RubyGems - canon - Versions diffs - 0.1.22 → 0.2.0 - Mend

canon 0.1.22 → 0.2.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Files changed (56) hide show

checksums.yaml +4 -4
data/.rubocop_todo.yml +174 -25
data/docs/INDEX.adoc +4 -0
data/docs/advanced/diff-classification.adoc +3 -2
data/docs/features/configuration-profiles.adoc +288 -0
data/docs/features/diff-formatting/character-visualization.adoc +153 -454
data/docs/features/diff-formatting/display-filtering.adoc +44 -0
data/docs/features/diff-formatting/display-preprocessing.adoc +656 -0
data/docs/features/diff-formatting/index.adoc +47 -0
data/docs/features/diff-formatting/pretty-diff-mode.adoc +154 -0
data/docs/features/environment-configuration/override-system.adoc +10 -3
data/docs/features/index.adoc +9 -0
data/docs/features/match-options/index.adoc +32 -42
data/docs/features/match-options/pretty-printed-fixtures.adoc +270 -0
data/docs/guides/choosing-configuration.adoc +22 -0
data/docs/reference/environment-variables.adoc +121 -1
data/docs/reference/options-across-interfaces.adoc +182 -2
data/lib/canon/cli.rb +20 -0
data/lib/canon/commands/diff_command.rb +7 -2
data/lib/canon/commands/format_command.rb +1 -1
data/lib/canon/comparison/html_comparator.rb +20 -15
data/lib/canon/comparison/html_compare_profile.rb +4 -4
data/lib/canon/comparison/markup_comparator.rb +12 -3
data/lib/canon/comparison/match_options/base_resolver.rb +29 -7
data/lib/canon/comparison/match_options/json_resolver.rb +9 -0
data/lib/canon/comparison/match_options/xml_resolver.rb +16 -2
data/lib/canon/comparison/match_options/yaml_resolver.rb +10 -0
data/lib/canon/comparison/match_options.rb +4 -1
data/lib/canon/comparison/whitespace_sensitivity.rb +189 -137
data/lib/canon/comparison/xml_comparator/child_comparison.rb +21 -4
data/lib/canon/comparison/xml_comparator.rb +14 -12
data/lib/canon/comparison/xml_node_comparison.rb +51 -6
data/lib/canon/comparison.rb +52 -9
data/lib/canon/config/env_schema.rb +32 -4
data/lib/canon/config/override_resolver.rb +16 -3
data/lib/canon/config/profile_loader.rb +135 -0
data/lib/canon/config/profiles/metanorma.yml +74 -0
data/lib/canon/config/profiles/metanorma_debug.yml +8 -0
data/lib/canon/config/type_converter.rb +8 -0
data/lib/canon/config.rb +469 -5
data/lib/canon/diff/diff_classifier.rb +41 -11
data/lib/canon/diff_formatter/diff_detail_formatter/dimension_formatter.rb +48 -17
data/lib/canon/diff_formatter/diff_detail_formatter/node_utils.rb +58 -0
data/lib/canon/diff_formatter/diff_detail_formatter.rb +22 -7
data/lib/canon/diff_formatter/theme.rb +24 -17
data/lib/canon/diff_formatter.rb +493 -36
data/lib/canon/pretty_printer/xml_normalized.rb +395 -0
data/lib/canon/rspec_matchers.rb +36 -0
data/lib/canon/tree_diff/matchers/hash_matcher.rb +26 -11
data/lib/canon/version.rb +1 -1
data/lib/canon/xml/nodes/namespace_node.rb +4 -0
data/lib/canon/xml/nodes/processing_instruction_node.rb +4 -0
data/lib/canon/xml/nodes/root_node.rb +4 -0
data/lib/canon/xml/nodes/text_node.rb +4 -0
data/lib/tasks/performance_helpers.rb +2 -2
metadata +24 -2

