wahwah 0.1.0 → 1.0.0

data/lib/wahwah/flac_tag.rb

@@ -0,0 +1,84 @@
+# frozen_string_literal: true
+
+module WahWah
+  class FlacTag < Tag
+    include Ogg::VorbisComment
+    include Flac::StreaminfoBlock
+
+    TAG_ID = 'fLaC'
+
+    private
+      # FLAC structure:
+      #
+      # The four byte string "fLaC"
+      # The STREAMINFO metadata block
+      # Zero or more other metadata blocks
+      # One or more audio frames
+      def parse
+        # Flac file maybe contain ID3 header on the start, so skip it if exists
+        id3_header = ID3::V2Header.new(@file_io)
+        id3_header.valid? ? @file_io.seek(id3_header.size) : @file_io.rewind
+
+        return if @file_io.read(4) != TAG_ID
+
+        loop do
+          block = Flac::Block.new(@file_io)
+          parse_block(block)
+
+          break if block.is_last? || @file_io.eof?
+        end
+      end
+
+      def parse_block(block)
+        return unless block.valid?
+
+        case block.type
+        when 'STREAMINFO'
+          parse_streaminfo_block(block.data)
+        when 'VORBIS_COMMENT'
+          parse_vorbis_comment(block.data)
+        when 'PICTURE'
+          parse_picture_block(block.data)
+        else
+          @file_io.seek(block.size, IO::SEEK_CUR)
+        end
+      end
+
+      # PICTURE block data structure:
+      #
+      # Length(bit)  Meaning
+      #
+      # 32           The picture type according to the ID3v2 APIC frame:
+      #
+      # 32           The length of the MIME type string in bytes.
+      #
+      # n*8          The MIME type string.
+      #
+      # 32           The length of the description string in bytes.
+      #
+      # n*8          The description of the picture, in UTF-8.
+      #
+      # 32           The width of the picture in pixels.
+      #
+      # 32           The height of the picture in pixels.
+      #
+      # 32           The color depth of the picture in bits-per-pixel.
+      #
+      # 32           For indexed-color pictures (e.g. GIF), the number of colors used, or 0 for non-indexed pictures.
+      #
+      # 32           The length of the picture data in bytes.
+      #
+      # n*8          The binary picture data.
+      def parse_picture_block(block_data)
+        block_content = StringIO.new(block_data)
+
+        type_index, mime_type_length = block_content.read(8).unpack('NN')
+        mime_type = Helper.encode_to_utf8(block_content.read(mime_type_length))
+        description_length = block_content.read(4).unpack('N').first
+        data_length = block_content.read(description_length + 20).unpack("#{'x' * (description_length + 16)}N").first
+        data = block_content.read(data_length)
+
+        @images.push({ data: data, mime_type: mime_type, type: ID3::ImageFrameBody::TYPES[type_index] })
+      end
+  end
+end

data/lib/wahwah/helper.rb

@@ -0,0 +1,37 @@
+# frozen_string_literal: true
+
+module WahWah
+  module Helper
+    def self.encode_to_utf8(string, source_encoding: '')
+      encoded_string = source_encoding.empty? ?
+        string.force_encoding('utf-8') :
+        string.encode('utf-8', source_encoding, invalid: :replace, undef: :replace, replace: '')
+
+      encoded_string.valid_encoding? ? encoded_string.strip : ''
+    end
+
+    # ID3 size is encoded with four bytes where may the most significant
+    # bit (bit 7) is set to zero in every byte,
+    # making a total of 28 bits. The zeroed bits are ignored
+    def self.id3_size_caculate(bits_string, has_zero_bit: true)
+      if has_zero_bit
+        bits_string.scan(/.{8}/).map { |byte_string| byte_string[1..-1] }.join.to_i(2)
+      else
+        bits_string.to_i(2)
+      end
+    end
+
+    def self.split_with_terminator(string, terminator_size)
+      string.split(Regexp.new(('\x00' * terminator_size).b), 2)
+    end
+
+    def self.file_format(file_path)
+      File.extname(file_path).downcase.delete('.')
+    end
+
+    def self.byte_string_to_guid(byte_string)
+      guid = byte_string.unpack('NnnA*').pack('VvvA*').unpack('H*').first
+      [guid[0..7], guid[8..11], guid[12..15], guid[16..19], guid[20..-1]].join('-').upcase
+    end
+  end
+end

data/lib/wahwah/id3/comment_frame_body.rb

@@ -0,0 +1,21 @@
+# frozen_string_literal: true
+
+module WahWah
+  module ID3
+    class CommentFrameBody < FrameBody
+      # Comment frame body structure:
+      #
+      # Text encoding             $xx
+      # Language                  $xx xx xx
+      # Short content description <textstring> $00 (00)
+      # The actual text           <textstring>
+      def parse
+        encoding_id, _language, reset_content = @content.unpack('CA3a*')
+        encoding = ENCODING_MAPPING[encoding_id]
+        _description, comment_text = Helper.split_with_terminator(reset_content, ENCODING_TERMINATOR_SIZE[encoding])
+
+        @value = Helper.encode_to_utf8(comment_text, source_encoding: encoding)
+      end
+    end
+  end
+end

