bool CScanner::NextToken()
{
SkipWhiteSpace();
if (ErrorString != "")
return false;
// if (Pos >= SourceLength)
// return false;
_off_t OldPos = Pos;
char c = cget();
if (isdigit(c) || // Thinks a little, while reading this condition, it might be a bit unreadable.
(
(
c == '.' &&
(
(IncPos(), isdigit(cget())) || (DecPos(), false)
)
) &&
(DecPos(), true)
)
)
return ParseNumber() && OldPos != Pos;
else if (isalpha(c) || c == '_')
return ParseIdentifier() && OldPos != Pos;
else if (MatchPrefix("'"))
return ParseCharacterLiteral() && OldPos != Pos;
else if (MatchPrefix("\""))
return ParseStringLiteral() && OldPos != Pos;
else if (SymbolicTokenMapping.find(string() += c))
return ParseSymbol() && OldPos != Pos;
else if (c == 0)
return true;
else
ErrorString = "Unrecognized character at line: ";// + itoa(CurrentLine) + ", Column: " + itoa(CurrentColumn);
return false;
}
bool CScanner::ParseNumber()
{
UpdateCurrentToken();
string TempString = "";
while (isalnum(cget()) || cget() == '.' || cget() == '+' || cget() == '-')
TempString += cget(), IncPos();
string ClearString = "uUlL";
if (
TempString[1] != 'x' && TempString[1] != 'X'
// TempString.find('e') != string::npos ||
// TempString.find('E') != string::npos ||
// TempString.find('.') != string::npos ||
)
ClearString += "fF";
while (ClearString.find(TempString[TempString.length() - 1]) != string::npos)
TempString.erase(TempString.length()- 1);
CurrentToken.Kind = DTM_TOKEN::LITERAL::NUMBER::INTEGER;
if (TempString[1] == 'x' || TempString[1] == 'X')
try
{
CurrentToken.Value = FromHexadecimalString(TempString);
return true;
}
catch (...)
{
ErrorString = "Bad hexadecimal integral literal.";
return false;
}
try
{
CurrentToken.Value = lexical_cast<int>(TempString);
return true;
}
catch (...)
{
}
try
{
CurrentToken.Value = lexical_cast<float>(TempString);
CurrentToken.Kind = DTM_TOKEN::LITERAL::NUMBER::FLOAT;
return true;
}
catch (...)
{
ErrorString = "Bad number literal.";
}
return true;
}
bool CScanner::ParseIdentifier()
{
UpdateCurrentToken();
string TempString = "";
while (isalnum(cget()) || cget() == '_')
TempString += cget(), IncPos();
if (isKeyword(TempString))
CurrentToken.Kind = GetKeywordKind(TempString);
else
CurrentToken.Kind = DTM_TOKEN::IDENTIFIER;
CurrentToken.Value = TempString;
return true;
}
bool CScanner::MatchPrefix(const string &Prefix, unsigned int Offset /*= 0*/)
{
if (Offset >= Prefix.length())
return true;
bool Result = false;
if (Source[Pos] == Prefix[Offset])
{
IncPos();
Result = MatchPrefix(Prefix, Offset + 1);
if (!Result)
DecPos();
}
return Result;
}
bool CScanner::ParseSymbol()
{
string TempString = "";
UpdateCurrentToken();
while (isSymbol(cget()) && TempString.length() < 3)
TempString += cget(), IncPos();
while(!SymbolicTokenMapping.find(TempString))
{
TempString.erase(TempString.end() - 1), DecPos();
}
CurrentToken.Kind = SymbolicTokenMapping[TempString];
CurrentToken.Value.reset();
CurrentToken.Value = TempString;
return true;
}
bool CScanner::ParseStringLiteral()
{
UpdateCurrentToken();
string TempString = "";
while (cget() != '"' && cget() != 0)
{
if (isNewline())
{
ErrorString = "Newline in string literal.";
return false;
}
if (cget() == '\\')
{
IncPos();
if (cget() == 'x')
{
IncPos();
string VeryTempString = "x";
while(isxdigit(cget()))
VeryTempString += cget(), IncPos();
try
{
TempString += EscapeSequenceToValue(VeryTempString);
continue;
}
catch (positive_overflow &e)
{
ErrorString = "Hexadecimal constant is invalid (may be too large for character).";
return false;
}
}
else
{
TempString += EscapeSequenceToValue(string() += cget());
IncPos();
continue;
}
}
TempString += cget();
IncPos();
}
IncPos();
CurrentToken.Kind = DTM_TOKEN::LITERAL::STRING;
CurrentToken.Value = TempString;
return true;
}
bool CScanner::ParseCharacterLiteral()
{
enum
{
BEGIN,
ESCAPE_SEQUENCE_FOUND,
HEXADECIMAL_ESCAPE_SEQUENCE_FOUND,
BEGIN_FINISHING,
FINISHED,
};
enum
{ // We discovered that between the ' ' is:
R_SIMPLE, // just a c-character
R_ESCHEX, // hexadecimal escape sequence
R_ESCSEQ, // simple escape sequence
R_NOTHIN, // <- Initial state.
