在获得MP3 ID3 v2实施时,我遇到了一些问题。除了这一个问题之外,我大部分时间都在工作,这可能与此根本无关。无论如何,我使用下面的代码来处理检索涉及文本的标题标记的数据。
我遇到的是(我猜?)我在一些不同的字符串中遇到Unicode字符。我试图在下面转换它,它的工作原理。但是我在字符串之前获得了3F的奖金,之后获得了$ 3F $ 3F。我可以对下面的代码做些什么来解析那些或者我必须自己做吗?文件由ITunes编码,如果这有助于任何。
function Id3v2_string(currp: pointer; datasize: integer): string;
{ handles string processing for ID3v2 data }
const
IS_TEXT_UNICODE_UNICODE_MASK = $0F;
var
outstr: string;
uscan: integer;
begin
outstr := '';
SetLength(outstr, datasize);
uscan := IS_TEXT_UNICODE_UNICODE_MASK;
if IsTextUnicode(currp, datasize, @uscan) then
outstr := WideCharToString(currp)
else
move(currp^, outstr[1], datasize);
Result := outstr;
end;
注意,我真的对媒体库不感兴趣,因为我要做的就是编辑ID3标签而不是播放文件 - 除了像这样的一些小问题之外,实现已经完成。
答案 0 :(得分:6)
根据正在使用的ID3 v2的版本,文本字符串可能会或可能不会以字节开头,以告诉您字符串的实际编码。请勿使用IsTextUnicode()
来猜测编码是什么(特别是因为它可以报告false results)。
在ID3 v2到v2.3之前,没有编码字节,文本是ISO-8859-1
或UCS-2
,UCS-2
字符串始终以BOM开头,因此您知道字节排序。例如:
// prior to Delphi 2009 - String is Ansi
function Id3v2_string(currp: Pointer; datasize: Integer): String;
var
W: WideString;
I: Integer;
Ch: WideChar;
begin
Result := '';
if (datasize >= SizeOf(Word)) and ((PWord(currp)^ = $FEFF) or (PWord(currp)^= $FFFE)) then begin
// UCS-2 with BOM
W := WideCharLenToString(PWideChar(Integer(currp) + SizeOf(Word)), (datasize - SizeOf(Word)) div SizeOf(WideChar));
if PWord(currp)^ = $FFFE then begin
// BE, convert to LE
for I := 1 to Length(W) do begin
Ch := W[I];
W[I] := WideChar(((Word(Ch) and $FF) shl 8) or (Word(Ch) shr 8));
end;
end;
end else begin
// ISO-8859-1
I := MultiByteToWideChar(28591, 0, PAnsiChar(currp), datasize, nil, 0);
if I > 0 then begin
SetLength(W, I);
MultiByteToWideChar(28591, 0, PAnsiChar(currp), datasize, PWideChar(W), I);
end;
end;
Result := TrimRight(W);
end;
// Delphi 2009+ - String is Unicode
function Id3v2_string(currp: Pointer; datasize: Integer): String;
var
Enc: TEncoding;
function Convert(P: Pointer; Size: Integer): String;
var
Buf: TBytes;
begin
SetLength(Buf, Size);
if Size > 0 then Move(P^, Buf[0], Size);
Result := Enc.GetString(Buf);
end;
begin
Result := '';
if (datasize >= SizeOf(Word)) and ((PWord(currp)^ = $FEFF) or (PWord(currp)^ = $FFFE)) then begin
// UCS-2 with BOM
if PWord(currp)^ = $FFFE then begin
// BE
Enc := TEncoding.BigEndianUnicode;
end else begin
// LE
Enc := TEncoding.Unicode;
end;
Result := Convert(PWord(currp)+1, datasize - SizeOf(Word));
end else begin
// ISO-8859-1
Enc := TEncoding.GetEncoding(28591);
try
Result := Convert(currp, datasize);
finally
Enc.Free;
end;
end;
end;
ID3 v2.4将UCS-2
切换为UTF-16
,并在没有BOM的情况下添加对UTF-8
和UTF-16BE
的支持,例如:
// prior to Delphi 2009 - String is Ansi
function Id3v2_string(currp: Pointer; datasize: Integer; Encoding: Byte): String;
var
W: WideString;
I: Integer;
Ch: WideChar;
begin
Result := '';
case Encoding of
$00: begin
// ISO-8859-1
I := MultiByteToWideChar(28591, 0, PAnsiChar(currp), datasize, nil, 0);
if I > 0 then begin
SetLength(W, I);
MultiByteToWideChar(28591, 0, PAnsiChar(currp), datasize, PWideChar(W), I);
end;
end;
$01: begin
// UTF-16 with BOM
SetString(W, PWideChar(Integer(currp) + SizeOf(Word)), (datasize - SizeOf(Word)) div SizeOf(WideChar));
if PWord(currp)^ = $FFFE then begin
// BE, convert to LE
for I := 1 to Length(W) do begin
Ch := W[I];
W[I] := WideChar(((Word(Ch) and $FF) shl 8) or (Word(Ch) shr 8));
end;
end;
end;
$02: begin
// UTF-16BE without BOM, convert to LE
SetString(W, PWideChar(currp), datasize div SizeOf(WideChar));
for I := 1 to Length(W) do begin
Ch := W[I];
W[I] := WideChar(((Word(Ch) and $FF) shl 8) or (Word(Ch) shr 8));
end;
end;
$03: begin
// UTF-8
I := MultiByteToWideChar(65001, 0, PAnsiChar(currp), datasize, nil, 0);
if I > 0 then begin
SetLength(W, I);
MultiByteToWideChar(65001, 0, PAnsiChar(currp), datasize, PWideChar(W), I);
end;
end;
end;
Result := TrimRight(W);
end;
// Delphi 2009+ - String is Unicode
function Id3v2_string(currp: Pointer; datasize: Integer; Encoding: Byte): String;
var
Enc: TEncoding;
function Convert(P: Pointer; Size: Integer): String;
var
Buf: TBytes;
begin
SetLength(Buf, Size);
if Size > 0 then Move(P^, Buf[0], Size);
Result := Enc.GetString(Buf);
end;
begin
Result := '';
case Encoding of
$00: begin
// ISO-8859-1
Enc := TEncoding.GetEncoding(28591);
try
Result := Convert(currp, datasize);
finally
Enc.Free;
end;
end;
$01: begin
// UTF-16 with BOM
if PWord(currp)^ = $FFFE then begin
// BE
Enc := TEncoding.BigEndianUnicode;
end else begin
// LE
Enc := TEncoding.Unicode;
end;
Result := Convert(PWord(currp)+1, datasize - SizeOf(Word));
end;
$02: begin
// UTF-16BE without BOM
Enc := TEncoding.BigEndianUnicode;
Result := Convert(currp, datasize);
end;
$03: begin
// UTF-8
Enc := TEncoding.UTF8;
Result := Convert(currp, datasize);
end;
end;
Result := TrimRight(Result);
end;