[antlr-interest] How to get ANTLR 3.2 to exit upon first error?

[Dan Becker]

Two things:

1) I incorrectly summarized the wiki page below; if you override
nextToken() you can't throw RecognitionException because nextToken()
also doesn't declare any exceptions. You have to throw
RuntimeException or Error

2) I had also posted this question on stackoverflow; I summarized
Andrew's answer there along with a sample grammar and test runs. See
http://stackoverflow.com/questions/2445008/how-to-get-antlr-3-2-to-exit-upon-first-error.

Thanks again for the help,

Dan

On Mon, Mar 15, 2010 at 8:48 PM, Dan Becker <db19606 at gmail.com> wrote:
> Thanks, Andrew! That works for the parser.
>
> For the lexer, it apparently isn't so simple in java, because
> reportError() doesn't declare any exceptions. There's a wiki page that
> explains what to do for the lexer:
>
> http://www.antlr.org/wiki/pages/viewpage.action?pageId=5341217
>
> The quick summary is either (1) override nextToken() to throw a
> Recognition Exception instead of call reportError() (2) use sneaky
> java tricks to let you throw RecognitionException anyway (3) throw the
> RecognitionException wrapped in a RuntimeException.
>
> Dan
>
> On Mon, Mar 15, 2010 at 4:25 PM, Andrew Haritonkin <thikone at gmail.com> wrote:
>> On Mon, Mar 15, 2010 at 9:41 PM, Kirby Bohling <kirby.bohling at gmail.com> wrote:
>>> You need to repeat some of that for the lexer.  Using the
>>> @lexer::members syntax if you're going to do it as a combined
>>> lexer/parser grammar (I always separate mine to keep my mental working
>>> set smaller).  If you don't do the lexer, you can have a lex error and
>>> recover from it, but this should catch all of the parse errors.
>>>
>>> Kirby
>>>
>>>
>>> On Mon, Mar 15, 2010 at 3:32 PM, Andrew Haritonkin <thikone at gmail.com> wrote:
>>>> For Java and C# target add this in the beginning of the grammar but
>>>> after grammar statement:
>>>>
>>>> grammar grammar1;
>>>>
>>>> @members {
>>>> protected override object RecoverFromMismatchedToken(IIntStream input,
>>>> int ttype, BitSet follow)
>>>> {
>>>>        throw new MismatchedTokenException(ttype, input);
>>>> }
>>>> public override object RecoverFromMismatchedSet(IIntStream input,
>>>> RecognitionException e, BitSet follow)
>>>> {
>>>>        throw e;
>>>> }
>>>> }
>>>>
>>>> @rulecatch {
>>>> catch (RecognitionException e)
>>>> {
>>>>    throw e;
>>>> }
>>>> }
>>>>
>>>> Hope it helps,
>>>> Andrew
>>>>
>>
>> Indeed, forgot about lexer... So, the full code which you need to add
>> to the grammar for C# target would be:
>>
>> grammar grammar1;
>>
>> @lexer::members {
>> public override void ReportError(RecognitionException e)
>> {
>>        throw e;
>> }
>> }
>>
>> @parser::members {
>> protected override object RecoverFromMismatchedToken(IIntStream input,
>> int ttype, BitSet follow)
>> {
>>        throw new MismatchedTokenException(ttype, input);
>> }
>> public override object RecoverFromMismatchedSet(IIntStream input,
>> RecognitionException e, BitSet follow)
>> {
>>        throw e;
>> }
>> }
>>
>> @rulecatch {
>> catch (RecognitionException e)
>> {
>>    throw e;
>> }
>> }
>>
>> Similar should be for Java target, only naming convention is different.
>>
>> Andrew
>>
>
>
>
> --
> Dan Becker
> 303/497-6824
>



-- 
Dan Becker
303/497-6824