44d439a2c218067f3aefe4a3b596d0f351b60c0c18a7f9fd0d5bcbf77192a0525c84dac7c1b35fe7f92369737f6dc2308
+  metadata.gz: b060895cd58c7321db7d0640a02ca166f3efb9b5cb4c90d5fafb7d615b3faf6c73f4e29d42533d56ac9fd393e625bac473d9cd122273dc81b3598b0be76b86f9
+  data.tar.gz: 4b160273200d229151e61842bc994cd8e1d23f495107f495edde383b2fdccda2e9de150a65160d34c5e3177a7bf236282a6e1bb675c72b2bf667deb200e3dab4

data/lib/canon/tree_diff/adapters/html_adapter.rb

@@ -196,7 +196,7 @@ module Canon
           return false unless Canon::Comparison::NodeInspector.element_node?(element)
 
           # List of HTML elements where whitespace is semantically significant
-          whitespace_sensitive_tags = %w[pre code textarea script style]
+          whitespace_sensitive_tags = Core::TreeNode::WHITESPACE_SENSITIVE_TAGS
           whitespace_sensitive_tags.include?(element.name.downcase)
         end
 

data/lib/canon/tree_diff/core/matching.rb

@@ -215,9 +215,10 @@ module Canon
 
           # Check prefix closure constraint
           # If ancestors are matched, they must be matched to each other
-          node1.ancestors.each_with_index do |anc1, idx|
+          ancestors1 = node1.ancestors
+          anc2_ancestors = node2.ancestors
+          ancestors1.each_with_index do |anc1, idx|
             # Get corresponding ancestor in tree2
-            anc2_ancestors = node2.ancestors
             return false if idx >= anc2_ancestors.size
 
             anc2 = anc2_ancestors[idx]

data/lib/canon/tree_diff/core/node_signature.rb

@@ -81,34 +81,7 @@ module Canon
         #
         # @return [Array<String>] Path components
         def compute_path
-          # Build path from root to node
-          ancestors = @node.ancestors.reverse
-          components = ancestors.map do |ancestor|
-            path_component(ancestor)
-          end
-
-          # Add the node itself
-          components << path_component(@node)
-
-          components
-        end
-
-        # Get path component for a node
-        #
-        # @param node [TreeNode] Node to get component for
-        # @return [String]
-        def path_component(node)
-          # For element nodes: use label with sorted attributes
-          # For text nodes: use "#text"
-          # CRITICAL: Text nodes should use "#text" not "text"
-          # Check the label - actual text nodes have no label or label == "text"
-          label_str = node.label.to_s.downcase
-          if node.label.nil? || label_str.empty? || label_str == "#text" || label_str == "text"
-            "#text"
-          else
-            # Memoized on the node — see TreeNode#signature_component.
-            node.signature_component(include_attributes: @include_attributes)
-          end
+          @node.signature_path(include_attributes: @include_attributes)
         end
 
         # Compute signature string from path

data/lib/canon/tree_diff/core/tree_node.rb

@@ -66,6 +66,20 @@ module Canon
           end
         end
 
+        # Signature path components from root to this node, built by
+        # parent chaining (see NodeSignature). Memoized per variant —
+        # signatures previously re-walked the ancestor chain per node.
+        #
+        # @param include_attributes [Boolean]
+        # @return [Array<String>]
+        def signature_path(include_attributes: true)
+          if include_attributes
+            @signature_path_attrs ||= build_signature_path(true)
+          else
+            @signature_path ||= build_signature_path(false)
+          end
+        end
+
         # Check if this is a leaf node (no children)
         #
         # @return [Boolean]
@@ -73,17 +87,40 @@ module Canon
           children.empty?
         end
 
-        # Elements where whitespace is semantically significant; their
-        # text value participates in the signature (see NodeSignature).
+        # Elements where whitespace is semantically significant (HTML):
+        # their text value participates in signatures and matching.
+        # Single source for the tree_diff components.
         WHITESPACE_SENSITIVE_TAGS = %w[pre code textarea script style].freeze
 
+        def build_signature_path(include_attributes)
+          component = signature_component(include_attributes: include_attributes)
+          if parent
+            parent.signature_path(include_attributes: include_attributes) + [component]
+          else
+            [component]
+          end
+        end
+
         def build_signature_component(include_attributes)
+          label_str = label.to_s.downcase
+          if label.nil? || label_str.empty? || label_str == "#text" || label_str == "text"
+            return "#text"
+          end
+
           component = label.to_s
 