data/lib/canon/comparison.rb CHANGED Viewed

@@ -273,8 +273,10 @@ module Canon
                              text_content structural_whitespace attribute_presence
                              attribute_order attribute_values element_position
                              comments format similarity_threshold hash_matching
-                             similarity_matching propagation sensitive_elements
-                             whitespace_sensitive_elements respect_xml_space]
+                             similarity_matching propagation
+                             preserve_whitespace_elements
+                             collapse_whitespace_elements
+                             strip_whitespace_elements respect_xml_space]
         match_options_only = match_opts_hash.slice(*match_only_keys)
         # Convert operations to DiffNodes for unified pipeline
@@ -596,10 +598,30 @@ module Canon
           format1 = format2 = opts[:format]
           # Parse HTML strings if format is html/html4/html5
           if %i[html html4 html5].include?(opts[:format])
-            obj1 = HtmlParser.parse(obj1, opts[:format]) if obj1.is_a?(String)
-            obj2 = HtmlParser.parse(obj2, opts[:format]) if obj2.is_a?(String)
-            # Note: We preserve html4/html5 format instead of normalizing to :html
-            # This allows HtmlComparator to use the correct parsing behavior
+            # Preserve original strings for display (HTML fragment
+            # parsers can mutate the DOM).
+            opts[:_original_str1] = obj1.dup if obj1.is_a?(String)
+            opts[:_original_str2] = obj2.dup if obj2.is_a?(String)
+            if opts[:format] == :html5
+              # HTML5 fragment parsing is safe — it normalizes without
+              # destructive content-model mutations.
+              obj1 = HtmlParser.parse(obj1, :html5) if obj1.is_a?(String)
+              obj2 = HtmlParser.parse(obj2, :html5) if obj2.is_a?(String)
+            else
+              # HTML4 fragment parsing mutates the DOM (strips <body>
+              # attributes, re-parents <h1> content, etc.).  Use XML
+              # fragment parsing which preserves structure faithfully.
+              if obj1.is_a?(String)
+                obj1 = Nokogiri::XML.fragment(
+                  strip_xml_preamble(obj1),
+                )
+              end
+              if obj2.is_a?(String)
+                obj2 = Nokogiri::XML.fragment(
+                  strip_xml_preamble(obj2),
+                )
+              end
+            end
           end
         else
           format1 = FormatDetector.detect(obj1)
@@ -638,12 +660,21 @@ module Canon
         # get match_profile if it is not defined in options
         # but defined in config
-        if opts[:match_profile].nil? &&
-            Canon::Config.instance.respond_to?(comparison_format)
+        if Canon::Config.instance.respond_to?(comparison_format)
           format_config = Canon::Config.instance.public_send(comparison_format)
-          if format_config.match.profile
+          if opts[:match_profile].nil? && format_config.match.profile
             opts[:match_profile] = format_config.match.profile
           end
+          # Pass YAML profile's extra match options (e.g., preserve_whitespace_elements)
+          # that are stored in MatchConfig's resolver but not exposed via the
+          # built-in MATCH_PROFILES system. These supplement the built-in profile.
+          profile_opts = format_config.match.profile_options
+          if profile_opts.any? && opts[:global_options].nil?
+            opts[:global_options] = profile_opts
+          elsif profile_opts.any?
+            # Merge: global_options already set (e.g., per-call) takes precedence
+            opts[:global_options] = opts[:global_options].merge(profile_opts)
+          end
         end
         case comparison_format
@@ -658,6 +689,18 @@ module Canon
         end
       end
+      # Strip XML declarations and DOCTYPE preambles from an HTML string
+      # so it can be safely parsed with Nokogiri::XML.fragment without
+      # generating processing-instruction nodes.
+      def strip_xml_preamble(str)
+        str = str.sub(/\A\s*<\?xml[^?]*\?>\s*/m, "")
+        if (i = str.index(/<!DOCTYPE/i))
+          j = str.index(">", i)
+          str = (str[0...i] + str[(j + 1)..]).strip if j
+        end
+        str
+      end
       # Detect the format of an object (delegates to FormatDetector)
       #
       # @param obj [Object] Object to detect format of

data/lib/canon/config/env_schema.rb CHANGED Viewed

@@ -15,9 +15,28 @@ module Canon
         verbose_diff: :boolean,
         algorithm: :symbol,
         show_raw_inputs: :boolean,
+        show_raw_expected: :boolean,
+        show_raw_received: :boolean,
         show_preprocessed_inputs: :boolean,
+        show_preprocessed_expected: :boolean,
+        show_preprocessed_received: :boolean,
+        show_prettyprint_inputs: :boolean,
+        show_prettyprint_expected: :boolean,
+        show_prettyprint_received: :boolean,
         show_line_numbered_inputs: :boolean,
+        character_visualization: :symbol,     # true, false, :content_only
         display_format: :symbol,
+        display_preprocessing: :symbol,       # :none, :pretty_print, :normalize_pretty_print, :c14n
+        pretty_printer_indent: :integer,
+        pretty_printer_indent_type: :symbol,  # :space or :tab
+        preserve_whitespace_elements: :string_array,
+        collapse_whitespace_elements: :string_array,
+        strip_whitespace_elements: :string_array,
+        pretty_printed_expected: :boolean,
+        pretty_printed_received: :boolean,
+        pretty_printer_sort_attributes: :boolean,
+        compact_semantic_report: :boolean,
+        expand_difference: :boolean,
         theme: :symbol,
         # MatchConfig attributes
@@ -47,13 +66,22 @@ module Canon
         def all_diff_attributes
           %i[mode use_color context_lines grouping_lines show_diffs
-             verbose_diff algorithm show_raw_inputs show_preprocessed_inputs
-             show_line_numbered_inputs display_format max_file_size max_node_count max_diff_lines
-             theme]
+             verbose_diff algorithm show_raw_inputs show_raw_expected show_raw_received
+             show_preprocessed_inputs show_preprocessed_expected show_preprocessed_received
+             show_prettyprint_inputs show_prettyprint_expected show_prettyprint_received
+             show_line_numbered_inputs character_visualization
+             display_format display_preprocessing
+             pretty_printer_indent pretty_printer_indent_type
+             preserve_whitespace_elements collapse_whitespace_elements strip_whitespace_elements
+             pretty_printed_expected pretty_printed_received
+             pretty_printer_sort_attributes
+             compact_semantic_report expand_difference
+             max_file_size max_node_count max_diff_lines theme]
         end
         def all_match_attributes
-          %i[profile]
+          %i[profile
+             preserve_whitespace_elements collapse_whitespace_elements strip_whitespace_elements]
         end
         def all_format_attributes

