canon 0.3.39 → 0.3.41

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 6b65514ddc21b35416e7856a21f17f084cde6f6176a0f84dd8ee03a3ec7c8604
-  data.tar.gz: e6390a45d84dc78708aa3f043f053c41bba8b68c30170b11a58282581365c09d
+  metadata.gz: 3b6ae46cc828d11dd9474dbaad7453e48cd8c6282d6523732683e06812c919ba
+  data.tar.gz: d9245d48440921bca331af7c3be00843c21647cbd043473580c6a218f6a15f77
 SHA512:
-  metadata.gz: 338db190efc246cb7a23f06b4a6f3dc620e969e3029cc2384f52bf8b082f6903dd21883b29316bcff676bb36f446e1fcad8cab434d8590f40b57a32a6d56baca
-  data.tar.gz: 602303bc515970ac9085e1a5184d8684604b80c5047d65cc3bf6bb6aa2854fdce1bd9eb3a61b4ee8e0f0c74f3f347d63d62133c9e8df93b51d146eca67807da9
+  metadata.gz: 7e9eb37f4285e047e2defa49566b19f3d2ddec40cb1fe55f6711c16674dcd076a805244c1716996a442b1bac99dc44af8e2c56ef89c49d677a5cf6649096adcc
+  data.tar.gz: f39f78aae5823e33033d328138da4dc5a748c3ee1c55079190457f926bcebeea4767ffc5e515132cd3e767347d80b4065c7fdd1d8e60e7a0833306053271b2e3

data/lib/canon/diff/diff_node_enricher.rb

@@ -1019,7 +1019,9 @@ module Canon
         opener = /<#{element_name}[>\s]/
         occurrences = SourceLocator.locate_all(value, text, line_map)
         occurrences.each do |occ|
-          count = text[0...occ[:char_offset]].scan(opener).length
+          count = count_elements_before_position_open(text,
+                                                      occ[:char_offset],
+                                                      opener)
           return occ if count == target_index
         end
 
@@ -1445,9 +1447,29 @@ range_start, range_end)
       # @param char_offset [Integer] character offset to check before
       # @param element_name [String] name of element to count
       # @return [Integer] element index (0-based) of the element containing the position
+      # Occurrence count with offsets AT the element's own opening
+      # tag (no inside-the-element correction — see
+      # locate_element_at_index).
+      def count_elements_before_position_open(text, char_offset, opener)
+        count = 0
+        pos = 0
+        while (hit = text.index(opener, pos)) && hit < char_offset
+          count += 1
+          pos = hit + 1
+        end
+        count
+      end
+
       def count_elements_before_position(text, char_offset, element_name)
-        prefix = text[0...char_offset]
-        count = prefix.scan(/<#{element_name}[>\s]/).length
+        # Index loop instead of copying the prefix per call — the
+        # copy was O(offset) on every occurrence check.
+        opener = /<#{element_name}[>\s]/
+        count = 0
+        pos = 0
+        while (hit = text.index(opener, pos)) && hit < char_offset
+          count += 1
+          pos = hit + 1
+        end
         # Subtract 1 because the count includes the element we are inside
         [count - 1, 0].max
       end

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

@@ -67,8 +67,8 @@ module Canon
           output = []
 
           # Detect non-ASCII characters
-          all_text = (lines1 + lines2).join
-          non_ascii = Legend.detect_non_ascii(all_text, @visualization_map)
+          non_ascii = Legend.detect_non_ascii(doc1, @visualization_map)
+          non_ascii.merge!(Legend.detect_non_ascii(doc2, @visualization_map))
 
           # Add Unicode legend if needed
           unless non_ascii.empty?

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

@@ -19,8 +19,8 @@ module Canon
           lines2 = doc2.split("\n", -1)
 
           # Detect non-ASCII characters in the diff
-          all_text = (lines1 + lines2).join
-          non_ascii = Legend.detect_non_ascii(all_text, @visualization_map)
+          non_ascii = Legend.detect_non_ascii(doc1, @visualization_map)
+          non_ascii.merge!(Legend.detect_non_ascii(doc2, @visualization_map))
 
           # Add Unicode legend if any non-ASCII characters detected
           unless non_ascii.empty?

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

@@ -72,9 +72,11 @@ module Canon
 
           output = []
 
