afp_mimemagic 0.3.6

data/script/generate-mime.rb

@@ -0,0 +1,164 @@
+#!/usr/bin/env ruby
+
+require 'nokogiri'
+
+class String
+  alias inspect_old inspect
+
+  def inspect
+    x = b.inspect_old.gsub(/\\x([0-9a-f]{2})/i) do
+      '\\%03o' % Regexp.last_match(1).to_i(16)
+    end
+    x =~ /[\\']/ ? x : x.gsub('"', '\'')
+  end
+end
+
+def str2int(s)
+  return s.to_i(16) if s[0..1].downcase == '0x'
+  return s.to_i(8) if s[0..0].downcase == '0'
+
+  s.to_i(10)
+end
+
+def get_matches(parent)
+  parent.elements.map do |match|
+    if match['mask']
+      nil
+    else
+      type = match['type']
+      value = match['value']
+      offset = match['offset'].split(':').map { |x| x.to_i }
+      offset = offset.size == 2 ? offset[0]..offset[1] : offset[0]
+      case type
+      when 'string'
+        value.gsub!(/\\(x[\dA-Fa-f]{1,2}|0\d{1,3}|\d{1,3}|.)/) { eval("\"\\#{Regexp.last_match(1)}\"") }
+      when 'big16'
+        value = str2int(value)
+        value = ((value >> 8).chr + (value & 0xFF).chr)
+      when 'big32'
+        value = str2int(value)
+        value = (((value >> 24) & 0xFF).chr + ((value >> 16) & 0xFF).chr + ((value >> 8) & 0xFF).chr + (value & 0xFF).chr)
+      when 'little16'
+        value = str2int(value)
+        value = ((value & 0xFF).chr + (value >> 8).chr)
+      when 'little32'
+        value = str2int(value)
+        value = ((value & 0xFF).chr + ((value >> 8) & 0xFF).chr + ((value >> 16) & 0xFF).chr + ((value >> 24) & 0xFF).chr)
+      when 'host16' # use little endian
+        value = str2int(value)
+        value = ((value & 0xFF).chr + (value >> 8).chr)
+      when 'host32' # use little endian
+        value = str2int(value)
+        value = ((value & 0xFF).chr + ((value >> 8) & 0xFF).chr + ((value >> 16) & 0xFF).chr + ((value >> 24) & 0xFF).chr)
+      when 'byte'
+        value = str2int(value)
+        value = value.chr
+      end
+      children = get_matches(match)
+      children.empty? ? [offset, value] : [offset, value, children]
+    end
+  end.compact
+end
+
+if ARGV.size != 1
+  puts "Usage: #{$0} <freedesktop.org.xml>"
+  exit 1
+end
+
+FILE = ARGV[0]
+file = File.new(FILE)
+doc = Nokogiri::XML(file)
+extensions = {}
+types = {}
+magics = []
+(doc / 'mime-info/mime-type').each do |mime|
+  comments = Hash[*(mime / 'comment').map { |comment| [comment['xml:lang'], comment.inner_text] }.flatten]
+  type = mime['type']
+  subclass = (mime / 'sub-class-of').map { |x| x['type'] }
+  exts = (mime / 'glob').map { |x| x['pattern'] =~ /^\*\.([^\[\]]+)$/ ? Regexp.last_match(1).downcase : nil }.compact
+  (mime / 'magic').each do |magic|
+    priority = magic['priority'].to_i
+    matches = get_matches(magic)
+    magics << [priority, type, matches]
+  end
+  next if exts.empty?
+
+  exts.each do |x|
+    extensions[x] = type unless extensions.include?(x)
+  end
+  types[type] = [exts, subclass, comments[nil]]
+end
+
+magics = magics.sort { |a, b| [-a[0], a[1]] <=> [-b[0], b[1]] }
+
+common_types = [
+  'image/jpeg',                                                              # .jpg
+  'image/png',                                                               # .png
+  'image/gif',                                                               # .gif
+  'image/tiff',                                                              # .tiff
+  'image/bmp',                                                               # .bmp
+  'image/vnd.adobe.photoshop',                                               # .psd
+  'image/webp',                                                              # .webp
+  'image/svg+xml',                                                           # .svg
+
+  'video/x-msvideo',                                                         # .avi
+  'video/x-ms-wmv',                                                          # .wmv
+  'video/mp4',                                                               # .mp4, .m4v
+  'video/quicktime',                                                         # .mov
+  'video/mpeg',                                                              # .mpeg
+  'video/ogg',                                                               # .ogv
+  'video/webm',                                                              # .webm
+  'video/x-matroska',                                                        # .mkv
+  'video/x-flv',                                                             # .flv
+
+  'audio/mpeg',                                                              # .mp3
+  'audio/x-wav',                                                             # .wav
+  'audio/aac',                                                               # .aac
+  'audio/flac',                                                              # .flac
+  'audio/mp4',                                                               # .m4a
+  'audio/ogg',                                                               # .ogg
+
+  'application/pdf',                                                         # .pdf
+  'application/msword',                                                      # .doc
+  'application/vnd.openxmlformats-officedocument.wordprocessingml.document', # .docx
+  'application/vnd.ms-powerpoint',                                           # .pps
+  'application/vnd.openxmlformats-officedocument.presentationml.slideshow',  # .ppsx
+  'application/vnd.ms-excel',                                                # .pps
+  'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet'       # .ppsx
+]
+
+common_magics = common_types.map do |common_type|
+  magics.find { |_, type, _| type == common_type }
+end
+
+magics = (common_magics.compact + magics).uniq
+
+puts '# -*- coding: binary -*-'
+puts '# frozen_string_literal: true'
+puts "# Generated from #{FILE}"
+puts 'class MimeMagic'
+puts '  # @private'
+puts '  # :nodoc:'
+puts '  EXTENSIONS = {'
+extensions.keys.sort.each do |key|
+  puts "    '#{key}' => '#{extensions[key]}',"
+end
+puts '  }'
+puts '  # @private'
+puts '  # :nodoc:'
+puts '  TYPES = {'
+types.keys.sort.each do |key|
+  exts = types[key][0].join(' ')
+  parents = types[key][1].sort.join(' ')
+  comment = types[key][2].inspect
+  puts "    '#{key}' => [%w(#{exts}), %w(#{parents}), #{comment}],"
+end
+puts '  }'
+puts '  # @private'
+puts '  # :nodoc:'
+puts '  MAGIC = ['
+magics.each do |_priority, type, matches|
+  puts "    ['#{type}', #{matches.inspect}],"
+end
+puts '  ]'
+puts 'end'

