mime-types 2.0

A substantial upgrade to the mime-types library. Major features:

- No longer compatible with Ruby 1.8.
- Several major API changes and deprecations.
- The default mime-types registry is now stored as JSON.
- Improved developer tools.
- Improved repackager tools.

See History.rdoc for full details.

1 files changed, 42 insertions, 185 deletions

diff --git a/Rakefile b/Rakefile
index bc2c25a..01e5c60 100644
--- a/Rakefile
+++ b/Rakefile
@@ -14,6 +14,9 @@ Hoe.plugin :travis
 
 spec = Hoe.spec 'mime-types' do
   developer('Austin Ziegler', 'austin@rubyforge.org')
+  self.need_tar = true
+
+  self.require_ruby_version '>= 1.9.2'
 
   self.remote_rdoc_dir = '.'
   self.rsync_args << ' --exclude=statsvn/'
@@ -31,202 +34,34 @@ spec = Hoe.spec 'mime-types' do
   self.extra_dev_deps << ['hoe-travis', '~> 1.2']
   self.extra_dev_deps << ['minitest', '~> 4.5']
   self.extra_dev_deps << ['rake', '~> 10.0']
-end
-
-def reload_mime_types(repeats = 1, force_load = false)
-  repeats.times {
-    Object.send(:remove_const, :MIME) if defined? MIME
-    load 'lib/mime/types.rb'
-    MIME::Types.send(:__types__) if force_load
-  }
+  self.extra_dev_deps << ['simplecov', '~> 0.7']
 end
 
 desc 'Benchmark'
 task :benchmark, :repeats do |t, args|
-  repeats = args.repeats.to_i
-  repeats = 50 if repeats.zero?
-  require 'benchmark'
-  $LOAD_PATH.unshift 'lib'
-
-  cache_file = File.expand_path('../cache.mtx', __FILE__)
-  rm cache_file if File.exist? cache_file
-
-  Benchmark.bm(17) do |x|
-    x.report("Normal:") { reload_mime_types repeats }
-
-    ENV['RUBY_MIME_TYPES_LAZY_LOAD'] = 'yes'
-    x.report("Lazy:") { reload_mime_types repeats }
-    x.report("Lazy+Load:") { reload_mime_types repeats, true }
+  $LOAD_PATH.unshift('support')
+  require 'benchmarker'
 
-    ENV.delete('RUBY_MIME_TYPES_LAZY_LOAD')
-
-    ENV['RUBY_MIME_TYPES_CACHE'] = cache_file
-    reload_mime_types
-
-    x.report("Cached:") { reload_mime_types repeats }
-    ENV['RUBY_MIME_TYPES_LAZY_LOAD'] = 'yes'
-    x.report("Lazy Cached:") { reload_mime_types repeats }
-    x.report("Lazy Cached Load:") { reload_mime_types repeats, true }
+  Benchmarker.benchmark(args.repeats)
+end
 
+namespace :test do
+  task :coverage do
+    spec.test_prelude = [
+      'require "simplecov"',
+      'SimpleCov.start("test_frameworks") { command_name "Minitest" }',
+      'gem "minitest"'
+    ].join('; ')
+    Rake::Task['test'].execute
   end
-
-  rm cache_file
 end
 
