/**
* Symbol parser generates token of type TokenType.Symbol.
* If successful, 'true' is returned, 'result' contains token, 'next' contains further index information.
* If not successful, 'false' is returned, other values can be arbitrary.
*/
public static bool ParseSymbol(string source, int index, out SymbolToken result, out int next)
{
result = null;
next = -1;
var specialDouble = new[]
{">>", "<<", "**", "<>", "><", "<=", ">=", ":=", "+=", "-=", "*=", "/=", "(*", "*)", "(.", ".)", "//"};
if (specialDouble.Any(s => TokenParseUtility.HasPrefix(source, index, s)))
{
result = new SymbolToken();
result._from = index;
result._to = index + 2;
next = index + 2;
return true;
}
if (index < source.Length)
{
return false;
}
var ch = source[index];
if ("ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789’+-*/=<>[].,():^@{}$#&%".Contains(ch))
{
result = new SymbolToken();
result._from = index;
result._to = index + 1;
next = index + 1;
return true;
}
return false;
}
/**
* Comment parser generates token of type TokenType.Comment.
* If successful, 'true' is returned, 'result' contains token, 'next' contains further index information.
* If not successful, 'false' is returned, other values can be arbitrary.
*
* Note, that list of comment tokens are returned!
*/
public static bool ParseComment(string source, int index, out List <CommentToken> result, out int next)
{
LinkedList <(int, string)> closeStack = new LinkedList <(int, string)>();
bool singleLine = false;
var startIndex = index;
if (TokenParseUtility.HasPrefix(source, index, "(*"))
{
closeStack.AddFirst((index, "*)"));
index += 2;
}
else if (TokenParseUtility.HasPrefix(source, index, "{"))
{
closeStack.AddFirst((index, "}"));
index += 1;
}
else if (TokenParseUtility.HasPrefix(source, index, "//"))
{
index += 2;
singleLine = true;
}
else
{
result = null;
next = -1;
return(false);
}
result = new List <CommentToken>();
while (index <= source.Length)
{
if (singleLine && (index == source.Length || source[index] == '\n'))
{
// single line comment ended
var singleCommentToken = new CommentToken();
singleCommentToken._from = startIndex;
singleCommentToken._to = index;
result.Add(singleCommentToken);
next = index;
return(true);
}
// for other cases, break the loop
if (index == source.Length)
{
break;
}
if (closeStack.Count > 0)
{
var i = 0;
var prevCommType = "\n";
foreach (var pair in closeStack)
{
if (pair.Item2 == "\n")
{
i++;
continue;
}
prevCommType = pair.Item2;
break;
}
if (TokenParseUtility.HasPrefix(source, index, prevCommType))
{
// prefix elimination
for (var k = 0; k <= i; k++)
{
// put all disposed inner comments (newlined mostly)
var pref = closeStack.First();
closeStack.RemoveFirst();
CommentToken ct = new CommentToken();
ct._from = pref.Item1;
ct._to = index;
if (k == i)
{
ct._to += pref.Item2.Length;
}
result.Add(ct);
}
index += prevCommType.Length;
// put intermediate comment
if (closeStack.Count == 0 && !singleLine)
{
// that's all ended
next = index;
return(true);
}
continue; // don't forget to continue main loop with new index
}
}
if (TokenParseUtility.HasPrefix(source, index, "(*"))
{
closeStack.AddFirst((index, "*)"));
index += 2;
continue;
}
else if (TokenParseUtility.HasPrefix(source, index, "{"))
{
closeStack.AddFirst((index, "}"));
index += 1;
continue;
}
else if (TokenParseUtility.HasPrefix(source, index, "//"))
{
closeStack.AddFirst((index, "\n"));
index += 1;
continue;
}
// else this is just the char, skip...
index++;
}
next = index;
return(closeStack.Count == 0);
}