           # Sorted attributes distinguish same-label nodes while
-          # ignoring attribute order.
+          # ignoring attribute order. Built directly into the
+          # component — no intermediate sorted/map/join arrays.
           if include_attributes && !attributes.empty?
-            component += "{#{attributes.sort.map { |k, v| "#{k}=#{v}" }.join(',')}}" # rubocop:disable Style/StringConcatenation
+            component += "{"
+            first = true
+            attributes.sort.each do |k, v|
+              component << "," unless first
+              first = false
+              component << k.to_s << "=" << v.to_s
+            end
+            component << "}"
           end
 
           if include_attributes &&
@@ -95,7 +132,7 @@ module Canon
 
           component
         end
-        private :build_signature_component
+        private :build_signature_component, :build_signature_path
 
         # Check if this is a text node (leaf with value)
         #
@@ -137,10 +174,14 @@ module Canon
         #
         # @return [Array<TreeNode>]
         def descendants
+          # Iterative pre-order walk — the recursive form allocated an
+          # intermediate array per level.
           result = []
-          children.each do |child|
-            result << child
-            result.concat(child.descendants)
+          stack = children.reverse_each.to_a
+          until stack.empty?
+            node = stack.pop
+            result << node
+            stack.concat(node.children.reverse_each.to_a) if node.children.any?
           end
           result
         end

data/lib/canon/tree_diff/matchers/hash_matcher.rb

@@ -118,11 +118,19 @@ module Canon
 
         def subtrees_match?(node1, node2)
           return false unless nodes_match?(node1, node2)
-          return false unless node1.children.size == node2.children.size
 
-          node1.children.zip(node2.children).all? do |child1, child2|
-            subtrees_match?(child1, child2)
+          children1 = node1.children
+          children2 = node2.children
+          return false unless children1.size == children2.size
+
+          index = 0
+          while index < children1.size
+            return false unless subtrees_match?(children1[index],
+                                                children2[index])
+
+            index += 1
           end
+          true
         end
 
         def nodes_match?(node1, node2)
@@ -142,6 +150,11 @@ module Canon
           return true if (text1.nil? || text1.empty?) && (text2.nil? || text2.empty?)
           return false if (text1.nil? || text1.empty?) || (text2.nil? || text2.empty?)
 
+          # Identical raw text is equivalent under every mode — skips
+          # the gsub normalization for the overwhelmingly common
+          # matched-leaf case.
+          return true if text1 == text2
+
           norm1 = normalize_text(text1)
           norm2 = normalize_text(text2)
           return true if norm1.empty? && norm2.empty?

data/lib/canon/tree_diff/matchers/similarity_matcher.rb

@@ -158,7 +158,7 @@ module Canon
           return false unless node
 
           # List of HTML elements where whitespace is semantically significant
-          whitespace_sensitive_tags = %w[pre code textarea script style]
+          whitespace_sensitive_tags = Core::TreeNode::WHITESPACE_SENSITIVE_TAGS
 
           # Check if this node is whitespace-sensitive
           if node.is_a?(Core::TreeNode)

data/lib/canon/tree_diff/matchers/structural_propagator.rb

@@ -211,7 +211,7 @@ module Canon
           return false unless node
 
           # List of HTML elements where whitespace is semantically significant
-          whitespace_sensitive_tags = %w[pre code textarea script style]
+          whitespace_sensitive_tags = Core::TreeNode::WHITESPACE_SENSITIVE_TAGS
 
           # Check if this node is whitespace-sensitive
           if node.is_a?(Core::TreeNode)

data/lib/canon/tree_diff/operations/operation_detector.rb

@@ -205,8 +205,10 @@ module Canon
           attrs1 = node1.attributes
           attrs2 = node2.attributes
 
-          # Check if attribute values differ (ignoring order)
-          if attrs1.sort.to_h != attrs2.sort.to_h
+          # Check if attribute values differ (ignoring order).
+          # Identical raw hashes (order included) are trivially equal
+          # after sorting — the overwhelmingly common matched case.
+          if attrs1 != attrs2 && attrs1.sort.to_h != attrs2.sort.to_h
             # Actual attribute value differences
             changes[:attributes] = {
               old: attrs1,
@@ -611,15 +613,12 @@ module Canon
         def whitespace_sensitive?(node)
           return false unless node
 
-          # List of HTML elements where whitespace is semantically significant
-          whitespace_sensitive_tags = %w[pre code textarea script style]
-
           # Check if this node or any ancestor is whitespace-sensitive
           current = node
           while current
             if current.is_a?(Core::TreeNode)
               label = current.label.to_s.downcase
-              return true if whitespace_sensitive_tags.include?(label)
+              return true if Core::TreeNode::WHITESPACE_SENSITIVE_TAGS.include?(label)
             end
 
             current = current.is_a?(Core::TreeNode) ? current.parent : nil

data/lib/canon/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Canon
-  VERSION = "0.3.58"
+  VERSION = "0.3.60"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: canon
 version: !ruby/object:Gem::Version
-  version: 0.3.58
+  version: 0.3.60
 platform: ruby
 authors:
 - Ribose Inc.