data/test/files/application.x-ruby

@@ -0,0 +1,2 @@
+#!/usr/bin/ruby
+print "Hello World"

data/test/mimemagic_test.rb

@@ -0,0 +1,160 @@
+require 'minitest/autorun'
+require 'mimemagic'
+require 'stringio'
+require 'forwardable'
+
+class TestMimeMagic < Minitest::Test
+  # Do deep copy for constants of initial state.
+  INIT_EXTENSIONS = Marshal.load(Marshal.dump(MimeMagic::EXTENSIONS))
+  INIT_TYPES = Marshal.load(Marshal.dump(MimeMagic::TYPES))
+  INIT_MAGIC = Marshal.load(Marshal.dump(MimeMagic::MAGIC))
+
+  def setup
+    extentions = Marshal.load(Marshal.dump(INIT_EXTENSIONS))
+    types = Marshal.load(Marshal.dump(INIT_TYPES))
+    magic = Marshal.load(Marshal.dump(INIT_MAGIC))
+    MimeMagic.send(:remove_const, :EXTENSIONS) if MimeMagic.const_defined?(:EXTENSIONS)
+    MimeMagic.send(:remove_const, :TYPES) if MimeMagic.const_defined?(:TYPES)
+    MimeMagic.send(:remove_const, :MAGIC) if MimeMagic.const_defined?(:MAGIC)
+    MimeMagic.const_set('EXTENSIONS', extentions)
+    MimeMagic.const_set('TYPES', types)
+    MimeMagic.const_set('MAGIC', magic)
+  end
+
+  def test_have_type_mediatype_and_subtype
+    assert_equal 'text/html', MimeMagic.new('text/html').type
+    assert_equal 'text', MimeMagic.new('text/html').mediatype
+    assert_equal 'html', MimeMagic.new('text/html').subtype
+  end
+
+  def test_have_mediatype_helpers
+    assert MimeMagic.new('text/plain').text?
+    assert MimeMagic.new('text/html').text?
+    assert MimeMagic.new('application/xhtml+xml').text?
+    refute MimeMagic.new('application/octet-stream').text?
+    refute MimeMagic.new('image/png').text?
+    assert MimeMagic.new('image/png').image?
+    assert MimeMagic.new('video/ogg').video?
+    assert MimeMagic.new('audio/mpeg').audio?
+  end
+
+  def test_have_hierarchy
+    assert MimeMagic.new('text/html').child_of?('text/plain')
+    assert MimeMagic.new('text/x-java').child_of?('text/plain')
+  end
+
+  def test_have_extensions
+    assert_equal %w[htm html], MimeMagic.new('text/html').extensions
+  end
+
+  def test_have_comment
+    assert_equal 'HTML document', MimeMagic.new('text/html').comment
+  end
+
+  def test_recognize_extensions
+    assert_equal 'text/html', MimeMagic.by_extension('.html').to_s
+    assert_equal 'text/html', MimeMagic.by_extension('html').to_s
+    assert_equal 'text/html', MimeMagic.by_extension(:html).to_s
+    assert_equal 'application/x-ruby', MimeMagic.by_extension('rb').to_s
+    assert_nil MimeMagic.by_extension('crazy')
+    assert_nil MimeMagic.by_extension('')
+  end
+
+  def test_recognize_by_a_path
+    assert_equal 'text/html', MimeMagic.by_path('/adsjkfa/kajsdfkadsf/kajsdfjasdf.html').to_s
+    assert_equal 'text/html', MimeMagic.by_path('something.html').to_s
+    assert_equal 'application/x-ruby', MimeMagic.by_path('wtf.rb').to_s
+    assert_nil MimeMagic.by_path('where/am.html/crazy')
+    assert_nil MimeMagic.by_path('')
+  end
+
+  def test_recognize_xlsx_as_zip_without_magic
+    file = 'test/files/application.vnd.openxmlformats-officedocument.spreadsheetml.sheet'
+    %w[msoffice rubyxl gdocs].each do |variant|
+      file = "test/files/application.vnd.openxmlformats-officedocument.spreadsheetml{#{variant}}.sheet"
+      assert_equal 'application/zip', MimeMagic.by_magic(File.read(file)).to_s
+      assert_equal 'application/zip', MimeMagic.by_magic(File.open(file, 'rb')).to_s
+    end
+  end
+
+  def test_recognize_by_magic
+    load 'mimemagic/overlay.rb'
+    Dir['test/files/*'].each do |file|
+      mime = file[11..-1].sub('.', '/').sub(/\{\w+\}/, '')
+      assert_equal mime, MimeMagic.by_magic(File.read(file)).to_s
