Re: Duplicate JNI files.

[Etienne M. Gagnon]

On Fri, Sep 07, 2001 at 08:11:50PM -0600, Tom Tromey wrote:
> Brian> I'm almost done eith a full javap replacement that uses GNU
> Brian> bytecode.  Writing javah is trivial as well with this package
> Brian> and I'll do it soon.  Getting the decompilation to java
> Brian> assembly to display as I want (just as Sun's javap does) has
> Brian> proven tricky.
> 
> FYI gcj includes jcf-dump, which is similar to javap.  It isn't
> exactly the same, but it does have bytecode disassembly.

Shameless plug... ;-)

Have you looked, guys, at the Soot project (http://www.sable.mcgill.ca/soot)?
It is a complete bytecode optimization framework.  It has proven to improve
java programs execution speed, even on HotSpot. [See the papers page for 
details].

Internally, it supports many representations, which can be pretty handy for
programmers that want to look at classes.  It even does type inference for
local variables (that's my bit, proven to be NP-hard, but we have a practical
cheat to get a fast polynomial solution: read the papers again;-)

But, more importantly, Jerome Miecznikowski added a code restructurer (e.g.
get rid of goto's) on top of all this.  He calls the result "dava", which is
kind of a pretty neat decompiler.  Last message I got about it, it was supposed
to be in "beta" stage.  You can have a look at 
http://www.sable.mcgill.ca/~jerome/public/

Just though the stuff might interest you...

Etienne
-- 
Etienne M. Gagnon                    http://www.info.uqam.ca/~egagnon/
SableVM:                                       http://www.sablevm.org/
SableCC:                                       http://www.sablecc.org/