kturtle/src/lexer.cpp

   1 /*
   2     Copyright (C) 2003 by Walter Schreppers
   3     Copyright (C) 2004 by Cies Breijs
   4
   5     This program is free software; you can redistribute it and/or
   6     modify it under the terms of version 2 of the GNU General Public
   7     License as published by the Free Software Foundation.
   8
   9     This program is distributed in the hope that it will be useful,
  10     but WITHOUT ANY WARRANTY; without even the implied warranty of
  11     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12     GNU General Public License for more details.
  13
  14     You should have received a copy of the GNU General Public License
  15     along with this program; if not, write to the Free Software
  16     Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
  17  */
  18
  19 #include <qdom.h>
  20 #include <qfile.h>
  21
  22 #include <kdebug.h>
  23 #include <klocale.h>
  24
  25 #include "settings.h"
  26
  27 #include "lexer.h"
  28
  29
  30 Lexer::Lexer(QTextIStream& iStream)
  31 {
  32         inputStream = &iStream;
  33         row = 1;
  34         col = 1;
  35         prevCol = 1;
  36         translate = new Translate();
  37 }
  38
  39
  40 Token Lexer::lex()
  41 {
  42         skipSpaces(); // skips the white space that it quite likely (indentation) infront of the Token
  43
  44         Token currentToken;
  45         currentToken.type = tokNotSet; // not really needed
  46         currentToken.look = "";
  47         currentToken.value = 0;
  48         currentToken.start.row = row;
  49         currentToken.start.col = col;
  50
  51         QChar currentChar = getChar();
  52
  53         if ( inputStream->atEnd() )
  54         {
  55                 kdDebug(0)<<"Lexer::lex(), got EOF."<<endl;
  56                 currentToken.type = tokEOF;
  57                 currentToken.look = "EOF";
  58                 ungetChar(currentChar); // unget the currentChar and fix the row/col values
  59                 return currentToken;
  60         }
  61
  62         if (currentChar == '#')
  63         {
  64                 while ( !inputStream->atEnd() && !(currentChar == '\x0a' || currentChar == '\n') )
  65                         currentChar = getChar();
  66         }
  67
  68         // if (currentChar.category() == QChar::Separator_Line) somehow doesnt work
  69         if (currentChar == '\x0a' || currentChar == '\n')
  70         {
  71                 currentToken.type = tokEOL;
  72                 currentToken.look = "EOL";
  73         }
  74         else if (currentChar.isLetter() || currentChar == '[' || currentChar == ']')
  75         {
  76                 ungetChar(currentChar);
  77                 // sets currentToken.look by reference, and set the currentToken.type to tokUnknown
  78                 currentToken.type = getWord(currentToken.look);
  79                 setTokenType(currentToken); // gets the actual tokenType
  80         }
  81         else if ( currentChar.isNumber() )
  82         {
  83                 ungetChar(currentChar);
  84                 // set currentToken.value/look by reference, and set the currentToken.type to tokNumber
  85                 currentToken.type = getNumber(currentToken.value, currentToken.look);
  86         }
  87         else if (currentChar == '>')
  88         {
  89                 currentChar = getChar();
  90                 if (currentChar == '=')
  91                 {
  92                         currentToken.type = tokGe;
  93                         currentToken.look = ">=";
  94                 }
  95                 else
  96                 {
  97                         ungetChar(currentChar);
  98                         currentToken.type = tokGt;
  99                         currentToken.look = ">";
 100                 }
 101         }
 102         else if (currentChar == '<')
 103         {
 104                 currentChar = getChar();
 105                 if ( currentChar == '=' )
 106                 {
 107                         currentToken.type = tokLe;
 108                         currentToken.look = "<=";
 109                 }
 110                 else
 111                 {
 112                         ungetChar(currentChar);
 113                         currentToken.type = tokLt;
 114                         currentToken.look = ">";
 115                 }
 116         }
 117         else if (currentChar == '!')
