Token Scanner::nextToken(Attribute& attr, int& lineno) {
int c;
lineno_ = 1;
//
// Eat whitespace and comments
//
top:
do {
c = in_.get();
} while (std::isspace(c));
if (c == '#') {
do {
c = in_.get();
if (c == EOF) return Token::EOT;
} while (c != '\n');
lineno_++;
goto top;
}
if (c == EOF) return Token::EOT;
if (c == '+') return Token::PLUS;
if (c == '-') return Token::MINUS;
if (c == '*') return Token::MULT;
if (c == '/') return Token::DIV;
if (c == '(') return Token::LPAREN;
if (c == ')') return Token::RPAREN;
if (c == '=') return Token::EQ;
if (c == '\n') lineno_++;
if (c == ':') { // assign :=
c = in_.get();
if (c != '=')
throw std::runtime_error("Unknown lexeme");
return Token::ASSIGN;
}
if (c == '>') { //greater or equal
c = in_.get();
if (c != '='){
in_.unget();
return Token::GT;
}
return Token::GE;
}
if (c == '<') { //less then, Not equal, less then equal
c = in_.get();
if (c == '>'){
return Token::NE;
} else if (c == '='){
return Token::LE;
} else {
in_.unget();
return Token::LT;
}
}
//
// REAL
//
if (std::isdigit(c)) {
std::string buf = "";
do {
buf.push_back(c);
c = in_.get();
} while (std::isdigit(c));
if (c == '.') {
buf.push_back(c);
c = in_.get();
while (std::isdigit(c)) {
buf.push_back(c);
c = in_.get();
}
}
in_.unget();
attr.f = std::stod(buf);
return Token::REAL;
}
//
// IDENT or a reserved word
//
if (std::isalpha(c) || c == '_') {
std::string buf = "";
do {
buf.push_back(c);
c = in_.get();
} while (std::isalnum(c) || c == '_');
in_.unget();
attr.s = buf;
Token T = stringToToken(buf);
return T;
}
throw std::runtime_error("Unknown lexeme");
}
/**
* function to get one word from the parsed file
* @param pointer to the word returned
* @return token token, if found, otherwise tokNone
*/
enum eToken parserGetWord(char * const pWord) {
int ch;
char * ptr = pWord;
unsigned short i = 0;
assert(NULL != pWord);
ch = ' ';
while(true)
{
while((' ' == ch) ||
('\t' == ch) ||
('\n' == ch) ||
('\r' == ch)) {
ch = parserGetc();
if('\n' == ch) {
inLineNr++;
}
}
/* we stand on a non empty char */
/* step 1: check if it is #, if so, don't care
* until end of line
*/
if('#' == ch)
{
do
{
ch = parserGetc();
} while('\n' != ch);
inLineNr++;
continue;
}
else
{
break;
}
}
* ptr = ch;
ptr++;
i++;
while(NAMELEN > i)
{
ch = parserGetc();
if((' ' == ch) ||
('\t' == ch) ||
('\n' == ch) ||
('\r' == ch))
{
parserUngetc(ch);
* ptr = '\0';
return stringToToken(pWord);
}
else
{
* ptr = ch;
ptr++;
i++;
}
}
parserExit("name too long");
/* do never come to here, but write it for compilers sake */
return tokNone;
}
