ruby-mp3info 0.6.4 → 0.6.5

data/History.txt

@@ -1,3 +1,10 @@
+=== 0.6.5 / 2008-04-19
+
+* added Mp3Info#audio_content method, to return "audio-only" boundaries from mp3, i.e. data without tags
+  (closes feature request #17230)
+* bugfix on reading id3 2.4 (size not syncsafed)
+* more robust tag decoding with bad tags
+
 === 0.6.4 / 2008-04-16
 
 *  added @tag2["disc_number"] and @tag2["disc_total"] mirroring TPOS attribute (thanks to Harry Ohlsen)

data/lib/mp3info.rb

@@ -1,4 +1,4 @@
-# $Id: mp3info.rb 77 2008-04-16 20:04:35Z moumar $
+# $Id: mp3info.rb 82 2008-04-19 10:59:02Z moumar $
 # License:: Ruby
 # Author:: Guillaume Pierronnet (mailto:moumar_AT__rubyforge_DOT_org)
 # Website:: http://ruby-mp3info.rubyforge.org/
@@ -17,7 +17,7 @@ end
 
 class Mp3Info
 
-  VERSION = "0.6.4"
+  VERSION = "0.6.5"
 
   LAYER = [ nil, 3, 2, 1]
   BITRATE = [
@@ -65,7 +65,7 @@ class Mp3Info
     "Christian Rock", "Merengue", "Salsa", "Thrash Metal", "Anime", "JPop",
     "SynthPop" ]
 
-  TAGSIZE = 128
+  TAG1_SIZE = 128
   #MAX_FRAME_COUNT = 6  #number of frame to read for encoder detection
   V1_V2_TAG_MAPPING = { 
     "title"    => "TIT2",
@@ -132,7 +132,7 @@ class Mp3Info
   # Test the presence of an id3v1 tag in file +filename+
   def self.hastag1?(filename)
     File.open(filename) { |f|
-      f.seek(-TAGSIZE, File::SEEK_END)
+      f.seek(-TAG1_SIZE, File::SEEK_END)
       f.read(3) == "TAG"
     }
   end
@@ -148,7 +148,7 @@ class Mp3Info
   # Remove id3v1 tag from +filename+
   def self.removetag1(filename)
     if self.hastag1?(filename)
-      newsize = File.size(filename) - TAGSIZE
+      newsize = File.size(filename) - TAG1_SIZE
       File.open(filename, "rb+") { |f| f.truncate(newsize) }
     end
   end
@@ -172,7 +172,6 @@ class Mp3Info
   # reload (or load for the first time) the file from disk
   def reload
     raise(Mp3InfoError, "empty file") unless File.size?(@filename)
-    @hastag1 = false
     
     @tag1 = {}
     @tag1.extend(HashKeys)
@@ -262,7 +261,7 @@ class Mp3Info
 	@vbr = true
       else
 	# for cbr, calculate duration with the given bitrate
-	@streamsize = @file.stat.size - (@hastag1 ? TAGSIZE : 0) - (@tag2.valid? ? @tag2.io_position : 0)
+	@streamsize = @file.stat.size - (hastag1? ? TAG1_SIZE : 0) - (@tag2.valid? ? @tag2.io_position : 0)
 	@length = ((@streamsize << 3)/1000.0)/@bitrate
 	if @tag2["TLEN"]
 	  # but if another duration is given and it isn't close (within 5%)
@@ -301,10 +300,7 @@ class Mp3Info
 
   # Remove id3v1 from mp3
   def removetag1
-    if hastag1?
-      Mp3Info.removetag1(@filename)
-      @tag1.clear
-    end
+    @tag1.clear
     self
   end
   
