/// <summary>
/// Scans a literal string, contained between "(" and ")".
/// </summary>
public Symbol ScanLiteralString()
{
// Reference: 3.2.3 String Objects / Page 53
// Reference: TABLE 3.32 String Types / Page 157
Debug.Assert(_currChar == Chars.ParenLeft);
_token = new StringBuilder();
int parenLevel = 0;
char ch = ScanNextChar(false);
// Phase 1: deal with escape characters.
while (ch != Chars.EOF)
{
switch (ch)
{
case '(':
parenLevel++;
break;
case ')':
if (parenLevel == 0)
{
ScanNextChar(false);
// Is goto evil? We could move Phase 2 code here or create a subroutine for Phase 1.
goto Phase2;
}
parenLevel--;
break;
case '\\':
{
ch = ScanNextChar(false);
switch (ch)
{
case 'n':
ch = Chars.LF;
break;
case 'r':
ch = Chars.CR;
break;
case 't':
ch = Chars.HT;
break;
case 'b':
ch = Chars.BS;
break;
case 'f':
ch = Chars.FF;
break;
case '(':
ch = Chars.ParenLeft;
break;
case ')':
ch = Chars.ParenRight;
break;
case '\\':
ch = Chars.BackSlash;
break;
// AutoCAD PDFs my contain such strings: (\ )
case ' ':
ch = ' ';
break;
case Chars.CR:
case Chars.LF:
ch = ScanNextChar(false);
continue;
default:
if (char.IsDigit(ch)) // First octal character.
{
// Octal character code.
if (ch >= '8')
{
ParserDiagnostics.HandleUnexpectedCharacter(ch);
}
int n = ch - '0';
if (char.IsDigit(_nextChar)) // Second octal character.
{
ch = ScanNextChar(false);
if (ch >= '8')
{
ParserDiagnostics.HandleUnexpectedCharacter(ch);
}
n = n * 8 + ch - '0';
if (char.IsDigit(_nextChar)) // Third octal character.
{
ch = ScanNextChar(false);
if (ch >= '8')
{
ParserDiagnostics.HandleUnexpectedCharacter(ch);
}
n = n * 8 + ch - '0';
}
}
ch = (char)n;
}
else
{
//TODO
// Debug.As sert(false, "Not implemented; unknown escape character.");
ParserDiagnostics.HandleUnexpectedCharacter(ch);
}
break;
}
break;
}
default:
break;
}
_token.Append(ch);
ch = ScanNextChar(false);
}
// Phase 2: deal with UTF-16BE if necessary.
// UTF-16BE Unicode strings start with U+FEFF ("þÿ"). There can be empty strings with UTF-16BE prefix.
Phase2:
if (_token.Length >= 2 && _token[0] == '\xFE' && _token[1] == '\xFF')
{
// Combine two ANSI characters to get one Unicode character.
StringBuilder temp = _token;
int length = temp.Length;
if ((length & 1) == 1)
{
// TODO What does the PDF Reference say about this case? Assume (char)0 or treat the file as corrupted?
temp.Append(0);
++length;
DebugBreak.Break();
}
_token = new StringBuilder();
for (int i = 2; i < length; i += 2)
{
_token.Append((char)(256 * temp[i] + temp[i + 1]));
}
return(_symbol = Symbol.UnicodeString);
}
// Adobe Reader also supports UTF-16LE.
if (_token.Length >= 2 && _token[0] == '\xFF' && _token[1] == '\xFE')
{
// Combine two ANSI characters to get one Unicode character.
StringBuilder temp = _token;
int length = temp.Length;
if ((length & 1) == 1)
{
// TODO What does the PDF Reference say about this case? Assume (char)0 or treat the file as corrupted?
temp.Append(0);
++length;
DebugBreak.Break();
}
_token = new StringBuilder();
for (int i = 2; i < length; i += 2)
{
_token.Append((char)(256 * temp[i + 1] + temp[i]));
}
return(_symbol = Symbol.UnicodeString);
}
return(_symbol = Symbol.String);
}
