ruby-mp3info 0.6.16 → 0.7

data/History.txt

@@ -1,3 +1,9 @@
+=== 0.7 / 2012-02-29
+
+* removed iconv for ruby >= 1.9
+* default encoding for decoded and encoded tags is UTF-8 now (removed :encoding params for ID3v2)
+* come back to a non senseless version numbering
+
 === 0.6.16 / 2011-11-10
 
 * fixed type error when inspecting mp3info (thanks to Jacob Lichner)

data/Manifest.txt

@@ -2,7 +2,6 @@ History.txt
 Manifest.txt
 README.txt
 Rakefile
-install.rb
 lib/mp3info.rb
 lib/mp3info/extension_modules.rb
 lib/mp3info/id3v2.rb

data/README.txt

@@ -14,9 +14,12 @@ mp3 files.
 * read, write, remove id3v1 and id3v2 tags
 * correctly read VBR files (with or without Xing header)
 * only 2.3 version is supported for writings id3v2 tags
+* id3v2 tags are always written in UTF-16 encoding
 
 == SYNOPSIS:
 
+  ### read and display infos & tags
+
   require "mp3info"
   # read and display infos & tags
   Mp3Info.open("myfile.mp3") do |mp3info|
@@ -35,6 +38,10 @@ mp3 files.
     mp3.tag.artist = "artist name"
   end
 
+  # tags are written when calling Mp3Info#close or at the end of the #open block
+
+  ### access id3v2 tags
+
   Mp3Info.open("myfile.mp3") do |mp3|
     # you can access four letter v2 tags like this
     puts mp3.tag2.TIT2
@@ -47,21 +54,12 @@ mp3 files.
     mp3.tag2.COMM = "my comment in french, correctly handled when reading and writing"
   end
 
-  # tags v2 will be read and written according to the :encoding settings
-  mp3 = Mp3Info.open("myfile.mp3", :encoding => 'utf-8')
-
 == REQUIREMENTS:
 
 * iconv
 
 == INSTALL:
 
-$ ruby install.rb config
-$ ruby install.rb setup
-# ruby install.rb install
-
- or
-
 * gem install ruby-mp3info
 
 == DEVELOPERS:

data/Rakefile

@@ -5,7 +5,7 @@ require 'hoe'
 Hoe.plugin :yard
 
 Hoe.spec('ruby-mp3info') do 
-  developer "Guillaume Pierronnet", "moumar@rubyforge.org"
+  developer "Guillaume Pierronnet", "guillaume.pierronnet@gmail.com"
   remote_rdoc_dir = ''
   rdoc_locations << "rubyforge.org:/var/www/gforge-projects/ruby-mp3info/"
 end

data/lib/mp3info.rb

@@ -4,6 +4,7 @@
 # Website:: http://ruby-mp3info.rubyforge.org/
 
 require "fileutils"
+require "stringio"
 require "mp3info/extension_modules"
 require "mp3info/id3v2"
 
@@ -17,7 +18,7 @@ end
 
 class Mp3Info
 
-  VERSION = "0.6.16"
+  VERSION = "0.7"
 
   LAYER = [ nil, 3, 2, 1]
   BITRATE = {
@@ -219,6 +220,9 @@ class Mp3Info
   def initialize(filename_or_io, options = {})
     warn("#{self.class}::new() does not take block; use #{self.class}::open() instead") if block_given?
     @filename_or_io = filename_or_io
+    if @filename_or_io.nil?
+      raise ArgumentError, "filename is nil"
+    end
     options = {:parse_mp3 => true, :parse_tags => true}.update(options)
     @tag_parsing_enabled = options.delete(:parse_tags)
     @mp3_parsing_enabled = options.delete(:parse_mp3)
@@ -230,22 +234,21 @@ class Mp3Info
   def reload
     @header = {}
 
-    if @filename_or_io.is_a?(String)
-      @io_is_a_file = true
-      @io = File.new(@filename_or_io, "rb")
-      @io_size = @io.stat.size
-      @filename = @filename_or_io
-    elsif @filename_or_io.is_a?(StringIO)
+    if @filename_or_io.is_a?(StringIO)
       @io_is_a_file = false
       @io = @filename_or_io
       @io_size = @io.size
       @filename = nil
+    else
+      @io_is_a_file = true
+      @io = File.new(@filename_or_io, "rb")
+      @io_size = @io.stat.size
+      @filename = @filename_or_io
     end
 