@@ -316,12 +312,12 @@ class Mp3Info
 
   # Does the file has an id3v1 or v2 tag?
   def hastag?
-    @hastag1 or @tag2.valid?
+    hastag1? or hastag2?
   end
 
   # Does the file has an id3v1 tag?
   def hastag1?
-    @hastag1
+    !@tag1.empty?
   end
 
   # Does the file has an id3v2 tag?
@@ -333,6 +329,23 @@ class Mp3Info
   def rename(new_filename)
     @filename = new_filename
   end
+
+  # this method returns the "audio-only" data boundaries of the file, 
+  # i.e. content stripped form tags. Useful to compare 2 files with the same 
+  # audio content but with differents tags. Returned value is an array 
+  # [position_in_the_file, length_of_the_data]
+  def audio_content
+    pos = 0
+    length = File.size(@filename)
+    if hastag1?
+      length -= TAG1_SIZE
+    end
+    if hastag2?
+      pos = @tag2.io_position
+      length -= @tag2.io_position
+    end
+    [pos, length]
+  end
   
   # Flush pending modifications to tags and close the file
   def close
@@ -358,24 +371,29 @@ class Mp3Info
       #@tag1_orig.update(@tag1)
       @tag1_orig = @tag1.dup
       File.open(@filename, 'rb+') do |file|
-	file.seek(-TAGSIZE, File::SEEK_END)
-	t = file.read(3)
-	if t != 'TAG'
-	  #append new tag
-	  file.seek(0, File::SEEK_END)
-	  file.write('TAG')
-	end
-	str = [
-	  @tag1_orig["title"]||"",
-	  @tag1_orig["artist"]||"",
-	  @tag1_orig["album"]||"",
-	  ((@tag1_orig["year"] != 0) ? ("%04d" % @tag1_orig["year"].to_i) : "\0\0\0\0"),
-	  @tag1_orig["comments"]||"",
-	  0,
-	  @tag1_orig["tracknum"]||0,
-	  @tag1_orig["genre"]||255
-	  ].pack("Z30Z30Z30Z4Z28CCC")
-	file.write(str)
+        if @tag1_orig.empty?
+          newsize = file.stat.size - TAG1_SIZE
+          file.truncate(newsize)
+        else
+          file.seek(-TAG1_SIZE, File::SEEK_END)
+          t = file.read(3)
+          if t != 'TAG'
+            #append new tag
+            file.seek(0, File::SEEK_END)
+            file.write('TAG')
+          end
+          str = [
+            @tag1_orig["title"]||"",
+            @tag1_orig["artist"]||"",
+            @tag1_orig["album"]||"",
+            ((@tag1_orig["year"] != 0) ? ("%04d" % @tag1_orig["year"].to_i) : "\0\0\0\0"),
+            @tag1_orig["comments"]||"",
+            0,
+            @tag1_orig["tracknum"]||0,
+            @tag1_orig["genre"]||255
+            ].pack("Z30Z30Z30Z4Z28CCC")
+          file.write(str)
+        end
       end
     end
 
@@ -412,11 +430,18 @@ class Mp3Info
     end
   end
 
+  # close and reopen the file, i.e. commit changes to disk and 
+  # reload it
+  def flush
+    close
+    reload
+  end
+
   # inspect inside Mp3Info
   def to_s
     s = "MPEG #{@mpeg_version} Layer #{@layer} #{@vbr ? "VBR" : "CBR"} #{@bitrate} Kbps #{@channel_mode} #{@samplerate} Hz length #{@length} sec. error protection #{@error_protection} "
-    s << "tag1: "+@tag1.inspect+"\n" if @hastag1
-    s << "tag2: "+@tag2.inspect+"\n" if @tag2.valid?
+    s << "tag1: "+@tag1.inspect+"\n" if hastag1?
+    s << "tag2: "+@tag2.inspect+"\n" if hastag2?
     s
   end
 
@@ -425,24 +450,28 @@ private
   
   ### parses the id3 tags of the currently open @file
   def parse_tags
-    return if @file.stat.size < TAGSIZE  # file is too small
+    return if @file.stat.size < TAG1_SIZE  # file is too small
+    @tag1_parsed = false
     @file.seek(0)
     f3 = @file.read(3)
-    gettag1 if f3 == "TAG"  # v1 tag at beginning
-    begin
-      @tag2.from_io(@file) if f3 == "ID3"  # v2 tag at beginning
-    rescue RuntimeError => e
-      raise(Mp3InfoError, e.message)
+    # v1 tag at beginning
+    if f3 == "TAG"  
+      gettag1 
+      @tag1_parsed = true
     end
+
+    @tag2.from_io(@file) if f3 == "ID3"  # v2 tag at beginning
       
