Scenario is:Though the source-code doesn't change, I'm seeing that the the compiled lib/bin files are having difference though It's compiled by same compiler and with same dependencies.
Since it's not a text file, I'm deprived of ideas, as we cannot use levenshtein distance, or pattern matching.
Another idea would be if, I'm able to add some salt to the source files, so that It is detectable in the compiled binaries?
Any ideas would be great, Since I have a huge no.of libs, I'm afraid it would be practically impossible for a Machine learning to be put in place.