canon 0.3.40 → 0.3.41

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 7ceb43d90065cc39a64f8c74b6317dd9f8752aaa807ec7869ae3188c9e494b43
-  data.tar.gz: d6375f29977244331faff0d8dd54396d52494133d77e4286011804a63bd1d23f
+  metadata.gz: 3b6ae46cc828d11dd9474dbaad7453e48cd8c6282d6523732683e06812c919ba
+  data.tar.gz: d9245d48440921bca331af7c3be00843c21647cbd043473580c6a218f6a15f77
 SHA512:
-  metadata.gz: bed4791632c2d75f6db4e89e5c991b05a92593fdfaa16d7b20c21dbf5be3070914b0b4214567e9ca3045db8f9d3a8312a218acca28574214d76d817ee09706fd
-  data.tar.gz: 0d3f8b7bad01bd3bd1e089f5488dac9cd23ca9761ab287ad3f5f0340e918262e5b4fcf63c4bb64ddb427e19c3e1bec878b0bef9aaf93b596c5a2a1f125dadd99
+  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/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 = "0.3.40"
+  VERSION = "0.3.41"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: canon
 version: !ruby/object:Gem::Version
-  version: 0.3.40
+  version: 0.3.41
 platform: ruby
 authors:
 - Ribose Inc.