[Soot-list] Theoretically sound call-graph construction with Soot

[Hossein Mehrfard]

Hello,

I am wondering about the possibility of building theoretically sound
inter-procedural call graph at application level with Soot. In other words,
I only wants those sequences of method call that I am absolutely certain,
in case of proper test case, they will happen at run time. Therefore, the
precision (i.e., completeness) of a call graph is not important for me.

I know SPARK provides a few algorithms for call graph construction (e.g.,
1-CFA or OnFlyCallGraphBuilder), but is there any way to distinguish
between approximated method calls and deterministic method calls (may
versus must)?

Thanks,
Hossein
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailman.CS.McGill.CA/pipermail/soot-list/attachments/20170210/c5566c56/attachment.html