sitediff 1.1.1 → 1.2.1

data/lib/sitediff/report.rb

@@ -96,7 +96,7 @@ class SiteDiff
       if @config.export
         package_report(dir)
       else
-        SiteDiff.log 'Report generated to ' + report_file.expand_path.to_s
+        SiteDiff.log "Report generated to #{report_file.expand_path}"
       end
     end
 
@@ -135,7 +135,7 @@ class SiteDiff
 
       write_settings dir
 
-      SiteDiff.log 'Report generated to ' + report_file.expand_path.to_s
+      SiteDiff.log "Report generated to #{report_file.expand_path}"
     end
 
     ##
@@ -147,7 +147,7 @@ class SiteDiff
       temp_path.mkpath
       report_path = temp_path + REPORT_DIR
       report_path.mkpath
-      files_path = report_path + 'files'
+      files_path = report_path + "files"
       files_path.mkpath
       diffs_path = dir + DIFFS_DIR
 
@@ -164,7 +164,7 @@ class SiteDiff
       end
       FileUtils.move(temp_path + REPORT_FILE_TAR, dir)
       temp_path.rmtree
-      SiteDiff.log 'Archived report generated to ' + dir.join(REPORT_FILE_TAR).to_s
+      SiteDiff.log "Archived report generated to #{dir.join(REPORT_FILE_TAR)}"
     end
 
     ##
@@ -182,7 +182,7 @@ class SiteDiff
       diff_dir.rmtree if diff_dir.exist?
 
       # Write diffs to the diff directory.
-      @results.each { |r| r.dump(dir, @config.export) if r.status == Result::STATUS_FAILURE }
+      @results.each { |r| r.dump(dir, relative: @config.export) if r.status == Result::STATUS_FAILURE }
       SiteDiff.log "All diff files written to #{diff_dir.expand_path}" unless @config.export
     end
 
@@ -245,7 +245,7 @@ class SiteDiff
       if File.exist? timestamp_file
         file = File::Stat.new(timestamp_file)
         time = file.mtime
-        time.class == Time ? time.strftime('%Y-%m-%d %H:%M') : ''
+        time.instance_of?(Time) ? time.strftime('%Y-%m-%d %H:%M') : ''
       else
         'unknown'
       end

data/lib/sitediff/result.rb

@@ -76,19 +76,19 @@ class SiteDiff
 
     # Filename to store diff
     def filename
-      File.join(Report::DIFFS_DIR, Digest::SHA1.hexdigest(path) + '.html')
+      File.join(Report::DIFFS_DIR, "#{Digest::SHA1.hexdigest(path)}.html")
     end
 
     # Returns a URL to the result diff.
     #
     # Returns nil if the result has no diffs.
-    def diff_url(relative = false)
+    def diff_url(relative: false)
       prefix = relative ? 'files/' : '/files/'
       return prefix + filename if status == STATUS_FAILURE
     end
 
     # Log the result to the terminal
-    def log(verbose = true)
+    def log(verbose: true)
       case status
       when STATUS_SUCCESS
         SiteDiff.log path, :success, 'UNCHANGED'
@@ -101,12 +101,12 @@ class SiteDiff
     end
 
     # Dump the result to a file
-    def dump(dir, relative = false)
+    def dump(dir, relative: false)
       dump_path = File.join(dir, filename)
       base = File.dirname(dump_path)
       FileUtils.mkdir_p(base) unless File.exist?(base)
       File.open(dump_path, 'w') do |f|
-        f.write(Diff.generate_diff_output(self, relative))
+        f.write(Diff.generate_diff_output(self, relative:))
       end
     end
   end

data/lib/sitediff/sanitize/dom_transform.rb

@@ -30,10 +30,10 @@ class SiteDiff
 
       ##
       # TODO: Document what this method does.
-      def targets(node)
+      def targets(node, &block)
         selectors = to_array(@rule['selector'])
         selectors.each do |sel|
-          node.css(sel).each { |n| yield n }
+          node.css(sel).each(&block)
         end
       end
 

data/lib/sitediff/sanitize/regexp.rb

@@ -45,9 +45,9 @@ class SiteDiff
 
         ##
         # TODO: Document what this method does.
-        def contexts(node)
+        def contexts(node, &block)
           selectors = @rule['selector']
-          node.css(selectors).each { |e| yield(e) }
+          node.css(selectors).each(&block)
         end
 
