Mismatched input after converting grammar from Antlr2 to Antlr4

Question

I have an existing grammar I'm converting to Antlr4 using the C# output. It's failing on the first token with the message mismatched input 'Begin' expecting ANY_WORD. Strangely though when I print out the tokens to see what token it is being recognised as, it indicates it has identified it correctly.

            string text = File.ReadAllText(fi.FullName);
            var fs = new Antlr4.Runtime.AntlrInputStream(text);
            var lexer = new TestLexer(fs);
            var tokens = new Antlr4.Runtime.CommonTokenStream(lexer);
            tokens.Fill();
            var tokenList = tokens.GetTokens();
            yRLogger.Info("Printing tokens");
            foreach(IToken tok in tokenList)
            {
                Logger.Info(tok.ToString()); //prints [@0,13:17='Begin',<28>,1:13] 28 In TestLexer.tokens, 28 corresponds to ANY_WORD

            }
            var parser = new TestParser(tokens);
            parser.start(); //throws 'mismatched input 'Begin' expecting ANY_WORD. '

Checking the same grammar and input in AntlrWorks also behaves as expected - it finds the first token without an error.

From TestParser.tokens ANY_WORD=28.

The input file is

Begin parsing now

The Antlr file is

   start : ANY_WORD+;

   ANY_WORD
: 
~('\r'|'\n'|' '|'\t'|'\u000C')+
;

    WS  : (' '|'\t'|'\u000C') -> skip 
;

Sam Harwell Sam Harwell · Accepted Answer · 2014-03-27T11:26:29

This belongs on the issue tracker for the C# target.
I just tried to reproduce this issue, configuring the C# target as described in the Readme instructions for the latest version of the project, and parsing worked correctly on the first try.

Mismatched input after converting grammar from Antlr2 to Antlr4

1 Answers