Can you detect a method that will be 65535 bytes?

Question

There's a well-known problem with the JVM that a method cannot be greater than 65535 bytes. Obviously, most developers know better than to write such a large method by hand. However, as automatic source code generators become more popular, this method size limit is becoming a major headache.
http://forum.java.sun.com/thread.jspa?forumID=37&threadID=171744
http://forum.java.sun.com/thread.jspa?forumID=52&threadID=94978
http://forum.java.sun.com/thread.jspa?threadID=483376&start=0&tstart=0
http://forum.java.sun.com/thread.jspa?forumID=31&threadID=510932
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=4309152
Apparently the only workaround to this problem is to split such a large method into multiple smaller methods that don't exceed the size limit. Well, that sounds great, but how do you actually do this?
For example, in an automatic source code generator, is it possible to detect if generated code would exceed this 65535 limit? Maybe it's impossible to tell ahead-of-time how big the Java compiler will make a method. If one could easily tell that a method would exceed the size limit, it would be easier to write a code generator.
Any thoughts?

Answer 1

Thank you, jschell, for your wonderully helpful
posts. Your positive attitude is just what these
forums need.
As I said before, this generator is generating
scientific code, the code has to be this long to
solve the given problem. And since the JVM spec
places a constraint on method length (and we all have
our opinions about this, but while our opinions are
interesting, they are irrelevant) I need to find a
way to split the big method into multiple smaller
ones. Many others have voiced the same frustration
with this (do a search for "code length longer than
65535 byes" in these forums).I was rather frustrated when I ran across the C code generator problem as well.
>
So this post is just an attempt to get ideas or see
if someone else has faced this problem before. Your
condescending posts are not only offensive, but they
do absolutely nothing to help solve the problem.Your posts seemed to suggest, rather specifically, that you thought this was a problem with the JVM (and the spec.)
I was pointing out that that is not the case.
You also seemed to be asking for something that would only be useful to a code generator. And I also pointed out that someone who is writing a code generator, rather than just using one, can fix this problem by simply not allowing that to occur in the first place. Which is how I write my code generators.
So are your writing a code generator? Then change it so it doesn't have this problem.
Are you using a code generator? If so write a 'breaker'. It takes the output from the code generator and restructures the code into smaller pieces. By the way that is exactly what I did with the C problem that I mentioned.
And those are the solutions.
Please refrain from posting here again unless you
u have something constructive to say. Thank you.If you wish to pay me money (negotiable of course) then I am more than willing to say anything you wish or not say anything as the case may be.
Or you can post your problem to your own web site where you can have complete control over the content of all the answers.
Other than that you are out of luck.