         ##

data/lib/sitediff/sanitize.rb

@@ -62,7 +62,7 @@ class SiteDiff
       (rules = @config[name]) || (return nil)
 
       # Already an array? Do nothing.
-      if rules[0]&.respond_to?('each') && rules[0]&.fetch('value')
+      if rules[0].respond_to?('each') && rules[0]&.fetch('value')
       # If it is a hash, put it in an array.
       elsif rules['value']
         rules = [rules]
@@ -146,7 +146,7 @@ class SiteDiff
     def select_regions(node, regions, output)
       regions = output.map do |name|
         selector = get_named_region(regions, name)['selector']
-        region = Nokogiri::XML.fragment('<region id="' + name + '"></region>').at_css('region')
+        region = Nokogiri::XML.fragment("<region id=\"#{name}\"></region>").at_css('region')
         matching = node.css(selector)
         matching.each { |m| region.add_child m }
         region
@@ -210,7 +210,7 @@ class SiteDiff
     end
 
     # Parse HTML into a node
-    def self.domify(str, force_doc = false)
+    def self.domify(str, force_doc: false)
       if force_doc || /<!DOCTYPE/.match(str[0, 512])
         Nokogiri::HTML(str)
       else
@@ -224,9 +224,9 @@ class SiteDiff
         obj
       # node or fragment
       elsif Nokogiri::XML::Node == obj.class || Nokogiri::HTML::DocumentFragment == obj.class
-        domify(obj.to_s, true)
+        domify(obj.to_s, force_doc: true)
       else
-        to_document(domify(obj, false))
+        to_document(domify(obj, force_doc: false))
       end
     end
 

data/lib/sitediff/uriwrapper.rb

@@ -48,12 +48,13 @@ class SiteDiff
 
     ##
     # Creates a UriWrapper.
-    def initialize(uri, curl_opts = DEFAULT_CURL_OPTS, debug = true)
+    def initialize(uri, curl_opts = DEFAULT_CURL_OPTS, debug: true, referrer: '')
       @uri = uri.respond_to?(:scheme) ? uri : Addressable::URI.parse(uri)
       # remove trailing '/'s from local URIs
       @uri.path.gsub!(%r{/*$}, '') if local?
       @curl_opts = curl_opts
       @debug = debug
+      @referrer = referrer
     end
 
     ##
@@ -103,10 +104,9 @@ class SiteDiff
     # Returns the encoding of an HTTP response from headers , nil if not
     # specified.
     def charset_encoding(http_headers)
-      if (content_type = http_headers['Content-Type'])
-        if (md = /;\s*charset=([-\w]*)/.match(content_type))
-          md[1]
-        end
+      content_type = http_headers['Content-Type']
+      if (md = /;\s*charset=([-\w]*)/.match(content_type))
+        md[1]
       end
     end
 
@@ -117,7 +117,7 @@ class SiteDiff
     def typhoeus_request
       params = @curl_opts.dup
       # Allow basic auth
-      params[:userpwd] = @uri.user + ':' + @uri.password if @uri.user
+      params[:userpwd] = "#{@uri.user}: #{@uri.password}" if @uri.user
 
       req = Typhoeus::Request.new(to_s, params)
 
@@ -137,32 +137,31 @@ class SiteDiff
           raise if @debug
 
           yield ReadResult.error(
-            "Parsing error for #{@uri}: #{e.message}"
+            "Parsing error for #{@uri}: #{e.message}  From page: #{@referrer}"
           )
         rescue StandardError => e
           raise if @debug
 
           yield ReadResult.error(
-            "Unknown parsing error for #{@uri}: #{e.message}"
+            "Unknown parsing error for #{@uri}: #{e.message}  From page: #{@referrer}"
           )
         end
       end
 
