[antlr-interest] [c target] some questions about errors catching

[Jim Idle]

On 09/27/2009 07:22 AM, Юрушкин Михаил wrote:
> good day.
> I have some questions about an errors-processing alogirthm in Antlr:
>
> 1) Antlr generates strange code. If no alternatives are right, it does the
> following:
>
> 	    ...
>               default:
>                   CONSTRUCTEX();
>                   EXCEPTION->type         = ANTLR3_NO_VIABLE_ALT_EXCEPTION;
>                   EXCEPTION->message      = (void *)"";
>                   EXCEPTION->decisionNum  = 34;
>                   EXCEPTION->state        = 0;
>
>
>                   goto ruleattr_specEx;
>               }
>
>
> - initializes exception... and CLEANS some information. Why?
>    

Error recovery. Read the docs for the exception handling and error 
recovery logic if you want to know more, but for error reporting and so 
on, you don't really need to know much - just use the pre-supplied error 
message display routine as a guide - I tried to make it access 
everything that one woudl need so you can see what to do. See 
displayRecognitionError.

In release 3.2.1 I am going to fold that code into an internal function 
to reduce the line count of the output code and possibly the object code 
size.

> 2) after the finishing of processing nonterminal
> rec->state->errors_count = 0 Why?? It has detected an error already.
>
>    
I presume you mean errorCount? Is this from the lexer? I fixed a bug 
whereby the lexer would not count errors - fixed in release 3.2. 
However, lexers should really be written so they don't throw recognition 
errors, but detect anomalies and report them more directly (such as 
missing terminating quotes and illegal characters. FInally don;t use 
intenrals directly, use the api calls: getNumberOfSyntaxErrors

Jim