[Soot-list] Jimple local to source expression

[Marc-Andre Laverdiere-Papineau]

I just wanted to throw some ideas around for this...

Is there a way we could just annotate the Jimple statements with the 
source reference? That would keep the Jimple smaller and easier to 
analyze, wouldn't it?
I guess that transformations would have to carry the annotations all 
over the place on a best effort basis. Or maybe some heuristics is needed.

Would that be making sense? And be implementable?

On 12-08-26 11:17 AM, Patrick Lam wrote:
> On 08/26/2012 03:53 AM, Saswat Anand wrote:
>>> I don't know what happens if you don't transform the Jimple at all. I
>>> think there's a switch to do that, but I can't remember what it is.
>>
>> I would be curious to know what happens and what is that flag. It
>> would be nice to be able to configure soot not to worry about
>> efficient bytecode generation, but preserve the mapping from jimple
>> elements to the .class or .java file.
>
> Well, give it a try:
>
> java soot.Main -f j -cp
> .:/usr/lib/jvm/java-6-openjdk-amd64/jre/lib/rt.jar -p jb enabled:false ...
>
> That's for Jimple from bytecode.
>
>> It might be nice to use Eclipse's ASTParser to parse Java code and
>> build jimple from the AST. Did anybody explore this route?
>
> Soot also has a Java parser, but not Eclipse's. If you want to disable
> cleanup for that parser, try -p jj enabled:false.
>
> Warning: the blowup in the code size may make it impossible to analyze
> the code in a reasonable amount of time and space, though. The code is
> also going to get a bunch of nops, etc.
>
> pat
>
> _______________________________________________
> Soot-list mailing list
> Soot-list at sable.mcgill.ca
> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>


-- 
Marc-André Laverdière-Papineau
Étudiant au doctorat - PhD Student