+      assert_equal mime, MimeMagic.by_magic(File.open(file, 'rb')).to_s
+    end
+  end
+
+  def test_recognize_all_by_magic
+    load 'mimemagic/overlay.rb'
+    %w[msoffice rubyxl gdocs].each do |variant|
+      file = "test/files/application.vnd.openxmlformats-officedocument.spreadsheetml{#{variant}}.sheet"
+      mimes = %w[application/vnd.openxmlformats-officedocument.spreadsheetml.sheet application/zip]
+      assert_equal mimes, MimeMagic.all_by_magic(File.read(file)).map(&:type)
+    end
+  end
+
+  def test_have_add
+    MimeMagic.add('application/mimemagic-test',
+                  extensions: %w[ext1 ext2],
+                  parents: 'application/xml',
+                  comment: 'Comment')
+    assert_equal 'application/mimemagic-test', MimeMagic.by_extension('ext1').to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_extension('ext2').to_s
+    assert_equal 'Comment', MimeMagic.by_extension('ext2').comment
+    assert_equal %w[ext1 ext2], MimeMagic.new('application/mimemagic-test').extensions
+    assert MimeMagic.new('application/mimemagic-test').child_of?('text/plain')
+  end
+
+  def test_process_magic
+    MimeMagic.add('application/mimemagic-test',
+                  magic: [[0, 'MAGICTEST'], # MAGICTEST at position 0
+                          [1, 'MAGICTEST'], # MAGICTEST at position 1
+                          [9..12, 'MAGICTEST'], # MAGICTEST starting at position 9 to 12
+                          [2, 'MAGICTEST', [[0, 'X'], [0, 'Y']]]]) # MAGICTEST at position 2 and (X at 0 or Y at 0)
+
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic('MAGICTEST').to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic('XMAGICTEST').to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(' MAGICTEST').to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic('123456789MAGICTEST').to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic('123456789ABMAGICTEST').to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic('123456789ABCMAGICTEST').to_s
+    assert_nil MimeMagic.by_magic('123456789ABCDMAGICTEST')
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic('X MAGICTEST').to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic('Y MAGICTEST').to_s
+    assert_nil MimeMagic.by_magic('Z MAGICTEST')
+
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(StringIO.new('MAGICTEST')).to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(StringIO.new('XMAGICTEST')).to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(StringIO.new(' MAGICTEST')).to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(StringIO.new('123456789MAGICTEST')).to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(StringIO.new('123456789ABMAGICTEST')).to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(StringIO.new('123456789ABCMAGICTEST')).to_s
+    assert_nil MimeMagic.by_magic(StringIO.new('123456789ABCDMAGICTEST'))
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(StringIO.new('X MAGICTEST')).to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(StringIO.new('Y MAGICTEST')).to_s
+    assert_nil MimeMagic.by_magic(StringIO.new('Z MAGICTEST'))
+  end
+
+  class IOObject
+    def initialize
+      @io = StringIO.new('MAGICTEST')
+    end
+
+    extend Forwardable
+    delegate %i[read size rewind eof? close] => :@io
+  end
+
+  class StringableObject
+    def to_s
+      'MAGICTEST'
+    end
+  end
+
+  def test_handle_different_file_objects
+    MimeMagic.add('application/mimemagic-test', magic: [[0, 'MAGICTEST']])
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(IOObject.new).to_s
+    assert_equal 'application/mimemagic-test', MimeMagic.by_magic(StringableObject.new).to_s
+  end
+end