-    unless @hastag1         # v1 tag at end
-        # this preserves the file pos if tag2 found, since gettag2 leaves
-        #  the file at the best guess as to the first MPEG frame
-        pos = (@tag2.valid? ? @file.pos : 0)
-        # seek to where id3v1 tag should be
-        @file.seek(-TAGSIZE, IO::SEEK_END) 
-        gettag1 if @file.read(3) == "TAG"
-        @file.seek(pos)
+    unless @tag1_parsed         # v1 tag at end
+      # this preserves the file pos if tag2 found, since gettag2 leaves
+      # the file at the best guess as to the first MPEG frame
+      pos = (@tag2.valid? ? @file.pos : 0)
+      # seek to where id3v1 tag should be
+      @file.seek(-TAG1_SIZE, IO::SEEK_END) 
+      if @file.read(3) == "TAG"
+        gettag1
+      end
+      @file.seek(pos)
     end
   end
 
@@ -465,7 +494,14 @@ private
   ### gets id3v1 tag information from @file
   ### assumes @file is pointing to char after "TAG" id
   def gettag1
-    @hastag1 = true
+    @tag1_parsed = true
+=begin
+    # FIXME remove that
+    pos = @file.pos
+    p @file.read
+    @file.seek(pos)
+    # FIXME remove that
+=end
     @tag1["title"] = read_id3_string(30)
     @tag1["artist"] = read_id3_string(30)
     @tag1["album"] = read_id3_string(30)
@@ -500,7 +536,7 @@ private
     dummyproof.times do |i|
       if @file.getc == 0xff
         data = @file.read(3)
-	raise Mp3InfoError if @file.eof?
+        raise(Mp3InfoError, "end of file reached") if @file.eof?
         head = 0xff000000 + (data[0] << 16) + (data[1] << 8) + data[2]
         if check_head(head)
             return head

data/lib/mp3info/extension_modules.rb

@@ -22,6 +22,4 @@ class Mp3Info
       (getc << 21) + (getc << 14) + (getc << 7) + getc
     end
   end
-
 end
-

data/lib/mp3info/id3v2.rb

@@ -3,6 +3,8 @@ require "iconv"
 
 require "mp3info/extension_modules"
 
+class ID3v2Error < StandardError ; end
+
 # This class is not intended to be used directly
 class ID3v2 < DelegateClass(Hash) 
   
@@ -146,7 +148,7 @@ class ID3v2 < DelegateClass(Hash)
     @io.extend(Mp3Info::Mp3FileMethods)
     version_maj, version_min, flags = @io.read(3).unpack("CCB4")
     @unsync, ext_header, experimental, footer = (0..3).collect { |i| flags[i].chr == '1' }
-    raise("can't find version_maj ('#{version_maj}')") unless [2, 3, 4].include?(version_maj)
+    raise(ID3v2Error, "can't find version_maj ('#{version_maj}')") unless [2, 3, 4].include?(version_maj)
     @version_maj, @version_min = version_maj, version_min
     @valid = true
     @tag_length = @io.get_syncsafe
@@ -196,7 +198,7 @@ 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 << to_syncsafe(tag.size)
+    tag_str << [to_syncsafe(tag.size)].pack("N")
     tag_str << tag
     p tag_str if $DEBUG
     tag_str
@@ -228,19 +230,19 @@ class ID3v2 < DelegateClass(Hash)
   end
 