data/lib/wahwah/id3/frame.rb

@@ -0,0 +1,180 @@
+# frozen_string_literal: true
+
+require 'zlib'
+
+module WahWah
+  module ID3
+    class Frame
+      ID_MAPPING = {
+        # ID3v2.2 frame id
+        COM: :comment,
+        TRK: :track,
+        TYE: :year,
+        TAL: :album,
+        TP1: :artist,
+        TT2: :title,
+        TCO: :genre,
+        TPA: :disc,
+        TP2: :albumartist,
+        TCM: :composer,
+        PIC: :image,
+
+        # ID3v2.3 and ID3v2.4 frame id
+        COMM: :comment,
+        TRCK: :track,
+        TYER: :year,
+        TALB: :album,
+        TPE1: :artist,
+        TIT2: :title,
+        TCON: :genre,
+        TPOS: :disc,
+        TPE2: :albumartist,
+        TCOM: :composer,
+        APIC: :image,
+
+        # ID3v2.4 use TDRC replace TYER
+        TDRC: :year
+      }
+
+      # ID3v2.3 frame flags field is defined as follows.
+      #
+      # %abc00000 %ijk00000
+      #
+      # a - Tag alter preservation
+      # b - File alter preservation
+      # c - Read only
+      # i - Compression
+      # j - Encryption
+      # k - Grouping identity
+      V3_HEADER_FLAGS_INDICATIONS = Array.new(16).tap do |array|
+        array[0] = :tag_alter_preservation
+        array[1] = :file_alter_preservation
+        array[2] = :read_only
+        array[8] = :compression
+        array[9] = :encryption
+        array[10] = :grouping_identity
+      end
+
+      # ID3v2.4 frame flags field is defined as follows.
+      #
+      # %0abc0000 %0h00kmnp
+      #
+      # a - Tag alter preservation
+      # b - File alter preservation
+      # c - Read only
+      # h - Grouping identity
+      # k - Compression
+      # m - Encryption
+      # n - Unsynchronisation
+      # p - Data length indicator
+      V4_HEADER_FLAGS_INDICATIONS = Array.new(16).tap do |array|
+        array[1] = :tag_alter_preservation
+        array[2] = :file_alter_preservation
+        array[3] = :read_only
+        array[9] = :grouping_identity
+        array[12] = :compression
+        array[13] = :encryption
+        array[14] = :unsynchronisation
+        array[15] = :data_length_indicator
+      end
+
+      attr_reader :name, :value
+
+      def initialize(file_io, version)
+        @file_io = file_io
+        @version = version
+
+        parse_frame_header
+
+        # In ID3v2.3 when frame is compressed using zlib
+        # with 4 bytes for 'decompressed size' appended to the frame header.
+        #
+        # In ID3v2.4 A 'Data Length Indicator' byte MUST be included in the frame
+        # when frame is compressed, and 'Data Length Indicator'represented as a 32 bit
+        # synchsafe integer
+        #
+        # So skip those 4 byte.
+        if compressed? || data_length_indicator?
+          @file_io.seek(4, IO::SEEK_CUR)
+          @size = @size - 4
+        end
+
+        parse_body
+      end
+
+      def valid?
+        @size > 0 && !@name.nil?
+      end
+
+      def compressed?
+        @flags.include? :compression
+      end
+
+      def data_length_indicator?
+        @flags.include? :data_length_indicator
+      end
+
+
+      private
+        # ID3v2.2 frame header structure:
+        #
+        # Frame ID      $xx xx xx(tree characters)
+        # Size          3 * %xxxxxxxx
+        #
+        # ID3v2.3 frame header structure:
+        #
+        # Frame ID      $xx xx xx xx (four characters)
+        # Size          4 * %xxxxxxxx
+        # Flags         $xx xx
+        #
+        # ID3v2.4 frame header structure:
+        #
+        # Frame ID      $xx xx xx xx (four characters)
+        # Size          4 * %0xxxxxxx
+        # Flags         $xx xx
+        def parse_frame_header
+          header_size = @version == 2 ? 6 : 10
+          header_formate = @version == 2 ? 'A3B24' : 'A4B32B16'
+          id, size_bits, flags_bits = @file_io.read(header_size).unpack(header_formate)
+
+          @name = ID_MAPPING[id.to_sym]
+          @size = Helper.id3_size_caculate(size_bits, has_zero_bit: @version == 4)
+          @flags = parse_flags(flags_bits)
+        end
+
+        def parse_flags(flags_bits)
+          return [] if flags_bits.nil?
+
+          frame_flags_indications = @version == 4 ?
+            V4_HEADER_FLAGS_INDICATIONS :
+            V3_HEADER_FLAGS_INDICATIONS
+
+          flags_bits.split('').map.with_index do |flag_bit, index|
+            frame_flags_indications[index] if flag_bit == '1'
+          end.compact
+        end
+
+        def parse_body
+          return unless @size > 0
+          (@file_io.seek(@size, IO::SEEK_CUR); return) if @name.nil?
+
+          content = compressed? ? Zlib.inflate(@file_io.read(@size)) : @file_io.read(@size)
+          frame_body = frame_body_class.new(content, @version)
+          @value = frame_body.value
+        end
+
+        def frame_body_class
+          case @name
+          when :comment
+            CommentFrameBody
+          when :genre
+            GenreFrameBody
+          when :image
+            ImageFrameBody
+          else
+            TextFrameBody
+          end
+        end
+    end
+  end
+end