     if @io_size == 0
       raise(Mp3InfoError, "empty file or IO")
     end
-    
 
     @io.extend(Mp3FileMethods)
     @tag1 = @tag = @tag1_orig = @tag_orig = {}
@@ -554,9 +557,10 @@ private
   ### assumes @io is pointing to char after "TAG" id
   def gettag1
     @tag1_parsed = true
-    @tag1["title"] = @io.read(30).unpack("A*").first
-    @tag1["artist"] = @io.read(30).unpack("A*").first
-    @tag1["album"] = @io.read(30).unpack("A*").first
+    %w{title artist album}.each do |tag|
+      v = @io.read(30).unpack("A*").first
+      @tag1[tag] = Mp3Info::EncodingHelper.convert_from_iso_8859_1(v)
+    end
     year_t = @io.read(4).to_i
     @tag1["year"] = year_t unless year_t == 0
     comments = @io.read(30)
@@ -564,7 +568,8 @@ private
       @tag1["tracknum"] = comments.getbyte(-1).to_i
       comments.chop! #remove the last char
     end
-    @tag1["comments"] = comments.unpack("A*").first
+    comment = comments.unpack("A*").first
+    @tag1["comments"] = Mp3Info::EncodingHelper.convert_from_iso_8859_1(comment)
     @tag1["genre"] = @io.getbyte
     @tag1["genre_s"] = GENRES[@tag1["genre"]] || ""
 

data/lib/mp3info/extension_modules.rb

@@ -21,10 +21,6 @@ class Mp3Info
   class ::String
     if RUBY_VERSION < "1.9.0"
       alias getbyte []
-    else
-      def getbyte(i)
-        self[i].ord unless self[i].nil?
-      end
     end
   end
 
@@ -43,4 +39,50 @@ class Mp3Info
       (getbyte << 21) + (getbyte << 14) + (getbyte << 7) + getbyte
     end                 
   end
+
+  class EncodingHelper
+    def self.convert_to(value, from, to)
+      if RUBY_1_8
+        if to == "iso-8859-1"
+          to = to + "//TRANSLIT"
+        end
+        ruby_18_encode(from, to, value)
+      else
+        if to == "utf-16"
+          ("\uFEFF" +  value).encode("UTF-16BE")
+        else
+          value.encode(to)
+        end
+      end
+    end
+
+    def self.convert_from_iso_8859_1(value)
+      if RUBY_1_8
+        ruby_18_encode("utf-8", "iso-8859-1", value)
+      else
+        value.force_encoding("iso-8859-1").encode("utf-8")
+      end
+    end
+
+    def self.ruby_18_encode(from, to, value)
+      begin
+        Iconv.iconv(to, from, value).first
+      rescue Iconv::Failure
+        value
+      end
+    end
+
+    def self.decode_utf16(out)
+      if RUBY_1_8
+        convert_to(out, "UTF-8", "UTF-16")
+      else
+        if out.bytes.first == 0xff
+          tag_encoding = "UTF-16LE"
+        else
+          tag_encoding = "UTF-16BE"
+        end
+        out = out.dup.force_encoding(tag_encoding)[1..-1]
+      end
+    end
+  end
 end

data/lib/mp3info/id3v2.rb

@@ -1,10 +1,16 @@
-# coding:utf-8
+# encoding: utf-8
 # License:: Ruby
 # Author:: Guillaume Pierronnet (mailto:moumar_AT__rubyforge_DOT_org)
 # Website:: http://ruby-mp3info.rubyforge.org/
 
 require "delegate"
-require "iconv"
+
+if RUBY_VERSION[0..2] == "1.8"
+  require "iconv"
+  RUBY_1_8 = true
+else
+  RUBY_1_8 = false
+end
 
 require "mp3info/extension_modules"
 
@@ -16,9 +22,6 @@ class ID3v2Error < StandardError ; end
 # Written version is always 2.3.
 class ID3v2 < DelegateClass(Hash) 
   
-  # Major version used when writing tags
-  WRITE_VERSION = 3
-  
   TAGS = {
     "AENC" => "Audio encryption",
     "APIC" => "Attached picture",
@@ -169,26 +172,22 @@ class ID3v2 < DelegateClass(Hash)
 
   # :+lang+: for writing comments
   #
-  # :+encoding+: one of the string of +TEXT_ENCODINGS+, 
-  # used as a source and destination encoding respectively
-  # for read and write tag2 values.
+  # [DEPRECATION] :+encoding+: one of the string of +TEXT_ENCODINGS+, 
+  # use of :encoding parameter is DEPRECATED. In ruby 1.8, use utf-8 encoded strings for tags.
+  # In ruby >= 1.9, strings are automatically transcoded from their originaloriginal  encoding.
   attr_reader :options
   
