systemutf8.pas

支持版本:Delphi 5-2009, C++Builder 5-2009 ATViewer特性： Text, Binary, Hex, Unicode：所有文件

// This code is from Delphi 7 System.pas unit source,
// should be used only under Delphi versions prior to 6,
// with missing UTF-8 functions.

{$I Compilers.inc}

unit SystemUTF8;

interface

{$ifndef COMPILER_6_UP}

type
  UTF8String = type string;
  PUTF8String = ^UTF8String;

function UnicodeToUtf8(Dest: PChar; Source: PWideChar; MaxBytes: Integer): Integer; overload;
function UnicodeToUtf8(Dest: PChar; MaxDestBytes: Cardinal; Source: PWideChar; SourceChars: Cardinal): Cardinal; overload;
function Utf8ToUnicode(Dest: PWideChar; Source: PChar; MaxChars: Integer): Integer; overload;
function Utf8ToUnicode(Dest: PWideChar; MaxDestChars: Cardinal; Source: PChar; SourceBytes: Cardinal): Cardinal; overload;
function Utf8Encode(const WS: WideString): UTF8String;
function Utf8Decode(const S: UTF8String): WideString;

{$endif}

implementation

{$ifndef COMPILER_6_UP}


// UnicodeToUTF8(3):
// Scans the source data to find the null terminator, up to MaxBytes
// Dest must have MaxBytes available in Dest.

function UnicodeToUtf8(Dest: PChar; Source: PWideChar; MaxBytes: Integer): Integer;
var
  len: Cardinal;
begin
  len := 0;
  if Source <> nil then
    while Source[len] <> #0 do
      Inc(len);
  Result := UnicodeToUtf8(Dest, MaxBytes, Source, len);
end;

// UnicodeToUtf8(4):
// MaxDestBytes includes the null terminator (last char in the buffer will be set to null)
// Function result includes the null terminator.
// Nulls in the source data are not considered terminators - SourceChars must be accurate

function UnicodeToUtf8(Dest: PChar; MaxDestBytes: Cardinal; Source: PWideChar; SourceChars: Cardinal): Cardinal;
var
  i, count: Cardinal;
  c: Cardinal;
begin
  Result := 0;
  if Source = nil then Exit;
  count := 0;
  i := 0;
  if Dest <> nil then
  begin
    while (i < SourceChars) and (count < MaxDestBytes) do
    begin
      c := Cardinal(Source[i]);
      Inc(i);
      if c <= $7F then
      begin
        Dest[count] := Char(c);
        Inc(count);
      end
      else if c > $7FF then
      begin
        if count + 3 > MaxDestBytes then
          break;
        Dest[count] := Char($E0 or (c shr 12));
        Dest[count+1] := Char($80 or ((c shr 6) and $3F));
        Dest[count+2] := Char($80 or (c and $3F));
        Inc(count,3);
      end
      else //  $7F < Source[i] <= $7FF
      begin
        if count + 2 > MaxDestBytes then
          break;
        Dest[count] := Char($C0 or (c shr 6));
        Dest[count+1] := Char($80 or (c and $3F));
        Inc(count,2);
      end;
    end;
    if count >= MaxDestBytes then count := MaxDestBytes-1;
    Dest[count] := #0;
  end
  else
  begin
    while i < SourceChars do
    begin
      c := Integer(Source[i]);
      Inc(i);
      if c > $7F then
      begin
        if c > $7FF then
          Inc(count);
        Inc(count);
      end;
      Inc(count);
    end;
  end;
  Result := count+1;  // convert zero based index to byte count
end;

function Utf8ToUnicode(Dest: PWideChar; Source: PChar; MaxChars: Integer): Integer;
var
  len: Cardinal;
begin
  len := 0;
  if Source <> nil then
    while Source[len] <> #0 do
      Inc(len);
  Result := Utf8ToUnicode(Dest, MaxChars, Source, len);
end;

function Utf8ToUnicode(Dest: PWideChar; MaxDestChars: Cardinal; Source: PChar; SourceBytes: Cardinal): Cardinal;
var
  i, count: Cardinal;
  c: Byte;
  wc: Cardinal;
begin
  if Source = nil then
  begin
    Result := 0;
    Exit;
  end;
  Result := Cardinal(-1);
  count := 0;
  i := 0;
  if Dest <> nil then
  begin
    while (i < SourceBytes) and (count < MaxDestChars) do
    begin
      wc := Cardinal(Source[i]);
      Inc(i);
      if (wc and $80) <> 0 then
      begin
        if i >= SourceBytes then Exit;          // incomplete multibyte char
        wc := wc and $3F;
        if (wc and $20) <> 0 then
        begin
          c := Byte(Source[i]);
          Inc(i);
          if (c and $C0) <> $80 then Exit;      // malformed trail byte or out of range char
          if i >= SourceBytes then Exit;        // incomplete multibyte char
          wc := (wc shl 6) or (c and $3F);
        end;
        c := Byte(Source[i]);
        Inc(i);
        if (c and $C0) <> $80 then Exit;       // malformed trail byte

        Dest[count] := WideChar((wc shl 6) or (c and $3F));
      end
      else
        Dest[count] := WideChar(wc);
      Inc(count);
    end;
    if count >= MaxDestChars then count := MaxDestChars-1;
    Dest[count] := #0;
  end
  else
  begin
    while (i < SourceBytes) do
    begin
      c := Byte(Source[i]);
      Inc(i);
      if (c and $80) <> 0 then
      begin
        if i >= SourceBytes then Exit;          // incomplete multibyte char
        c := c and $3F;
        if (c and $20) <> 0 then
        begin
          c := Byte(Source[i]);
          Inc(i);
          if (c and $C0) <> $80 then Exit;      // malformed trail byte or out of range char
          if i >= SourceBytes then Exit;        // incomplete multibyte char
        end;
        c := Byte(Source[i]);
        Inc(i);
        if (c and $C0) <> $80 then Exit;       // malformed trail byte
      end;
      Inc(count);
    end;
  end;
  Result := count+1;
end;

function Utf8Encode(const WS: WideString): UTF8String;
var
  L: Integer;
  Temp: UTF8String;
begin
  Result := '';
  if WS = '' then Exit;
  SetLength(Temp, Length(WS) * 3); // SetLength includes space for null terminator

  L := UnicodeToUtf8(PChar(Temp), Length(Temp)+1, PWideChar(WS), Length(WS));
  if L > 0 then
    SetLength(Temp, L-1)
  else
    Temp := '';
  Result := Temp;
end;

function Utf8Decode(const S: UTF8String): WideString;
var
  L: Integer;
  Temp: WideString;
begin
  Result := '';
  if S = '' then Exit;
  SetLength(Temp, Length(S));

  L := Utf8ToUnicode(PWideChar(Temp), Length(Temp)+1, PChar(S), Length(S));
  if L > 0 then
    SetLength(Temp, L-1)
  else
    Temp := '';
  Result := Temp;
end;

function AnsiToUtf8(const S: string): UTF8String;
begin
  Result := Utf8Encode(S);
end;

function Utf8ToAnsi(const S: UTF8String): string;
begin
  Result := Utf8Decode(S);
end;

{$endif}

end.