[Soot-list] Confused by SOOT tags: in source file or in .jimple file? And how to remove them?

Z zell08v at orange.fr
Wed Jun 20 05:37:37 EDT 2012


Sorry for that obscure quesiton !

I would like to get rid of tags that appeared in Eclipse once the
information is obtained. I meant,  after analyzing the code with, say,  a
parity analysis,  the source code are colored, on the left side of Eclipse
windows, we have a "SA" labels, and a bullet pops up if we put the mouse
over a colored variable. Then, we may want to get ride of those tags, to
clean up a bit.  But I find they are always there, even when I recompiled
the program later.

%30 is an approximate rate of the  tags appearing in the source code
(displayed in Eclipse). It must be a wrong operation from my part for such
a deterministic situation. The issue is, the problems of tags I mentioned
seemed to be hyper-sensitive to som operation that I am not aware of.

Otherwise, I would like know how that

1.  Are those tags supposed to appear in source or in jimple code ? It is
clear that, in my machine the tags of the built-in nullness analysis
appeared in source code, and the tags of the built-in liveness analysis
appeared in jimple but not in  source code.

2.  Could you explain me the differences between some very basic terms:
tag/attribute/annotation ?  These seem to be too evident that I cannot find
their exact meaning in Google or the most helpful survivor´s guide. (I
suppose, an annotation is a Java term I suppose, a tag is something
attached to "hosts" to give a visual effect, an attribute is only a SOOT
option "-xml-attribute", is that right)

3. I also would like to know on which pack should I inject my third-part
liveness analysis, jtp or jap, Or whatever ?

Excuse me for those superficial questions. I hope I am clear now. Thank
you.

Zell.

On Wed, Jun 20, 2012 at 7:35 AM, Eric Bodden <eric.bodden at ec-spride.de>wrote:

> Zell, could you elaborate a bit?
>
> What do you mean with 30%? 30% of all runs? Or the same 30% of all
> tags in each run? Also what do you mean by getting rid of tags? Why
> would one want to do that, and at which stage of processing the code?
> Do you mean "not showing the tags in Eclipse"?
>
> Eric
>
> On 20 June 2012 06:45, Z <zell08v at orange.fr> wrote:
> > Hello,
> >
> > I feel quite confused about the SOOT tags. Sometimes they are given in
> the
> > source code (in the case of built-in nullness colorer launched from
> eclipse)
> > , sometimes they are shown in .jimple code (in the case of built-in
> liveness
> > analysis lauched from eclipse). Worse, for an user-defined analysis like
> a
> > third-part liveness analysis, I got those tags appearing in my source
> with a
> > probability of 30% using the following soot-options.
> >
> >     Args="-xml-attributes -f j -p jtp.myLiveness on tester.Snippet"
> >
> > I certainly get something wrong here. Could you help?
> >
> >  And my another question is, how to get rid of those tags once
> generated?? I
> > generated some tags from nullness analysis, and they are always there
> with
> > the "SA" labels aside.
> >
> > Thanks for your help. Great tool!
> >
> > Zell.
> >
> >
> >
> > _______________________________________________
> > Soot-list mailing list
> > Soot-list at sable.mcgill.ca
> > http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
> >
>
>
>
> --
> Eric Bodden, Ph.D., http://bodden.de/
> Head of Secure Software Engineering Group at EC SPRIDE
> Principal Investigator in Secure Services at CASED
> Tel: +49 6151 16-75422    Fax: +49 6151 16-72051
> Room 3.2.14, Mornewegstr. 30, 64293 Darmstadt
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.mcgill.ca/pipermail/soot-list/attachments/20120620/3088a6d4/attachment.html 


More information about the Soot-list mailing list