/// <summary>
/// This is the method used by Piglets parserfactory to obtain preconfigured lexers.
/// </summary>
/// <param name="grammar">Grammar to generate lexers from</param>
/// <param name="lexerSettings">Additional lexing settings</param>
/// <returns>A lexer compatibe with the given grammars tokenizing rules</returns>
internal static ILexer <T> ConfigureFromGrammar(IGrammar <T> grammar, ILexerSettings lexerSettings) =>
// This works because the grammar tokens will recieve the same token number
// since they are assigned to this list in just the same way. AND BECAUSE the
// end of input token is LAST. if this is changed it WILL break.
// This might be considered dodgy later on, since it makes it kinda sorta hard to
// use other lexers with Piglet. Let's see what happens, if anyone ever wants to write their
// own lexer for Piglet.
Configure(c =>
{
c.Runtime = lexerSettings.Runtime;
c.IgnoreCase = lexerSettings.IgnoreCase;
List <ITerminal <T> > terminals = grammar.AllSymbols.OfType <ITerminal <T> >().ToList();
foreach (ITerminal <T> terminal in terminals)
{
if (terminal.RegExp != null)
{
c.Token(terminal.RegExp, terminal.OnParse);
}
}
c.EndOfInputTokenNumber = terminals.FindIndex(f => f == grammar.EndOfInputTerminal);
foreach (string ignored in lexerSettings.Ignore)
{
c.Ignore(ignored);
}
});
public ParserConfigurator()
{
nonTerminals = new List <NonTerminal <T> >();
terminals = new LinkedList <Terminal <T> >();
lexerSettings = new LexerSettingsImpl();
terminalPrecedences = new List <TerminalPrecedence>();
currentPrecedence = 0;
// Create the Error token. This will create it as terminal 0, but in the end it will be the LAST terminal
// second last is EndOfInput. This is sort of hackish and mainly due to the way the lexer is configured.
ErrorToken = CreateTerminal(null, s => default(T));
ErrorToken.DebugName = "%ERROR%";
// Set some default settings
LexerSettings.CreateLexer = true;
LexerSettings.EscapeLiterals = true;
LexerSettings.Ignore = new[] { "\\s+" }; // Ignore all whitespace by default
}