   # possible options are described above ('options' attribute)
   # you can access this object like an hash, with [] and []= methods
   # special cases are ["disc_number"] and ["disc_total"] mirroring TPOS attribute
   def initialize(options = {})
-    @options = { 
-      :lang => "ENG",
-      :encoding => "iso-8859-1"
-    }
+    @options = { :lang => "ENG" }
+    if @options[:encoding]
+      warn("use of :encoding parameter is DEPRECATED. In ruby 1.8, use utf-8 encoded strings for tags.\n" +
+           "In ruby >= 1.9, strings are automatically transcoded from their original encoding.")
+    end
 
     @options.update(options)
-    @text_encoding_index = TEXT_ENCODINGS.index(@options[:encoding])
-    
-    unless @text_encoding_index
-      raise(ArgumentError, "bad id3v2 text encoding specified")
-    end
 
     @hash = {}
     #TAGS.keys.each { |k| @hash[k] = nil }
@@ -250,7 +249,7 @@ class ID3v2 < DelegateClass(Hash)
     @io = nil
   end
 
-  # dump tag for writing. Version is always 2.#{WRITE_VERSION}.0.
+  # dump tag for writing. Version is always 2.3.0
   def to_bin
     #TODO handle of @tag2[TLEN"]
     #TODO add of crc
@@ -269,13 +268,13 @@ class ID3v2 < DelegateClass(Hash)
       
       # Output one flag for each array element, or one only if it's not an array
       [v].flatten.each do |value|
-        data = encode_tag(k, value.to_s, WRITE_VERSION)
+        data = encode_tag(k, value.to_s)
         #data << "\x00"*2 #End of tag
 
         tag << k[0,4]   #4 characte max for a tag's key
         #tag << to_syncsafe(data.size) #+1 because of the language encoding byte
         size = data.size
-        if RUBY_VERSION >= "1.9.0"
+        unless RUBY_1_8
           size = data.dup.force_encoding("binary").size
         end
         tag << [size].pack("N") #+1 because of the language encoding byte
@@ -286,7 +285,7 @@ class ID3v2 < DelegateClass(Hash)
 
     tag_str = "ID3"
     #version_maj, version_min, unsync, ext_header, experimental, footer 
-    tag_str << [ WRITE_VERSION, 0, "0000" ].pack("CCB4")
+    tag_str << [ 3, 0, "0000" ].pack("CCB4")
     tag_str << [to_syncsafe(tag.size)].pack("N")
     tag_str << tag
     puts "tag in binary format: #{tag_str.inspect}" if $DEBUG
@@ -295,60 +294,83 @@ class ID3v2 < DelegateClass(Hash)
 
   private
 
-  def encode_tag(name, value, version)
-    puts "encode_tag(#{name.inspect}, #{value.inspect}, #{version})" if $DEBUG
-
-    text_encoding_index = @text_encoding_index 
-    if (name.index("T") == 0 || name == "COMM" ) && version == 3
-      # in id3v2.3 tags, there is only 2 encodings possible
-      transcoded_value = value
-      if text_encoding_index >= 2
-        begin
-	  transcoded_value = Iconv.iconv(TEXT_ENCODINGS[1], 
-					 TEXT_ENCODINGS[text_encoding_index], 
-					 value).first
-        rescue Iconv::Failure
-	  transcoded_value = value
-	end
-	text_encoding_index = 1
-      end
-    end
+  def encode_tag(name, value)
+    puts "encode_tag(#{name.inspect}, #{value.inspect})" if $DEBUG
+    name = name.to_s
 
+    if name =~ /^(COM|T)/
+      transcoded_value = Mp3Info::EncodingHelper.convert_to(value, "utf-8", "utf-16")
+    end
     case name
       when "COMM"
-      puts "encode COMM: enc: #{text_encoding_index}, lang: #{@options[:lang]}, str: #{transcoded_value.dump}" if $DEBUG
-	[ text_encoding_index , @options[:lang], 0, transcoded_value ].pack("ca3ca*")
+        puts "encode COMM: lang: #{@options[:lang]}, value #{transcoded_value.inspect}" if $DEBUG
+	s = [ 1, @options[:lang], "\xFE\xFF\x00\x00", transcoded_value].pack("ca3a*a*")
+	return s
       when /^T/
-	text_encoding_index.chr + transcoded_value
+        unless RUBY_1_8
+          transcoded_value.force_encoding("BINARY")
+        end
+	return "\x01" + transcoded_value
       else
-        value
+        return value
     end
   end
 