data/lib/wahwah/id3/frame_body.rb

@@ -0,0 +1,36 @@
+# frozen_string_literal: true
+
+module WahWah
+  module ID3
+    class FrameBody
+      # Textual frames are marked with an encoding byte.
+      #
+      # $00   ISO-8859-1 [ISO-8859-1]. Terminated with $00.
+      # $01   UTF-16 [UTF-16] encoded Unicode [UNICODE] with BOM.
+      # $02   UTF-16BE [UTF-16] encoded Unicode [UNICODE] without BOM.
+      # $03   UTF-8 [UTF-8] encoded Unicode [UNICODE].
+      ENCODING_MAPPING = %w(ISO-8859-1 UTF-16 UTF-16BE UTF-8)
+
+      ENCODING_TERMINATOR_SIZE = {
+        'ISO-8859-1' => 1,
+        'UTF-16' => 2,
+        'UTF-16BE' => 2,
+        'UTF-8' => 1
+      }
+
+
+      attr_reader :value
+
+      def initialize(content, version)
+        @content = content
+        @version = version
+
+        parse
+      end
+
+      def parse
+        raise WahWahNotImplementedError, 'The parse method is not implemented'
+      end
+    end
+  end
+end

data/lib/wahwah/id3/genre_frame_body.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+module WahWah
+  module ID3
+    class GenreFrameBody < TextFrameBody
+      def parse
+        super
+
+        # If value is numeric value, or contain numeric value in parens
+        # can use as index for ID3v1 genre list
+        @value = ID3::V1::GENRES[$1.to_i] if @value =~ /^\((\d+)\)$/ || @value =~ /^(\d+)$/
+      end
+    end
+  end
+end

data/lib/wahwah/id3/image_frame_body.rb

@@ -0,0 +1,60 @@
+# frozen_string_literal: true
+
+module WahWah
+  module ID3
+    class ImageFrameBody < FrameBody
+      TYPES = %i(
+        other
+        file_icon
+        other_file_icon
+        cover_front
+        cover_back
+        leaflet
+        media
+        lead_artist
+        artist
+        conductor
+        band
+        composer
+        lyricist
+        recording_location
+        during_recording
+        during_performance
+        movie_screen_capture
+        bright_coloured_fish
+        illustration
+        band_logotype
+        publisher_logotype
+      )
+
+      def mime_type
+        mime_type = @mime_type.downcase.yield_self { |type| type == 'jpg' ? 'jpeg' : type }
+        @version > 2 ? mime_type : "image/#{mime_type}"
+      end
+
+      # ID3v2.2 image frame structure:
+      #
+      # Text encoding  $xx
+      # Image format   $xx xx xx
+      # Picture type   $xx
+      # Description    <text string according to encoding> $00 (00)
+      # Picture data   <binary data>
+      #
+      # ID3v2.3 and ID3v2.4 image frame structure:
+      #
+      # Text encoding $xx
+      # MIME type     <text string> $00
+      # Picture type  $xx
+      # Description   <text string according to encoding> $00 (00)
+      # Picture data  <binary data>
+      def parse
+        frame_format = @version > 2 ? 'CZ*Ca*' : 'Ca3Ca*'
+        encoding_id, @mime_type, type_index, reset_content = @content.unpack(frame_format)
+        encoding = ENCODING_MAPPING[encoding_id]
+        _description, data = Helper.split_with_terminator(reset_content, ENCODING_TERMINATOR_SIZE[encoding])
+
+        @value = { data: data, mime_type: mime_type, type: TYPES[type_index] }
+      end
+    end
+  end
+end

data/lib/wahwah/id3/text_frame_body.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+
+module WahWah
+  module ID3
+    class TextFrameBody < FrameBody
+      # Text frame boby structure:
+      #
+      # Text encoding  $xx
+      # Information    <text string according to encoding>
+      def parse
+        encoding_id, text = @content.unpack('Ca*')
+        @value = Helper.encode_to_utf8(text, source_encoding: ENCODING_MAPPING[encoding_id])
+      end
+    end
+  end
+end