canon 0.3.41 → 0.3.43

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3b6ae46cc828d11dd9474dbaad7453e48cd8c6282d6523732683e06812c919ba
-  data.tar.gz: d9245d48440921bca331af7c3be00843c21647cbd043473580c6a218f6a15f77
+  metadata.gz: 88bb496bdbab5cc434e73d3f353563e110bbaf0b73a76f009886e45743206984
+  data.tar.gz: c00284c8cc531ed825633e8e2250a1dbead21ec4db083529a6bef5d3b909bd4b
 SHA512:
-  metadata.gz: 7e9eb37f4285e047e2defa49566b19f3d2ddec40cb1fe55f6711c16674dcd076a805244c1716996a442b1bac99dc44af8e2c56ef89c49d677a5cf6649096adcc
-  data.tar.gz: f39f78aae5823e33033d328138da4dc5a748c3ee1c55079190457f926bcebeea4767ffc5e515132cd3e767347d80b4065c7fdd1d8e60e7a0833306053271b2e3
+  metadata.gz: 7e32bb5f265281f3ba69a7b9c130025d84f27578084a8a90267edaefc7e5cc6cedd8b2aa72db0a930d38def70ac07f826fc80e767cda72089d3ebedbc88219f4
+  data.tar.gz: 82a1b117a2f121dbd436962937b1aec7de7ef691b8d15f51fbf69155d6ff9fe09af5ed5f69efdbfa4724af65dbd3858525a56ac22ebb6d603c91ad9b032ab590

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

@@ -59,10 +59,55 @@ module Canon
             grouping_lines: @diff_grouping_lines,
           )
 
+          # #85/#86: DiffNodes the enricher could not locate were
+          # previously dropped silently — the absorption mechanism.
+          # Their serialized content renders as a trailing context so
+          # the change is always visible, without perturbing the
+          # located walk's line accounting.
+          append_unlocated_context(report)
+
           # Layer 6: Format the report
           format_report(report, doc1, doc2)
         end
 
+        # Trailing context for DiffNodes with no located char ranges
+        # (multiline reflowed text, relocated content): one removed
+        # DiffLine per serialized_before line, one added per
+        # serialized_after line. Nil positions render as blank line
+        # numbers.
+        def append_unlocated_context(report)
+          unlocated = @differences.select do |dn|
+            dn.char_ranges.nil? || dn.char_ranges.empty?
+          end
+          return if unlocated.empty?
+
+          unlocated = unlocated.reject do |dn|
+            (@show_diffs == :normative && !dn.normative?) ||
+              (@show_diffs == :informative && !dn.informative?)
+          end
+          return if unlocated.empty?
+
+          lines = []
+          unlocated.each do |dn|
+            fmt = dn.formatting?
+            dn.serialized_before&.split("\n")&.each do |line|
+              lines << Canon::Diff::DiffLine.new(
+                line_number: nil, new_position: nil, content: line,
+                type: :removed, diff_node: dn, formatting: fmt
+              )
+            end
+            next unless dn.serialized_after
+
+            dn.serialized_after.split("\n").each do |line|
+              lines << Canon::Diff::DiffLine.new(
+                line_number: nil, new_position: nil, content: line,
+                type: :added, diff_node: dn, formatting: fmt
+              )
+            end
+          end
+          report.contexts << Canon::Diff::DiffContext.new(lines: lines)
+        end
+
         # Format a DiffReport for display
         def format_report(report, doc1, doc2)
           return "" if report.contexts.empty?
@@ -100,12 +145,12 @@ module Canon
           context.lines.each do |diff_line|
             case diff_line.type
             when :unchanged
-              old_num = diff_line.line_number + 1
-              new_num = (diff_line.new_position || diff_line.line_number) + 1
+              old_num = diff_line.line_number&.+(1)
+              new_num = (diff_line.new_position || diff_line.line_number)&.+(1)
               output << format_unified_line(old_num, new_num, " ",
                                             diff_line.content)
             when :removed
-              line_num = diff_line.line_number + 1
+              line_num = diff_line.line_number&.+(1)
               formatting = diff_line.formatting?
               informative = diff_line.informative?
 
@@ -147,7 +192,7 @@ module Canon
                                               theme_color(:removed, :content))
                         end
             when :added
-              line_num = (diff_line.new_position || diff_line.line_number) + 1
+              line_num = (diff_line.new_position || diff_line.line_number)&.+(1)
               formatting = diff_line.formatting?
               informative = diff_line.informative?
 
@@ -255,8 +300,8 @@ module Canon
         # Format a changed diff line using DiffCharRanges for character-level highlighting.
         # Reads pre-computed char ranges from the DiffLine — NO tokenization, NO LCS.
         def format_changed_line(diff_line, _lines1)
-          old_line_num = diff_line.line_number + 1
-          new_line_num = (diff_line.new_position || diff_line.line_number) + 1
+          old_line_num = diff_line.line_number&.+(1)
+          new_line_num = (diff_line.new_position || diff_line.line_number)&.+(1)
           formatting = diff_line.formatting?
           informative = diff_line.informative?
           old_content = diff_line.old_content || diff_line.content

data/lib/canon/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Canon
-  VERSION = "0.3.41"
+  VERSION = "0.3.43"
 end

data/lib/canon/xml/c14n.rb

@@ -22,13 +22,20 @@ module Canon
         processor.process(root_node)
       end
 
-      # leptris' C-side C14N 1.1 — ~50x faster than the Ruby processor
-      # but NOT yet byte-identical on canon's edge-case corpus
-      # (attribute ordering, `>` escaping, document-level PIs, one
-      # prefix case — leptris#1015; the parity spec pins
-      # each). Opt-in via CANON_C14N_BACKEND=leptris until those
-      # close; comments mode keeps the Ruby path regardless (the
-      # native seam exposes no with-comments form).
+      # leptris' C-side C14N 1.1 — 23x faster than the Ruby processor
+      # through canon's own API (1MB document). libleptris 1.9.164
+      # closed four of the five leptris#1015 families (attribute
+      # ordering, prefix loss and rebinding, `>` and TAB escaping —
+      # unpinned in the parity spec since gem 1.9.174.0). A local
+      # default-flip attempt surfaced three MORE divergences the
+      # edge corpus did not cover, so the lane stays opt-in
+      # (CANON_C14N_BACKEND=leptris) until they close (leptris#1096):
+      # redundant namespace redeclarations are kept (Ruby removes
+      # them), xmlns:xml with the standard URI is kept (Ruby omits
+      # it), and document-level processing instructions are dropped.
+      # The Ruby lane additionally validates relative namespace URIs
+      # at parse time — a future flip must carry that check into the
+      # native wrapper. Comments mode keeps the Ruby path regardless.
       def self.native_canonicalize(xml, with_comments)
         return nil if with_comments
         return nil if RUBY_ENGINE == "opal"

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: canon
 version: !ruby/object:Gem::Version
-  version: 0.3.41
+  version: 0.3.43
 platform: ruby
 authors:
 - Ribose Inc.