Replace simple regex with equivalent Ruby methods (#6736)

Merge pull request 6736

benchmark/sanitize-url.rb

@@ -0,0 +1,26 @@
+#!/usr/bin/env ruby
+
+require "benchmark/ips"
+
+PATH = "/../../..../...//.....//lorem/ipsum//dolor///sit.xyz"
+
+def sanitize_with_regex
+  "/" + PATH.gsub(%r!/{2,}!, "/").gsub(%r!\.+/|\A/+!, "")
+end
+
+def sanitize_with_builtin
+  "/#{PATH}".gsub("..", "/").gsub("./", "").squeeze("/")
+end
+
+if sanitize_with_regex == sanitize_with_builtin
+  Benchmark.ips do |x|
+    x.report("sanitize w/ regexes") { sanitize_with_regex }
+    x.report("sanitize w/ builtin") { sanitize_with_builtin }
+    x.compare!
+  end
+else
+  puts "w/ regexes: #{sanitize_with_regex}"
+  puts "w/ builtin: #{sanitize_with_builtin}"
+  puts ""
+  puts "Thank you. Do try again :("
+end

lib/jekyll/document.rb

@@ -417,7 +417,7 @@ module Jekyll
     def merge_categories!(other)
       if other.key?("categories") && !other["categories"].nil?
         if other["categories"].is_a?(String)
-          other["categories"] = other["categories"].split(%r!\s+!).map(&:strip)
+          other["categories"] = other["categories"].split
         end
         other["categories"] = (data["categories"] || []) | other["categories"]
       end

lib/jekyll/static_file.rb

@@ -137,7 +137,7 @@ module Jekyll
                    :template     => @collection.url_template,
                    :placeholders => placeholders,
                  })
-               end.to_s.gsub(%r!/$!, "")
+               end.to_s.chomp("/")
     end
 
     # Returns the type of the collection if present, nil otherwise.

lib/jekyll/theme_builder.rb

@@ -8,7 +8,7 @@ class Jekyll::ThemeBuilder
   attr_reader :name, :path, :code_of_conduct
 
   def initialize(theme_name, opts)
-    @name = theme_name.to_s.tr(" ", "_").gsub(%r!_+!, "_")
+    @name = theme_name.to_s.tr(" ", "_").squeeze("_")
     @path = Pathname.new(File.expand_path(name, Dir.pwd))
     @code_of_conduct = !!opts["code_of_conduct"]
   end

lib/jekyll/url.rb

@@ -72,9 +72,9 @@ module Jekyll
         break result if result.index(":").nil?
         if token.last.nil?
           # Remove leading "/" to avoid generating urls with `//`
-          result.gsub(%r!/:#{token.first}!, "")
+          result.gsub("/:#{token.first}", "")
         else
-          result.gsub(%r!:#{token.first}!, self.class.escape_path(token.last))
+          result.gsub(":#{token.first}", self.class.escape_path(token.last))
         end
       end
     end
@@ -109,14 +109,14 @@ module Jekyll
         replacement = self.class.escape_path(value)
 
         match.sub(":#{winner}", replacement)
-      end.gsub(%r!//!, "/".freeze)
+      end.squeeze("/")
     end
 
     # Returns a sanitized String URL, stripping "../../" and multiples of "/",
     # as well as the beginning "/" so we can enforce and ensure it.
 
     def sanitize_url(str)
-      "/" + str.gsub(%r!/{2,}!, "/").gsub(%r!\.+/|\A/+!, "")
+      "/#{str}".gsub("..", "/").gsub("./", "").squeeze("/")
     end
 
     # Escapes a path to be a valid URL path segment