Another route is to generate bytecodes, but make sure the generated 
bytecodes are the same as what a Java compiler would generate.  Then you 
can use a java decompiler on them.  Probably doesn't help in the 
debugger, but does help you understand what your compiler generates.

Best,
Martin

John Wilson wrote:
> 2009/11/22 Matt Fowles <[email protected]>:
>> Robert~
>>
>> Ease of debugging mostly.  Our production environment always goes straight
>> to bytecode, but when we are debugging the compiler it is much nicer to be
>> able to step through java code and get links from it back to compiler code
>> the generated it.
> 
> 
> We looked at this in the early days of Groovy. It my be a decent route
> for your language but it wasn't great for Groovy. The problem is that
> there are some Java features (e.g. checked exceptions) which are
> enforced by the compiler but not the JVM. If your language shares all
> the restrictions that the Java compiler enforces then it's a perfectly
> good option. Groovy doesn't share many of the Java restrictions so it
> didn't seem worth it.
> 
> Also generating bytecodes is not that hard once you get up the
> learning curve a bit so I think most people end up just going straight
> to bytecode because it's easier in the end.
> 
> John Wilson
> 
> --
> 
> You received this message because you are subscribed to the Google Groups 
> "JVM Languages" group.
> To post to this group, send email to [email protected].
> To unsubscribe from this group, send email to 
> [email protected].
> For more options, visit this group at 
> http://groups.google.com/group/jvm-languages?hl=.
> 
> 

--

You received this message because you are subscribed to the Google Groups "JVM 
Languages" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/jvm-languages?hl=.


Reply via email to