};
int State = BEGIN;
int ResultKind = R_NOTHIN;
UpdateCurrentToken();
CurrentToken.Kind = DTM_TOKEN::LITERAL::NUMBER::INTEGER;
string TempString = "";
bool ErrFlag = cget() == 0;
while (State != FINISHED && !ErrFlag)
{
switch(State)
{
case BEGIN:
ErrFlag = cget() == '\'';
TempString = cget();
if (cget() == '\\')
State = ESCAPE_SEQUENCE_FOUND;
else
State = BEGIN_FINISHING, ResultKind = R_SIMPLE;
break;
case ESCAPE_SEQUENCE_FOUND:
ErrFlag = cget() == '\'';
TempString = cget();
if (cget() == 'x')
State = HEXADECIMAL_ESCAPE_SEQUENCE_FOUND, ResultKind = R_ESCHEX;
else
State = BEGIN_FINISHING, ResultKind = R_ESCSEQ;
break;
case HEXADECIMAL_ESCAPE_SEQUENCE_FOUND:
ErrFlag = cget() == '\'';
while (isxdigit(cget()))
TempString += cget(), IncPos();
State = BEGIN_FINISHING, DecPos();
break;
case BEGIN_FINISHING:
ErrFlag = cget() != '\'';
State = FINISHED;
break;
}
IncPos();
}
if (ErrFlag)
{
ErrorString = "Invalid character constant.";
return false;
}
switch (ResultKind)
{
case R_SIMPLE:
CurrentToken.Value = TempString[0];
break;
case R_ESCHEX: case R_ESCSEQ:
try
{
CurrentToken.Value = EscapeSequenceToValue(TempString);
}
catch (...) // positive_overflow ?
{
ErrFlag = true;
}
break;
case R_NOTHIN:
break;
}
if (ErrFlag)
ErrorString = "Hexadecimal constant is invalid (may be too large for character).";
return !ErrFlag;
}
bool CScanner::SkipWhiteSpace()
{
enum EScannerState
{
SEARCH_FOR_WHITE_SPACE,
NEXT_TOKEN,
BEGIN_COMMENT,
C_COMMENT_BODY,
CPP_COMMENT_BODY,
CPP_COMMENT_GOING_NEWLINE,
PREPARE_TO_EXIT_C_COMMENT_BLOCK,
PREPARE_TO_EXIT_CPP_COMMENT_BLOCK,
BEGIN_PP,
PP_GOING_NEW_LINE,
BEGIN_SYMBOL,
BEGIN_IDENTIFIER,
BEGIN_NUMBER,
END_SCAN,
};
char c;
EScannerState State = SEARCH_FOR_WHITE_SPACE;
while (cget() != 0)
{
c = cget();
if (c == '\n')
IncLineCounter();
switch (State)
{
case SEARCH_FOR_WHITE_SPACE:
switch(c)
{
case '\t': case'\n': case '\v': case '\r': case ' ': case '\f':
break;
case '/':
State = BEGIN_COMMENT;
break;
case '#':
State = BEGIN_PP;
break;
default:
return true;
}
break;
case BEGIN_COMMENT:
if (c == '*')
State = C_COMMENT_BODY;
else if (c == '/')
State = CPP_COMMENT_BODY;
else
{
State = BEGIN_SYMBOL;
Pos--;
return true;
}
break;
case C_COMMENT_BODY:
if (c != '*')
break;
else
State = PREPARE_TO_EXIT_C_COMMENT_BLOCK;
break;
case PREPARE_TO_EXIT_C_COMMENT_BLOCK:
switch (c)
{
case '*':
break;
case '/':
State = SEARCH_FOR_WHITE_SPACE;
break;
default:
State = C_COMMENT_BODY;
break;
}
break;
case CPP_COMMENT_BODY:
if (c == '\\')
State = CPP_COMMENT_GOING_NEWLINE;
else if (c != '\r' && c != '\n')
break;
else if (c == '\n')
State = SEARCH_FOR_WHITE_SPACE;
break;
case CPP_COMMENT_GOING_NEWLINE:
if (c == '\r' || c == '\\')
break;
State = CPP_COMMENT_BODY;
break;
case BEGIN_PP:
switch(c)
{
case '\\':
State = PP_GOING_NEW_LINE;
break;
case '\n':
State = SEARCH_FOR_WHITE_SPACE;
break;
default:
break;
}
break;
case PP_GOING_NEW_LINE:
if (c == '\r' || c == '\\')
break;
State = BEGIN_PP;
break;
default:
assert(false);
break;
}
IncPos();
}
if (State == C_COMMENT_BODY || State == PREPARE_TO_EXIT_C_COMMENT_BLOCK)
{
ErrorString = "Unclosed multi-line comment.";
return false;
}
return true;
}
const CUILine* CUILine::CutByLength(CGameFont* pFont, float length, BOOL cut_word)
{
R_ASSERT(GetSize() > 0);
// if first sub line is void then delete it
Position pos;
InitPos(pos);
if (!pos.word_1.exist()) // void string
{
if (m_subLines[0].m_last_in_line)
{
m_subLines.erase(m_subLines.begin());
return GetEmptyLine();
}
else
{
m_subLines.erase(m_subLines.begin());
InitPos(pos);
}
}
float len2w1 = GetLength_inclusiveWord_1(pos, pFont);
//. * scale; bacause of our fonts not scaled
if (!pos.word_2.exist())
{
if (cut_word && len2w1 > length)
return CutWord(pFont, length);
else
return Cut2Pos(pos);
}
float len2w2 = GetLength_inclusiveWord_2(pos, pFont);
if (len2w1 > length)
{
if (cut_word)
return CutWord(pFont, length);
else
return Cut2Pos(pos);
}
else if (len2w1 <= length && len2w2 > length)
{
// cut whole first word
return Cut2Pos(pos); // all right :)
}
else // if (len2w1 > length && len2w2 > length)
{
while (IncPos(pos))
{
len2w1 = GetLength_inclusiveWord_1(pos, pFont);
if(!pos.word_2.exist())
{
return Cut2Pos(pos);
}
len2w2 = GetLength_inclusiveWord_2(pos, pFont);
if (len2w1 <= length && len2w2 > length)
return Cut2Pos(pos);
}
return Cut2Pos(pos, false);
}
}
