There is dependency analysis: Strender etc.
Bitwiser
There is tool analysis.
Bitwiser applied to the file, or alternatively, code coverage tools, which are an easier way of achieving almost the same thing.
- Java
- http://cobertura.github.io/cobertura/
- C/C++
- http://gcc.gnu.org/onlinedocs/gcc/Gcov.html
- http://covtool.sourceforge.net/
Image Testsuite is a good example of the way this can work out for corpus generation https://code.google.com/p/imagetestsuite/wiki/AboutTestSuite
On Tools