   ### Read a tag from file and perform UNICODE translation if needed
   def decode_tag(name, raw_value)
     puts("decode_tag(#{name.inspect}, #{raw_value.inspect})") if $DEBUG
-    case name
-      when /^COM/
+    if name =~ /^(T|COM)/
+      if name =~ /^COM/
         #FIXME improve this
-	encoding, lang, str = raw_value.unpack("ca3a*") 
-	out = raw_value.split(0.chr).last
-      when /^T/
-	encoding = raw_value.getbyte(0) # language encoding (see TEXT_ENCODINGS constant)   
-	out = raw_value[1..-1] 
-	# we need to convert the string in order to match
-	# the requested encoding
-	if encoding && TEXT_ENCODINGS[encoding] && out && encoding != @text_encoding_index
-	  begin
-	    out = Iconv.iconv(@options[:encoding], TEXT_ENCODINGS[encoding], out).first
-	  rescue Iconv::Failure
-	  end
-	end
-        # remove padding zeros for textual tags
-        out.sub!(/\0*$/, '') unless out.nil?
-        return out
+        encoding_index, lang, raw_tag = raw_value.unpack("ca3a*")
+        if encoding_index == 1
+=begin
+          comment = Mp3Info::EncodingHelper.decode_utf16(raw_tag)
+          e = comment.encoding
+          out = comment.force_encoding("BINARY").split("\x00\x00").last.force_encoding(e)
+          p out
+=end
+          comment = Mp3Info::EncodingHelper.decode_utf16(raw_tag)
+          split_val = RUBY_1_8 ? "\x00\x00" : "\x00".encode(comment.encoding)
+          out = comment.split(split_val).last rescue ""
+        else
+          comment, out = raw_tag.split("\x00", 2)
+        end
+        puts "COM tag found. encoding: #{encoding_index} lang: #{lang} str: #{out.inspect}" if $DEBUG
       else
-        return raw_value
+        encoding_index = raw_value.getbyte(0) # language encoding (see TEXT_ENCODINGS constant)   
+        out = raw_value[1..-1]
+      end
+      # we need to convert the string in order to match
+      # the requested encoding
+      if encoding_index && TEXT_ENCODINGS[encoding_index] && out
+        if RUBY_1_8
+          out = Mp3Info::EncodingHelper.convert_to(out, TEXT_ENCODINGS[encoding_index], "utf-8")
+        else
+          if encoding_index == 1
+            out = Mp3Info::EncodingHelper.decode_utf16(out)
+          else
+            out.force_encoding(TEXT_ENCODINGS[encoding_index])
+          end
+          if out
+            out.encode!("utf-8")
+          end
+        end
+      end
+
+      if out
+        # remove padding zeros for textual tags
+        if RUBY_1_8
+          r = /\0*$/
+        else
+          r = Regexp.new("\x00*$".encode(out.encoding))
+        end
+        out.sub!(r, '') 
+      end
+
+      return out
+    else
+      return raw_value
     end
   end
 
@@ -404,21 +426,26 @@ class ID3v2 < DelegateClass(Hash)
       
     data_io = @io.read(size)
     data = decode_tag(name, data_io)
+    if data && !data.empty?
+      if self.keys.include?(name) 
+        if self[name].is_a?(Array)
+          unless self[name].include?(data)
+            self[name] << data
+          end
+        else
+          self[name] = [ self[name], data ]
+        end
+      else
+        self[name] = data 
+      end
 
-    if self["TPOS"] =~ /(\d+)\s*\/\s*(\d+)/
-      self["disc_number"] = $1.to_i
-      self["disc_total"] = $2.to_i
-    end
-
-    if self.keys.include?(name)
-      unless self[name].is_a?(Array)
-        self[name] = [ self[name] ]
+      if name == "TPOS" && data =~ /(\d+)\s*\/\s*(\d+)/
+        self["disc_number"] = $1.to_i
+        self["disc_total"] = $2.to_i
       end
-      self[name] << data
-    else
-      self[name] = data 
     end
-    p data if $DEBUG
+
+    puts "self[#{name.inspect}] = #{self[name].inspect}" if $DEBUG
   end
   
   ### runs thru @file one char at a time looking for best guess of first MPEG
@@ -434,5 +461,6 @@ class ID3v2 < DelegateClass(Hash)
   def to_syncsafe(num)
     ( (num<<3) & 0x7f000000 )  + ( (num<<2) & 0x7f0000 ) + ( (num<<1) & 0x7f00 ) + ( num & 0x7f )
   end
+
 end