Dennis07

What you want is called a "Lexer" or "Parser" (depends on whether or not it is persistent), and should not be hard to get actually. Below you find one that I just wrote myself, feel free to use/modify/delete/eat/drink it or do whatever you want with it if it helps you (untested!). unit Lexer; interface uses System.SysUtils, System.Types, System.Generics.Collections; type TLexer = class abstract public type ETokenError = class(Exception); TToken = record private FText: String; FPosition: TPoint; FKind: Byte; public property Text: String read FText; property Position: TPoint read FPosition; property Kind: Byte read FKind; end; private FTokens: TList<TToken>; function GetTokenCount: Integer; function GetTokens(const AIndex: Integer): TToken; protected // Check if end of text is reached function EndsText(const AChar: PChar): Boolean; virtual; // Check if end of line is reached function BreaksLine(const AChar: PChar): Boolean; virtual; // Check if Char is valid (abort if not) function IsValidChar(const AChar: Char; const AKind: Byte): Boolean; virtual; // Get kind of new token function TokenKind(var AChar: PChar): Byte; virtual; abstract; // Check if token ends here function EndsToken(var AChar: PChar; const AKind: Byte): Boolean; virtual; abstract; // Convert token kind if necessary procedure ConvertToken(var AChar: PChar; var AKind: Byte); virtual; abstract; public property Tokens[const AIndex: Integer]: TToken read GetTokens; property TokenCount: Integer read GetTokenCount; constructor Create(const AText: String); destructor Destroy; override; end; implementation { TLexer } function TLexer.BreaksLine(const AChar: PChar): Boolean; begin Result := String.Create([AChar[0], AChar[1]]).Equals(sLineBreak); end; constructor TLexer.Create(const AText: String); procedure Parse; var Current: PChar; Previous: PChar; Token: TToken; Kind: Byte; Position: TPoint; StringBuilder: TStringBuilder; begin Position := Default(TPoint); Current := PChar(AText); StringBuilder := TStringBuilder.Create; try while not EndsText(Current) do begin Previous := Current; Kind := TokenKind(Current); while not (EndsText(Current) or EndsToken(Current, Kind)) do begin if not IsValidChar(Current[0], Kind) then begin raise ETokenError.CreateFmt('Invalid character: ', [String.Create([Current[0]]).QuotedString]); end; ConvertToken(Current, Kind); if BreaksLine(Current) then begin Inc(Position.Y); end; StringBuilder.Append(Current); Inc(Current); end; Token.FText := StringBuilder.ToString; Token.FKind := Kind; Inc(Position.X, (Current - Previous) div SizeOf(Char)); Token.FPosition := Position; StringBuilder.Clear; FTokens.Add(Token); end; finally StringBuilder.Free; end; end; begin inherited Create; FTokens := TList<TToken>.Create; Parse; end; destructor TLexer.Destroy; begin FTokens.Free; inherited; end; function TLexer.EndsText(const AChar: PChar): Boolean; begin Result := AChar[0] = #0; end; function TLexer.GetTokenCount: Integer; begin Result := FTokens.Count; end; function TLexer.GetTokens(const AIndex: Integer): TToken; begin Result := FTokens[AIndex]; end; function TLexer.IsValidChar(const AChar: Char; const AKind: Byte): Boolean; begin Result := CharInSet(AChar, [Low(AnsiChar) .. High(AnsiChar)]); end; end. Note however, that this is not optimal performance-wise.

This should happen by default if you are using the latest version of Delphi. What version are you using?