-          # Detect non-ASCII characters
-          all_text = (lines1 + lines2).join
-          non_ascii = Legend.detect_non_ascii(all_text, @visualization_map)
+          # Detect non-ASCII characters — scan the source strings
+          # directly; joining the split lines back into one document
+          # per render was pure garbage.
+          non_ascii = Legend.detect_non_ascii(doc1, @visualization_map)
+          non_ascii.merge!(Legend.detect_non_ascii(doc2, @visualization_map))
 
           # Add Unicode legend if needed
           unless non_ascii.empty?

data/lib/canon/diff_formatter/diff_detail_formatter/text_utils.rb

@@ -113,8 +113,7 @@ module Canon
         def self.needs_escaping?(text)
           return false if text.nil?
 
-          text.each_char.any? do |c|
-            codepoint = c.ord
+          text.each_codepoint.any? do |codepoint|
             codepoint < 32 || codepoint >= 127 || codepoint == 34 || codepoint == 92
           end
         end

data/lib/canon/diff_formatter/legend.rb

@@ -12,12 +12,23 @@ module Canon
       # @param visualization_map [Hash] Character visualization map
       # @return [Hash] Hash of characters with their metadata
       def self.detect_non_ascii(text, visualization_map)
+        # each_char allocates a 1-char String per character — over two
+        # full documents per render that is the presentation stage's
+        # biggest allocation source. ASCII-only text (the common case)
+        # needs no scan at all; otherwise scan codepoints and
+        # materialize the character String only at non-ASCII offsets.
+        return {} if text.ascii_only?
+
         detected = {}
         category_map = DiffFormatter::CHARACTER_CATEGORY_MAP
         metadata = DiffFormatter::CHARACTER_METADATA
 
-        text.each_char do |char|
-          next if char.ord <= 127
+        index = -1
+        text.each_codepoint do |codepoint|
+          index += 1
+          next if codepoint <= 127
+
+          char = text[index, 1]
           next if detected.key?(char)
 
           visualization = visualization_map.fetch(char, char)

data/lib/canon/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Canon
-  VERSION = "0.3.39"
+  VERSION = "0.3.41"
 end

data/lib/tasks/benchmark_runner.rb

@@ -225,6 +225,13 @@ compare_against: nil)
       failure_means: "Slow formatting affects serialization performance. C14N is critical for digital signatures and XML canonicalization.",
       compare_against: "Previous branch (main).",
     },
+    by_line_rendering: {
+      name: "By-Line Rendering",
+      icon: "🖨️",
+      description: "Verbose by_line diff rendering (enricher, line builder, formatters). Referees the #86 rendering lane.",
+      failure_means: "Slow diff display affects developer-facing CLI output and CI failure reports.",
+      compare_against: "Previous branch (main). Inputs: documents with scattered value differences.",
+    },
     data_comparison: {
       name: "Data Comparison",
       icon: "🧮",
@@ -272,6 +279,9 @@ compare_against: nil)
       { name: "JSON", method: :json_format, desc: "JSON formatting" },
       { name: "YAML", method: :yaml_format, desc: "YAML formatting" },
     ],
+    by_line_rendering: [
+      { name: "XML", method: :by_line_render_xml, desc: "by_line XML render" },
+    ],
     data_comparison: [
       { name: "JSON", method: :json_compare_equivalent,
         desc: "JSON equivalence" },
@@ -568,6 +578,18 @@ compare_against: nil)
       yaml = DataGenerator.generate_yaml(items: @items)
       data = YAML.safe_load(yaml, permitted_classes: [Time])
       measure { Canon.format_yaml(data) }
+    when :by_line_render_xml
+      xml1 = DataGenerator.generate_xml(items: @items)
+      xml2 = DataGenerator.generate_xml(items: @items)
+      formatter = Canon::DiffFormatter.new(mode: :by_line,
+                                           display_preprocessing: :pretty_print,
+                                           use_color: false)
+      result = Canon::Comparison.equivalent?(xml1, xml2, format: :xml,
+                                                         verbose: true)
+      measure do
+        formatter.format(result.differences, :xml,
+                         doc1: xml1, doc2: xml2)
+      end
     when :json_compare_equivalent
       json1 = DataGenerator.generate_json(items: @items)
       json2 = DataGenerator.generate_json(items: @items)

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: canon
 version: !ruby/object:Gem::Version
-  version: 0.3.39
+  version: 0.3.41
 platform: ruby
 authors:
 - Ribose Inc.