 118         {
 119                 currentChar = getChar();
 120                 if (currentChar == '=')
 121                 {
 122                         currentToken.type = tokNe;
 123                         currentToken.look = "!=";
 124                 }
 125                 else
 126                 {
 127                         ungetChar(currentChar);
 128                         currentToken.type = tokNot;
 129                         currentToken.look = "!";
 130                 }
 131         }
 132         else if (currentChar == '=')
 133         {
 134                 currentChar = getChar();
 135                 if (currentChar == '=')
 136                 {
 137                         currentToken.type = tokEq;
 138                         currentToken.look = "==";
 139                 }
 140                 else
 141                 {
 142                         ungetChar(currentChar);
 143                         currentToken.type = tokAssign;
 144                         currentToken.look = "=";
 145                 }
 146         }
 147         else if (currentChar == '(')
 148         {
 149                 currentToken.type = tokBraceOpen;
 150                 currentToken.look = "(";
 151         }
 152         else if (currentChar == ')')
 153         {
 154                 currentToken.type = tokBraceClose;
 155                 currentToken.look = ")";
 156         }
 157         else if (currentChar == '+')
 158         {
 159                 currentToken.type = tokPlus;
 160                 currentToken.look = "+";
 161         }
 162         else if (currentChar == '-')
 163         {
 164                 currentToken.type = tokMinus;
 165                 currentToken.look = "-";
 166         }
 167         else if (currentChar == '*')
 168         {
 169                 currentToken.type = tokMul;
 170                 currentToken.look = "*";
 171         }
 172         else if (currentChar == '/')
 173         {
 174                 currentToken.type = tokDev;
 175                 currentToken.look = "/";
 176         }
 177         else if (currentChar == ',')
 178         {
 179                 currentToken.type = tokComma;
 180                 currentToken.look = ",";
 181         }
 182         else if (currentChar == '"')
 183         {
 184                 getString(currentToken);
 185         }
 186         else
 187         {
 188                 currentToken.type = tokUnknown;
 189                 currentToken.look = currentChar;
 190         }
 191
 192         currentToken.end.row = row;
 193         currentToken.end.col = col;
 194         return currentToken;
 195 }
 196
 197
 198
 199 // PRIVATEs
 200
 201 QChar Lexer::getChar()
 202 {
 203         QChar c;
 204         if ( !putBackChar.isNull() )
 205         {
 206                 c = putBackChar; // use the char that is stored to be put back
 207                 // kdDebug(0)<<"Lexer::getChar(), restored: '"<<c<<"' @ ("<<row<<", "<<col<<")"<<endl;
 208                 putBackChar = QChar(); // and set putBackChar back to NULL
 209                 if (c == '\x0a' || c == '\n')
 210                 {
 211                         row++;
 212                         prevCol = col;
 213                         col = 1;
 214                 }
 215                 else
 216                 {
 217                         col++;
 218                 }
 219         }
 220         else
 221         {
 222                 *inputStream >> c; // take a QChar of the inputStream
 223                 // kdDebug(0)<<"Lexer::getChar(): '"<<c<<"' @ ("<<row<<", "<<col<<")"<<endl;
 224                 if (c == '\x0a' || c == '\n')
 225                 {
 226                         row++;
 227                         prevCol = col;
 228                         col = 1;
 229                 }
 230                 else
 231                 {
 232                         col++;
 233                 }
 234         }
 235         return c;
 236 }
 237
 238 void Lexer::ungetChar(QChar c)
 239 {
 240         if (c == '\x0a' || c == '\n')
 241         {
 242                 row--;
 243                 col = prevCol;
 244         }
 245         else
 246         {
 247                 col--;
 248         }
 249         putBackChar = c;
 250         // kdDebug(0)<<"Lexer::ungetChar(), saved char: '"<<c<<"' and steped back to ("<<row<<", "<<col<<")"<<endl;
 251 }
 252
 253 int Lexer::getWord(QString& word)
 254 {
 255         // kdDebug(0)<<"Lexer::getWord()"<<endl;
 256         QChar currentChar = getChar();
 257         if ( currentChar.isLetter() || currentChar == '[' || currentChar == ']' ) {
 258                 while ( ( currentChar.isLetterOrNumber() || currentChar == '_' || currentChar == '[' || currentChar == ']' ) && !inputStream->atEnd() )
 259                 {