data/lib/canon/config/override_resolver.rb CHANGED Viewed

@@ -3,14 +3,15 @@
 module Canon
   class Config
     # Resolves configuration values using priority chain
-    # Priority: ENV > programmatic > defaults
+    # Priority: ENV > programmatic > profile > defaults
     class OverrideResolver
-      attr_reader :defaults, :programmatic, :env
+      attr_reader :defaults, :programmatic, :env, :profile
-      def initialize(defaults: {}, programmatic: {}, env: {})
+      def initialize(defaults: {}, programmatic: {}, env: {}, profile: {})
         @defaults = defaults
         @programmatic = programmatic
         @env = env
+        @profile = profile
       end
       # Resolve a single value using priority chain
@@ -18,6 +19,7 @@ module Canon
       def resolve(key)
         return @env[key] if @env.key?(key)
         return @programmatic[key] if @programmatic.key?(key)
+        return @profile[key] if @profile.key?(key)
         @defaults[key]
       end
@@ -32,6 +34,16 @@ module Canon
         @env[key] = value
       end
+      # Update profile value
+      def set_profile(key, value)
+        @profile[key] = value
+      end
+      # Clear all profile values
+      def clear_profile!
+        @profile = {}
+      end
       # Check if value is set by ENV
       def env_set?(key)
         @env.key?(key)
@@ -46,6 +58,7 @@ module Canon
       def source_for(key)
         return :env if @env.key?(key)
         return :programmatic if @programmatic.key?(key)
+        return :profile if @profile.key?(key)
         return :default if @defaults.key?(key)
         nil

data/lib/canon/config/profile_loader.rb ADDED Viewed

@@ -0,0 +1,135 @@
+# frozen_string_literal: true
+require "yaml"
+require_relative "env_schema"
+module Canon
+  class Config
+    # Loads configuration profiles from YAML files.
+    # Supports built-in profiles (shipped with the gem) and external file paths.
+    # Profiles can inherit from other profiles via the +inherits+ key.
+    class ProfileLoader
+      PROFILES_DIR = File.expand_path("profiles", __dir__).freeze
+      class << self
+        # Load a profile by name (Symbol for built-in) or file path (String).
+        # Returns a merged Hash with inheritance resolved.
+        def load(name_or_path)
+          key = cache_key(name_or_path)
+          cache[key] ||= resolve(name_or_path, [])
+        end
+        # List available built-in profile names.
+        def available_profiles
+          return [] unless Dir.exist?(PROFILES_DIR)
+          Dir.glob(File.join(PROFILES_DIR, "*.yml")).map do |path|
+            File.basename(path, ".yml").to_sym
+          end.sort
+        end
+        def reset_cache!
+          @cache = nil
+        end
+        private
+        def cache
+          @cache ||= {}
+        end
+        def cache_key(name_or_path)
+          if name_or_path.is_a?(Symbol)
+            name_or_path
+          else
+            File.expand_path(name_or_path.to_s)
+          end
+        end
+        # Resolve a profile, following inheritance chain.
+        # +seen+ tracks visited profiles for cycle detection.
+        def resolve(name_or_path, seen)
+          path = resolve_path(name_or_path)
+          resolve_from_path(path, seen)
+        end
+        def resolve_from_path(path, seen)
+          canonical = File.expand_path(path)
+          if seen.include?(canonical)
+            chain = seen.map { |s| File.basename(s, ".yml") }.join(" -> ")
+            raise Canon::Error,
+                  "Profile inheritance cycle detected: #{chain} -> #{File.basename(
+                    canonical, '.yml'
+                  )}"
+          end
+          seen = seen + [canonical]
+          data = load_yaml(path)
+          if data["inherits"]
+            parent_path = resolve_inherits_path(data["inherits"])
+            parent = resolve_from_path(parent_path, seen)
+            data = deep_merge(parent, data)
+          end
+          data.delete("inherits")
+          data
+        end
+        # Determine the YAML file path from a name or path value.
+        # Symbols are looked up as built-in profiles; strings are treated
+        # as file paths.
+        def resolve_path(name_or_path)
+          if name_or_path.is_a?(Symbol)
+            path = File.join(PROFILES_DIR, "#{name_or_path}.yml")
+            unless File.exist?(path)
+              available = available_profiles.join(", ")
+              raise Canon::Error,
+                    "Unknown config profile: #{name_or_path}. Available: #{available}"
+            end
+            path
+          else
+            expanded = File.expand_path(name_or_path.to_s)
+            unless File.exist?(expanded)
+              raise Canon::Error, "Profile file not found: #{expanded}"
+            end
+            expanded
+          end
+        end
+        # Resolve an +inherits+ value from YAML (always a string).
+        # Tries built-in profile name first, then file path.
+        def resolve_inherits_path(value)
+          builtin = File.join(PROFILES_DIR, "#{value}.yml")
+          return builtin if File.exist?(builtin)
+          expanded = File.expand_path(value)
+          return expanded if File.exist?(expanded)
+          raise Canon::Error, "Inherited profile not found: #{value}"
+        end
+        def load_yaml(path)
+          content = File.read(path)
+          YAML.safe_load(content, permitted_classes: [Symbol]) || {}
+        end
+        # Deep merge two hashes. Arrays are replaced (not concatenated).
+        def deep_merge(base, overlay)
+          result = base.dup
+          overlay.each do |key, value|
+            result[key] = if result[key].is_a?(Hash) && value.is_a?(Hash)
+                            deep_merge(result[key], value)
+                          else
+                            value
+                          end
+          end
+          result
+        end
+      end
+    end
+  end
+end

