Question

Is it possible to look at an object code and tell which language has been used originally to produce it? or does the language leaves a trace or a stamp on the object code ? do the compilers of various languages use a fixed format for a given ISA to develop the object code?

Answer 1

There is no general algorithm, but in practice it is often possible. Usually you can just look at the libraries that the application depends on – if a Windows application depends on msvcrt.dll, for example, then there’s a high chance that it’s a C or C++ program compiled with Visual C++. Sometimes a compiler leaves traces of evidence in the .data section. Here is what I see when opening a “Hello, World!”-like Haskell binary (compiled with GHC) in a hex editor:

Here’s what GCC’s “copyright notice” looks like:

A trained eye can even recognize compiler version by looking at disassembly (every compiler optimizes code slightly differently and has its own implementation quirks). If you need to automate this, I suggest looking at machine learning techniques.