[antlr-interest] philosophy about translation

Andy Tripp antlr at jazillian.com
Sat Oct 7 10:44:34 PDT 2006


Terence Parr wrote:

>
> On Oct 6, 2006, at 11:35 AM, Andy Tripp wrote:
>
>> The distinction because I think a lot of people mistakenly believe  
>> that tools like ANTLR and
>> ASTs will "scale up" from simple stuff to real, complete translation.
>
>
> I have done some nasty translators, but none on the scale of your  
> COBOL to Java translator.  Gary Funck and I worked on a Pascal like  
> language to C translator (he did the work) that use tree  
> transformation and it had lots of nasty things to do like  
> topologically sorting the declarations so that C was happy with them.  
> so clearly it's possible--we are only talking about convenience in  
> this thread probably. of course that can make the difference between  
> finishing something and not. ;)
>
> Ter
>
I've found that, at least for C to Java, I spend most of my time mapping 
the library calls, not the core language itself.
And most of the core language work seems to be in getting rid of pointers.
The libraries are where a lot - probably most- of the hard cases are.
memset() and memcpy() are not my friends :(



More information about the antlr-interest mailing list