 namespace :mime do
   desc "Download the current MIME type registrations from IANA."
-  task :iana, :save, :destination do |t, args|
-    save_type = (args.save || :text).to_sym
-
-    case save_type
-    when :text, :both, :html
-      nil
-    else
-      raise "Unknown save type provided. Must be one of text, both, or html."
-    end
-
-    destination = args.destination || "type-lists"
-
-    require 'open-uri'
-    require 'nokogiri'
-    require 'cgi'
-
-    class IANAParser
-      include Comparable
-
-      INDEX = %q(http://www.iana.org/assignments/media-types/)
-      CONTACT_PEOPLE = %r{http://www.iana.org/assignments/contact-people.html?#(.*)}
-      RFC_EDITOR = %r{http://www.rfc-editor.org/rfc/rfc(\d+).txt}
-      IETF_RFC = %r{http://www.ietf.org/rfc/rfc(\d+).txt}
-      IETF_RFC_TOOLS = %r{http://tools.ietf.org/html/rfc(\d+)}
-
-      class << self
-        def load_index
-          @types ||= {}
-
-          Nokogiri::HTML(open(INDEX) { |f| f.read }).xpath('//p/a').each do |tag|
-            href_match = %r{^/assignments/media-types/(.+)/$}.match(tag['href'])
-            next if href_match.nil?
-            type = href_match.captures[0]
-            @types[tag.content] = IANAParser.new(tag.content, type)
-          end
-        end
-
-        attr_reader :types
-      end
-
-      def initialize(name, type)
-        @name = name
-        @type = type
-        @url  = File.join(INDEX, @type)
-      end
-
-      attr_reader :name
-      attr_reader :type
-      attr_reader :url
-      attr_reader :html
-
-      def download(name = nil)
-        @html = Nokogiri::HTML(open(name || @url) { |f| f.read })
-      end
-
-      def save_html
-        File.open("#@name.html", "wb") { |w| w.write @html }
-      end
-
-      def <=>(o)
-        self.name <=> o.name
-      end
-
-      def parse
-        nodes = html.xpath("//table//table//tr")
-
-        # How many <td> children does the first node have?
-        node_count = nodes.first.children.select { |n| n.elem? }.size
-
-        if node_count == 1
-          # The title node doesn't have what we expect. Let's try it based
-          # on the first real node.
-          node_count = nodes.first.next.children.select { |n| n.elem? }.size
-        end
-
-        @mime_types = nodes.map do |node|
-          next if node == nodes.first
-          elems = node.children.select { |n| n.elem? }
-          next if elems.size.zero?
-
-          if elems.size != node_count
-            warn "size mismatch (#{elems.size} != #{node_count}) in node: #{node}"
-            next
-          end
-
-          case elems.size
-          when 3
-            subtype_index = 1
-            refnode_index = 2
-          when 4
-            subtype_index = 1
-            refnode_index = 3
-          else
-            raise "Unknown element size."
-          end
-
-          subtype   = elems[subtype_index].content.chomp.strip
-          refnodes  = elems[refnode_index].children.select { |n| n.elem? }.map { |ref|
-            case ref['href']
-            when CONTACT_PEOPLE
-              tag = CGI::unescape($1).chomp.strip
-              if tag == ref.content
-                "[#{ref.content}]"
-              else
-                "[#{ref.content}=#{tag}]"
-              end
-            when RFC_EDITOR, IETF_RFC, IETF_RFC_TOOLS
-              "RFC#$1"
-            when %r{(https?://.*)}
-              "{#{ref.content}=#$1}"
-            else
-              ref
-            end
-          }
-          refs = refnodes.join(',')
-
-          "#@type/#{subtype} 'IANA,#{refs}"
-        end.compact
-
-        @mime_types
-      end
-
-      def save_text
-        File.open("#@name.txt", "wb") { |w| w.write @mime_types.join("\n") }
-      end
-    end
-
-    puts "Downloading index of MIME types from #{IANAParser::INDEX}."
-    IANAParser.load_index
-
-    require 'fileutils'
-    FileUtils.mkdir_p destination
-    Dir.chdir destination do
-      IANAParser.types.values.sort.each do |parser|
-        next if parser.name == "example" or parser.name == "mime"
-        puts "Downloading #{parser.name} from #{parser.url}"
-        parser.download
-
-        if :html == save_type || :both == save_type
-          puts "Saving #{parser.name}.html"
-          parser.save_html
-        end
-
-        if :text == save_type || :both == save_type
-          puts "Parsing #{parser.name} HTML"
-          parser.parse
-
-          puts "Saving #{parser.name}.txt"
-          parser.save_text
-        end
-      end
-    end
+  task :iana, :destination do |t, args|
+    $LOAD_PATH.unshift('support')
+    require 'iana_downloader'
+    IANADownloader.download_to(args.destination)
   end
 
   desc "Shows known MIME type sources."
@@ -242,4 +77,26 @@ http://www.iana.org/assignments/media-types/
   end
 end
 
+Rake::Task['gem'].prerequisites.unshift("convert:yaml:json")
+
+namespace :convert do
+  namespace :yaml do
+    desc "Convert from YAML to JSON"
+    task :json, :source, :destination, :multiple_files do |t, args|
+      $LOAD_PATH.unshift('support')
+      require 'convert'
+      Convert.from_yaml_to_json(args)
+    end
+  end
+
+  namespace :json do
+    desc "Convert from JSON to YAML"
+    task :yaml, :source, :destination, :multiple_files do |t, args|
+      $LOAD_PATH.unshift('support')
+      require 'convert'
+      Convert.from_json_to_yaml(args)
+    end
+  end
+end
+
 # vim: syntax=ruby