metadata

@@ -0,0 +1,101 @@
+--- !ruby/object:Gem::Specification
+name: afp_mimemagic
+version: !ruby/object:Gem::Version
+  version: 0.3.6
+platform: ruby
+authors:
+- Daniel Mendler
+autorequire:
+bindir: bin
+cert_chain: []
+date: 2021-12-20 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: minitest
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.11'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.11'
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+description: Fast mime detection by extension or content in pure ruby (Uses freedesktop.org.xml
+  shared-mime-info database)
+email:
+- mail@daniel-mendler.de
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- ".gitignore"
+- ".travis.yml"
+- ".yardopts"
+- CHANGELOG.md
+- Gemfile
+- LICENSE
+- README.md
+- Rakefile
+- lib/mimemagic.rb
+- lib/mimemagic/overlay.rb
+- lib/mimemagic/tables.rb
+- lib/mimemagic/version.rb
+- mimemagic.gemspec
+- script/freedesktop.org.xml
+- script/generate-mime.rb
+- test/files/application.gzip
+- test/files/application.vnd.openxmlformats-officedocument.spreadsheetml{gdocs}.sheet
+- test/files/application.vnd.openxmlformats-officedocument.spreadsheetml{msoffice}.sheet
+- test/files/application.vnd.openxmlformats-officedocument.spreadsheetml{rubyxl}.sheet
+- test/files/application.x-bzip
+- test/files/application.x-ruby
+- test/files/application.x-tar
+- test/files/application.zip
+- test/files/image.jpeg
+- test/files/image.png
+- test/mimemagic_test.rb
+homepage: https://github.com/minad/mimemagic
+licenses:
+- GPL-2.0
+metadata:
+  changelog_uri: https://github.com/minad/mimemagic/blob/master/CHANGELOG.md
+  source_code_uri: https://github.com/minad/mimemagic
+  bug_tracker_uri: https://github.com/minad/mimemagic/issues
+post_install_message:
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.0.8
+signing_key:
+specification_version: 4
+summary: Fast mime detection by extension or content
+test_files: []