   ### Read a tag from file and perform UNICODE translation if needed
-  def decode_tag(name, value)
-    puts("decode_tag(#{name.inspect}, #{value.inspect})") if $DEBUG
+  def decode_tag(name, raw_value)
+    puts("decode_tag(#{name.inspect}, #{raw_value.inspect})") if $DEBUG
     case name
       when "COMM"
         #FIXME improve this
-	encoding, lang, str = value.unpack("ca3a*")
-	out = value.split(0.chr).last
+	encoding, lang, str = raw_value.unpack("ca3a*")
+	out = raw_value.split(0.chr).last
       when /^T/
-	encoding = value[0]     # language encoding (see TEXT_ENCODINGS constant)
-	out = value[1..-1] 
+	encoding = raw_value[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_encoding_index
+	if out && encoding != @text_encoding_index
 	  begin
 	    Iconv.iconv(@options[:encoding], TEXT_ENCODINGS[encoding], out)[0]
 	  rescue Iconv::Failure
@@ -250,7 +252,7 @@ class ID3v2 < DelegateClass(Hash)
 	  out
 	end
       else
-        value
+        raw_value
     end
   end
 
@@ -264,11 +266,13 @@ class ID3v2 < DelegateClass(Hash)
 	seek_to_v2_end
         break
       else
-        #size = @io.get_syncsafe #this seems to be a bug
-        size = @io.get32bits
-	@io.read(2)
+	if @version_maj == 4
+	  size = @io.get_syncsafe
+	else
+	  size = @io.get32bits
+	end
+        @io.seek(2, IO::SEEK_CUR)     # skip flags
         puts "name '#{name}' size #{size}" if $DEBUG
-        #@io.seek(2, IO::SEEK_CUR)     # skip flags
         add_value_to_tag2(name, size)
       end
       break if @io.pos >= @tag_length # 2. reach length from header
@@ -297,7 +301,6 @@ class ID3v2 < DelegateClass(Hash)
   ### create an array if the key already exists in the tag
   def add_value_to_tag2(name, size)
     puts "add_value_to_tag2" if $DEBUG
-    raise("tag size too big for tag #{name.inspect} unsync #{@unsync} ") if size > 50_000_000
     data_io = @io.read(size)
     data = decode_tag(name, data_io)
 
@@ -328,8 +331,7 @@ class ID3v2 < DelegateClass(Hash)
   
   ### convert an 32 integer to a syncsafe string
   def to_syncsafe(num)
-    n = ( (num<<3) & 0x7f000000 )  + ( (num<<2) & 0x7f0000 ) + ( (num<<1) & 0x7f00 ) + ( num & 0x7f )
-    [n].pack("N")
+    ( (num<<3) & 0x7f000000 )  + ( (num<<2) & 0x7f0000 ) + ( (num<<1) & 0x7f00 ) + ( num & 0x7f )
   end
 
 #  def method_missing(meth, *args)

data/test/test_ruby-mp3info.rb

@@ -11,7 +11,8 @@ require "tempfile"
 class Mp3InfoTest < Test::Unit::TestCase
 
   TEMP_FILE = File.join(File.dirname($0), "test_mp3info.mp3")
-  BASIC_TAG2 = {
+
+  DUMMY_TAG2 = {
     "COMM" => "comments",
     #"TCON" => "genre_s" 
     "TIT2" => "title",
@@ -20,7 +21,17 @@ class Mp3InfoTest < Test::Unit::TestCase
     "TYER" => "year",
     "TRCK" => "tracknum"
   }
-  
+
+  DUMMY_TAG1 = {
+    "title"    => "toto",
+    "artist"   => "artist 123", 
+    "album"    => "ALBUMM",
+    "year"     => 1934,
+    "tracknum" => 14,
+    "comments" => "comment me",
+    "genre" => 233
+  }
+
   def setup
     # Command to create a dummy MP3
     # dd if=/dev/zero bs=1024 count=15 | lame --preset cbr 128 -r -s 44.1 --bitwidth 16 - - | ruby -rbase64 -e 'print Base64.encode64($stdin.read)'
@@ -229,7 +240,7 @@ EOF
   end
 
   def test_id3v2_version
-    written_tag = write_temp_file(BASIC_TAG2)
+    written_tag = write_temp_file(DUMMY_TAG2)
     assert_equal( "2.#{ID3v2::WRITE_VERSION}.0", written_tag.version )
   end
 