data/lib/canon/config/profiles/metanorma.yml ADDED Viewed

@@ -0,0 +1,74 @@
+---
+name: metanorma
+description: Standard Metanorma spec configuration for XML/HTML comparison
+shared:
+  preprocessing: format
+  match:
+    profile: spec_friendly
+  diff:
+    show_diffs: normative
+    verbose_diff: false
+    show_line_numbered_inputs: false
+    show_raw_inputs: false
+    show_raw_expected: false
+    show_raw_received: false
+    show_prettyprint_expected: false
+    context_lines: 5
+    mode: pretty_diff
+    algorithm: dom
+    display_format: canonical
+    display_preprocessing: normalize_pretty_print
+    compact_semantic_report: true
+    pretty_printed_expected: true
+    expand_difference: true
+    pretty_printer_sort_attributes: true
+formats:
+  xml:
+    match:
+      profile: spec_friendly
+      # Elements where whitespace is PRESERVED exactly (no manipulation)
+      # All whitespace characters are significant in these elements
+      preserve_whitespace_elements:
+        - body
+        - passthrough
+      # Elements where whitespace is COLLAPSED (HTML-style behavior)
+      # Multiple whitespace chars collapse to single space; boundaries preserved
+      collapse_whitespace_elements:
+        - p
+        - title
+        - name
+        - td
+        - th
+        - dt
+        - form
+        - floating-title
+        - variant-title
+        - field-of-application
+        - usage-info
+        - pronunciation
+        - domain
+        - subject
+        - fmt-title
+        - fmt-name
+        - semx
+        - fmt-identifier
+        - fmt-xref-label
+        - fmt-definition
+        - fmt-fn-label
+        - fmt-sourcecode
+        - fmt-preferred
+        - fmt-admitted
+        - fmt-deprecates
+        - note
+        - abstract
+        - formattedref
+        - description
+        - identifier
+      # Elements where whitespace-only nodes are STRIPPED
+      # Default: none (elements not in any list use xml:space or format defaults)
+      strip_whitespace_elements: []
+  html:
+    diff:
+      show_raw_inputs: true

data/lib/canon/config/profiles/metanorma_debug.yml ADDED Viewed

@@ -0,0 +1,8 @@
+---
+name: metanorma_debug
+description: Metanorma profile with debug output enabled
+inherits: metanorma
+shared:
+  diff:
+    show_prettyprint_received: true

data/lib/canon/config/type_converter.rb CHANGED Viewed

@@ -28,6 +28,8 @@ module Canon
             convert_symbol(value)
           when :string
             value
+          when :string_array
+            convert_string_array(value)
           else
             value
           end
@@ -53,6 +55,12 @@ module Canon
         def convert_symbol(value)
           value.to_sym
         end
+        # Parse comma-separated element names from an ENV string.
+        # e.g. "p,li,td,th" → ["p", "li", "td", "th"]
+        def convert_string_array(value)
+          value.split(",").map(&:strip).reject(&:empty?)
+        end
       end
     end
   end