[antlr-interest] Using range operator (INT .. INT)

[Jiwon Seo]

I'm attaching the complete grammar file.

I'd really appreciate if you take a look at it.
As is mentioned in the previous email, the following is the main change I made.

FLOAT
    : ('0'..'9')+
    (
        {input.LA(2) != '.'}? => '.' ('0'..'9')+ EXPONENT? {$type = FLOAT;}
        |  {$type = INT;}
    )
    |  ('0'..'9')+ EXPONENT
    ;

Thanks.


On Sun, Nov 13, 2011 at 11:06 PM, Jiwon Seo <jiwon at stanford.edu> wrote:
> I did the following change:
>
> FLOAT
>    : ('0'..'9')+
>    (
>        {input.LA(2) != '.'}? => '.' ('0'..'9')+ EXPONENT? {$type = FLOAT;}
>        |  {$type = INT;}
>    )
>    |  ('0'..'9')+ EXPONENT
>    ;
>
> I tested it inside antlrworks, and very strangely, it works fine with
> the debugger(generates the right parse tree),
> but raises Mismatched token exception with the interpreter.
>
> Is the above right way to do this?
>
>
> On Sun, Nov 13, 2011 at 5:10 PM, Jim Idle <jimi at temporal-wave.com> wrote:
>> Please read the wiki articles - your rules conflict and there are simple
>> ways around this :)
>>
>> Jim
>>
>>> -----Original Message-----
>>> From: antlr-interest-bounces at antlr.org [mailto:antlr-interest-
>>> bounces at antlr.org] On Behalf Of Jiwon Seo
>>> Sent: Sunday, November 13, 2011 9:58 AM
>>> To: Christian
>>> Cc: antlr-interest at antlr.org
>>> Subject: Re: [antlr-interest] Using range operator (INT .. INT)
>>>
>>> I attached a simplified grammar file with the problem.
>>> If you run the interpreter in antlrworks 1.4.3 the following will give
>>> an error
>>>
>>> int n:0..10
>>>
>>> But the following won't give an error:
>>>
>>> int n:0 .. 10    // white space before and after ..
>>>
>>>
>>> Note that my definition of FLOAT does not conflict with the range
>>> operator:
>>>
>>> FLOAT
>>>     :   ('0'..'9')+ '.' ('0'..'9')+ EXPONENT?
>>>     |   ('0'..'9')+ EXPONENT
>>>     ;
>>>
>>>
>>> Thanks.
>>>
>>> On Sun, Nov 13, 2011 at 3:09 AM, Christian <chwchw at gmx.de> wrote:
>>> > Hello Jiwon,
>>> >
>>> > first, try to write a simple minimal grammar that represents your
>>> > usecase because I do not see anything that explains the behavior you
>>> > described. Maybe there are some other rules that influence the
>>> parsing.
>>> >
>>> > Tell us afterwards.
>>> >
>>> > Regards,
>>> > Christian
>>> >
>>> > Am 13.11.2011 10:31, schrieb Jiwon Seo:
>>> >> I'm trying to add range operator in my language, and I'm having a
>>> >> problem with it.
>>> >>
>>> >> The range operator is defined as following:
>>> >>
>>> >> range: INT '..' INT ;
>>> >>
>>> >> where INT is '0'..'9'+
>>> >>
>>> >> I do have FLOAT, but it is defined such that it does not have
>>> >> conflict with the range operator:
>>> >>
>>> >> FLOAT: ('0'..'9')+'.'('0'..'9')+ EXPONENT?
>>> >>             | ('0'..'9')+ EXPONENT
>>> >>           ;
>>> >>
>>> >> The thing is, if I use the range operator with white spaces before &
>>> >> after the dots(such as 10 .. 100), it parses fine But without the
>>> >> white spaces(such as 10..100), it says 'MismatchedTokenException' .
>>> >>
>>> >> Can anyone see why it's acting like that? I do not use dot ('.') in
>>> >> any other places.
>>> >>
>>> >> I'm using antlrworks 1.4.3 for writing grammar and testing it
>>> >> (interpreter & debugger)
>>> >>
>>> >> Thanks.
>>> >>
>>> >> List: http://www.antlr.org/mailman/listinfo/antlr-interest
>>> >> Unsubscribe:
>>> >> http://www.antlr.org/mailman/options/antlr-interest/your-email-
>>> addres
>>> >> s
>>> >>
>>> >
>>> >
>>> > List: http://www.antlr.org/mailman/listinfo/antlr-interest
>>> > Unsubscribe:
>>> > http://www.antlr.org/mailman/options/antlr-interest/your-email-
>>> address
>>> >
>>
>> List: http://www.antlr.org/mailman/listinfo/antlr-interest
>> Unsubscribe: http://www.antlr.org/mailman/options/antlr-interest/your-email-address
>>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: simple.g
Type: application/octet-stream
Size: 1564 bytes
Desc: not available
Url : http://www.antlr.org/pipermail/antlr-interest/attachments/20111113/516fc660/attachment.obj