@@ -244,9 +255,9 @@ EOF
   end
 
   def test_id3v2_basic
-    w = write_temp_file(BASIC_TAG2)
-    assert_equal(BASIC_TAG2, w)
-    id3v2_prog_test(BASIC_TAG2, w)
+    w = write_temp_file(DUMMY_TAG2)
+    assert_equal(DUMMY_TAG2, w)
+    id3v2_prog_test(DUMMY_TAG2, w)
   end
 
   #test the tag with the "id3v2" program
@@ -295,7 +306,7 @@ EOF
   end
 
   def test_leading_char_gets_chopped
-    tag2 = BASIC_TAG2.dup
+    tag2 = DUMMY_TAG2.dup
     tag2["WOAR"] = "http://foo.bar"
     w = write_temp_file(tag2)
     assert_equal("http://foo.bar", w["WOAR"])
@@ -316,27 +327,11 @@ EOF
       mp3.reload
       assert !mp3.tag1.empty?, "tag is empty"
       mp3.removetag1
-      mp3.close
-      mp3.reload
+      mp3.flush
       assert mp3.tag1.empty?, "tag is not empty"
     end
   end
 
-    #test the tag with php getid3
-#    prog = %{
-#    <?php
-#      require("/var/www/root/netjuke/lib/getid3/getid3.php");
-#      $mp3info = GetAllFileInfo('#{TEMP_FILE}');
-#      echo $mp3info;
-#    ?>
-#    }
-#
-#    open("|php", "r+") do |io|
-#      io.puts(prog)
-#      io.close_write
-#      p io.read
-#    end
-
   def test_good_parsing_of_a_pathname
     fn = "Freak On `(Stone´s Club Mix).mp3"
     File.rename(TEMP_FILE, fn)
@@ -394,6 +389,42 @@ EOF
     end
   end
 
+  def test_audio_content
+    require "digest/md5"
+
+    expected_digest = nil
+    Mp3Info.open(TEMP_FILE) do |mp3|
+      mp3.tag1.update(DUMMY_TAG1)
+      mp3.tag2.update(DUMMY_TAG2)
+      mp3.flush
+      assert mp3.hastag1?
+      assert mp3.hastag2?
+      assert mp3.tag2.io_position != 0
+      expected_digest = compute_audio_content_mp3_digest(mp3)
+    end
+
+    Mp3Info.open(TEMP_FILE) do |mp3|
+      mp3.removetag1
+      mp3.removetag2
+      mp3.flush
+      assert !mp3.hastag1?
+      assert !mp3.hastag2?
+      got_digest = compute_audio_content_mp3_digest(mp3)
+      assert_equal expected_digest, got_digest
+    end
+  end
+
+  def compute_audio_content_mp3_digest(mp3)
+    pos, size = mp3.audio_content
+    data = File.open(mp3.filename) do |f|
+      f.seek(pos, IO::SEEK_SET)
+      f.read(size)
+    end
+    Digest::MD5.new.update(data).hexdigest
+  end
+  def compute_md5(content)
+  end
+
   def write_temp_file(tag)
     Mp3Info.open(TEMP_FILE) do |mp3|
       mp3.tag2.update(tag)

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: ruby-mp3info
 version: !ruby/object:Gem::Version 
-  version: 0.6.4
+  version: 0.6.5
 platform: ruby
 authors: 
 - Guillaume Pierronnet
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2008-04-16 00:00:00 +02:00
+date: 2008-04-19 00:00:00 +02:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -64,7 +64,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
 requirements: []
 
 rubyforge_project: ruby-mp3info
-rubygems_version: 1.0.1
+rubygems_version: 1.1.1
 signing_key: 
 specification_version: 2
 summary: ruby-mp3info is a pure-ruby library to retrieve low level informations on mp3 files and manipulate id3v1 and id3v2 tags