-      req.on_failure do |resp|
+      req.on_failure do |resp|           
         if resp&.status_message
-          msg = resp.status_message
           yield ReadResult.error(
-            "HTTP error when loading #{@uri}: #{msg}",
+            "HTTP error when loading #{@uri} : [#{resp.response_code}] #{resp.status_message}  From page: #{@referrer}",
             resp.response_code
           )
         elsif (msg = resp.options[:return_code])
           yield ReadResult.error(
-            "Connection error when loading #{@uri}: #{msg}",
+            "Connection error when loading #{@uri} : [#{resp.options[:return_code]}] #{msg}  From page: #{@referrer}",
             resp.response_code
           )
         else
           yield ReadResult.error(
-            "Unknown error when loading #{@uri}: #{msg}",
+            "Unknown error when loading #{@uri} : [#{resp.response_code}] #{resp.status_message}  From page: #{@referrer}",
             resp.response_code
           )
         end

data/lib/sitediff/webserver/resultserver.rb

@@ -14,6 +14,7 @@ class SiteDiff
         ##
         # Creates a Cache Servlet.
         def initialize(_server, cache)
+          super
           @cache = cache
         end
 
@@ -41,6 +42,7 @@ class SiteDiff
         ##
         # Creates a Side By Side Servlet.
         def initialize(_server, cache, settings)
+          super
           @cache = cache
           @settings = settings
         end

data/lib/sitediff/webserver.rb

@@ -77,15 +77,18 @@ class SiteDiff
       BASE = 'spec/sites/ruby-doc.org'
       NAMES = %w[core-1.9.3 core-2.0].freeze
 
+      # Initialize web server.
       def initialize(port = PORT, base = BASE, names = NAMES)
         dirs = names.map { |n| File.join(base, n) }
         super(port, dirs, quiet: true)
       end
 
+      # Get the before site uri.
       def before
         uris.first
       end
 
+      # Get the after site uri.
       def after
         uris.last
       end

data/lib/sitediff.rb

@@ -47,7 +47,7 @@ class SiteDiff
         bg = :yellow
       end
 
-      label = '[' + label.to_s + ']'
+      label = "[#{label}]"
       label = Rainbow(label)
       label = label.bg(bg) if bg
       label = label.fg(fg) if fg
@@ -76,7 +76,7 @@ class SiteDiff
   end
 
   # Initialize SiteDiff.
-  def initialize(config, cache, verbose = true, debug = false)
+  def initialize(config, cache, verbose: true, debug: false)
     @cache = cache
     @verbose = verbose
     @debug = debug
@@ -97,7 +97,7 @@ class SiteDiff
   end
 
   # Sanitize HTML.
-  def sanitize(path, read_results)
+  def sanitize(path_passed, read_results)
     %i[before after].map do |tag|
       html = read_results[tag].content
       # TODO: See why encoding is empty while running tests.
@@ -107,8 +107,8 @@ class SiteDiff
       # during rspec tests for some reason.
       encoding = read_results[tag].encoding
       if encoding || html.length.positive?
-        section = @config.send(tag, true)
-        opts = { path: path }
+        section = @config.send(tag, apply_preset: true)
+        opts = { path: path_passed }
         opts[:output] = @config.output if @config.output
         Sanitizer.new(html, section, opts).sanitize
       else
@@ -144,7 +144,7 @@ class SiteDiff
 
     # Print results in order!
     while (next_diff = @results[@ordered.first])
-      next_diff.log(@verbose)
+      next_diff.log(verbose: @verbose)
       @ordered.shift
     end
   end
@@ -160,7 +160,7 @@ class SiteDiff
     @ordered = @config.paths.dup
 
     unless @cache.read_tags.empty?
-      SiteDiff.log('Using sites from cache: ' + @cache.read_tags.sort.join(', '))
+      SiteDiff.log("Using sites from cache: #{@cache.read_tags.sort.join(', ')}")
     end
 
     # TODO: Fix this after config merge refactor!
@@ -175,7 +175,7 @@ class SiteDiff
       @config.setting(:interval),
       @config.setting(:concurrency),
       curl_opts,
-      @debug,
+      debug: @debug,
       before: @config.before_url,
       after: @config.after_url
     )
@@ -203,7 +203,7 @@ class SiteDiff
   ##
   # Get SiteDiff gemspec.
   def self.gemspec
-    file = ROOT_DIR + '/sitediff.gemspec'
+    file = "#{ROOT_DIR}/sitediff.gemspec"
     Gem::Specification.load(file)
   end