 260                         word += currentChar;
 261                         currentChar = getChar();
 262                 }
 263                 kdDebug(0)<<"Lexer::getWord(), got NAME: '"<<word<<"'"<<endl;
 264                 ungetChar(currentChar); //read one too much
 265                 return tokUnknown; // returns tokUnknown, actual token is to be determained later in Lexer::setTokenType
 266         }
 267         else return tokError;
 268 }
 269
 270 void Lexer::setTokenType(Token& currentToken)
 271 {
 272         if (currentToken.type == tokUnknown)
 273         {
 274                 // make lowercase copy of the word as it was found in the inputStream
 275                 QString k = currentToken.look.lower();
 276                 // if the key is an alias translate that alias to a key
 277                 if ( !translate->alias2key(k).isEmpty() ) k = translate->alias2key(k);
 278
 279                 if      (k == translate->name2key("begin")          ) currentToken.type = tokBegin;
 280                 else if (k == translate->name2key("end")            ) currentToken.type = tokEnd;
 281                 else if (k == translate->name2key("while")          ) currentToken.type = tokWhile;
 282                 else if (k == translate->name2key("if")             ) currentToken.type = tokIf;
 283                 else if (k == translate->name2key("else")           ) currentToken.type = tokElse;
 284                 else if (k == translate->name2key("for")            ) currentToken.type = tokFor;
 285                 else if (k == translate->name2key("to")             ) currentToken.type = tokTo;
 286                 else if (k == translate->name2key("step")           ) currentToken.type = tokStep;
 287                 else if (k == translate->name2key("and")            ) currentToken.type = tokAnd;
 288                 else if (k == translate->name2key("or")             ) currentToken.type = tokOr;
 289                 else if (k == translate->name2key("not")            ) currentToken.type = tokNot;
 290                 else if (k == translate->name2key("return")         ) currentToken.type = tokReturn;
 291                 else if (k == translate->name2key("break")          ) currentToken.type = tokBreak;
 292                 else if (k == translate->name2key("run")            ) currentToken.type = tokRun;
 293                 else if (k == translate->name2key("foreach")        ) currentToken.type = tokForEach;
 294                 else if (k == translate->name2key("in")             ) currentToken.type = tokIn;
 295
 296                 else if (k == translate->name2key("learn")          ) currentToken.type = tokLearn;
 297
 298                 else if (k == translate->name2key("clear")          ) currentToken.type = tokClear;
 299                 else if (k == translate->name2key("go")             ) currentToken.type = tokGo;
 300                 else if (k == translate->name2key("gox")            ) currentToken.type = tokGoX;
 301                 else if (k == translate->name2key("goy")            ) currentToken.type = tokGoY;
 302                 else if (k == translate->name2key("forward")        ) currentToken.type = tokForward;
 303                 else if (k == translate->name2key("backward")       ) currentToken.type = tokBackward;
 304                 else if (k == translate->name2key("direction")      ) currentToken.type = tokDirection;
 305                 else if (k == translate->name2key("turnleft")       ) currentToken.type = tokTurnLeft;
 306                 else if (k == translate->name2key("turnright")      ) currentToken.type = tokTurnRight;
 307                 else if (k == translate->name2key("center")         ) currentToken.type = tokCenter;
 308                 else if (k == translate->name2key("setpenwidth")    ) currentToken.type = tokSetPenWidth;
 309                 else if (k == translate->name2key("penup")          ) currentToken.type = tokPenUp;
 310                 else if (k == translate->name2key("pendown")        ) currentToken.type = tokPenDown;
 311                 else if (k == translate->name2key("setfgcolor")     ) currentToken.type = tokSetFgColor;
 312                 else if (k == translate->name2key("setbgcolor")     ) currentToken.type = tokSetBgColor;
 313                 else if (k == translate->name2key("resizecanvas")   ) currentToken.type = tokResizeCanvas;
 314                 else if (k == translate->name2key("spriteshow")     ) currentToken.type = tokSpriteShow;
 315                 else if (k == translate->name2key("spritehide")     ) currentToken.type = tokSpriteHide;
 316                 else if (k == translate->name2key("spritepress")    ) currentToken.type = tokSpritePress;
 317                 else if (k == translate->name2key("spritechange")   ) currentToken.type = tokSpriteChange;
 318
