[Soot-list] Extending Soot

[Eric Bodden]

Marcos, most of the generic data-flow components are very generic. For
instance, the FlowAnalysis interface just acts on a graph of nodes,
where nodes are *usually* Jimple Units but in general they could be
anything.

Also, Unit is a generic interface:
http://www.sable.mcgill.ca/soot/doc/soot/Unit.html
Any statement of your target language could be a unit.

All actual built-in analyses, however, are specific to a single IR.
Most are specific to Jimple. So my guess would be that if you just
want to program and use your own analyses on the new IR then this
should well be possible. However, if your goal is to reuse many of the
existing Soot-internal analyses then you would have to stick to
Jimple.

Eric

--
Eric Bodden
Software Technology Group, Technische Universität Darmstadt, Germany
Tel: +49 6151 16-5478    Fax: +49 6151 16-5410
Mailing Address: S2|02 A209, Hochschulstraße 10, 64289 Darmstadt



2009/12/4 Marcos Chicote <totochicote at gmail.com>:
> Hi!
> I'm considering Soot to develop some dataflow analysis in a language other
> than Java.
> My idea is to parse this language X and generate a intermediate
> representation of the language like Jimple and then use Soot's dataflow
> engine for the analysis.
>
> Is this possible? Has anyone done it? Is there documentation on how to do
> this?
>
> Thanks!
>
> _______________________________________________
> Soot-list mailing list
> Soot-list at sable.mcgill.ca
> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>
>