 319                 else if (k == translate->name2key("do")             ) currentToken.type = tokDo; // dummy commands
 320
 321                 else if (k == translate->name2key("message")        ) currentToken.type = tokMessage;
 322                 else if (k == translate->name2key("inputwindow")    ) currentToken.type = tokInputWindow;
 323                 else if (k == translate->name2key("print")          ) currentToken.type = tokPrint;
 324                 else if (k == translate->name2key("fonttype")       ) currentToken.type = tokFontType;
 325                 else if (k == translate->name2key("fontsize")       ) currentToken.type = tokFontSize;
 326                 else if (k == translate->name2key("repeat")         ) currentToken.type = tokRepeat;
 327                 else if (k == translate->name2key("random")         ) currentToken.type = tokRandom;
 328                 else if (k == translate->name2key("wait")           ) currentToken.type = tokWait;
 329                 else if (k == translate->name2key("wrapon")         ) currentToken.type = tokWrapOn;
 330                 else if (k == translate->name2key("wrapoff")        ) currentToken.type = tokWrapOff;
 331                 else if (k == translate->name2key("reset")          ) currentToken.type = tokReset;
 332                 else
 333                 {
 334                         kdDebug(0)<<"Lexer::setTokenType, found UNKNOWN word @ ("<<currentToken.start.row<<", "<<currentToken.start.col<<"), can be anything"<<endl;
 335                         // t.type = tokUnknown; is allready
 336                 }
 337
 338                 kdDebug(0)<<"Lexer::setTokenType, found tok-number: '"<<currentToken.type<<"' with the key: '"<<k<<"' @ ("<<currentToken.start.row<<", "<<currentToken.start.col<<")"<<endl;
 339         }
 340 }
 341
 342
 343 void Lexer::skipSpaces()
 344 {
 345         // kdDebug(0)<<"Lexer::skipSpaces(), skipping SPACES."<<endl;
 346         QChar currentChar = getChar();
 347         // when the Separator_* groups can be identified in the QChar thing would be easier
 348         while ( !inputStream->atEnd() && ( currentChar.isSpace() && !(currentChar == '\x0a' || currentChar == '\n') ) )
 349         {
 350                 currentChar = getChar();
 351         }
 352         ungetChar(currentChar); // unget the tokEOL we likely just found
 353 }
 354
 355
 356 int Lexer::getNumber(Value& num, QString& look)
 357 {
 358         // by reference the value (Value) and look part are set
 359         // kdDebug(0)<<"Lexer::getNumber()"<<endl;
 360         bool hasPoint = false;
 361         QChar currentChar = getChar();
 362         if ( currentChar.isNumber() )
 363         {
 364                 while ( ( currentChar.isNumber() || (currentChar == '.' && !hasPoint) ) && !inputStream->atEnd() )
 365                 {
 366                         if (currentChar == '.')
 367                         {
 368                                 hasPoint = true;
 369                         }
 370                         look += currentChar;
 371                         currentChar = getChar();
 372                 }
 373                 ungetChar(currentChar); //read one too much
 374                 num.setNumber( look.toDouble() );
 375                 kdDebug(0)<<"Lexer::getNumber(), got NUMBER: '"<<num.Number()<<"'"<<endl;
 376                 return tokNumber;
 377         }
 378         else return tokError;
 379 }
 380
 381 void Lexer::getString(Token& currentToken)
 382 {
 383         QString str = "\""; // start with a " cauz it just got lost
 384         QChar currentChar = QChar(); // start empty
 385         while ( currentChar != '"' && !(currentChar == '\x0a' || currentChar == '\n') && !inputStream->atEnd() )
 386         {
 387                 currentChar = getChar();
 388                 if (currentChar == '\\') // escape sequence
 389                 {
 390                         currentChar = getChar();
 391                         switch (currentChar)
 392                         {
 393                                 case 'n': str += '\n'; break;
 394                                 case 't': str += '\t'; break;
 395                                 case 'f': str += '\f'; break;
 396                                 case '"': str += '"';  break;
 397                         }
 398                 }
 399                 else if (currentChar == '\x0a' || currentChar == '\n') // if the user forgot to delimit the string
 400                 {
 401                         ungetChar(currentChar);
 402                         break;
 403                 }
 404                 else str += currentChar;
 405         }
 406         currentToken.type = tokString;
 407         currentToken.look = str;
 408
 409         kdDebug(0)<<"Lexer::getStringConstant, got STRINGCONSTANT: "<<currentToken.look<<"'"<<endl;
 410 }
 411