What function does an interpreter perform with the instructions in a high level programming language?

Answer 1

All programming languages need to be converted to machine code. Machine code is the native language of the machine; it is the only language understood by the machine. When we write programs in other languages they must be translated into machine code. To achieve that, we need to write a program to perform the translation for us.

For this we can either use an assembler, a compiler or an interpreter, all of which are machine code programs. Assemblers can only be used with assembly languages, which are low-level symbolic languages that translate near 1:1 with the machine code. The machine code can be saved and executed without further translation. However, assembly languages are not portable; each machine architecture has its own specific version of machine code and therefore requires its own specific assemblers to produce it. To port the code to another architecture you essentially have to re-write the entire source code. There are tools (pre-processors) that can help make the task easier but it's by no means a straightforward process.

[The reverse of assembling is disassembling, which converts machine code into a low-level language that is similar to assembly but is known as disassembly. However, since an assembler strips out all variable names and user comments, a disassembler cannot restore the original assembly language code. Indeed it is impossible to restore the original source code no matter what language was used. Disassembly is as good as it gets.]

All high-level languages are either compiled or interpreted or are both compiled and interpreted.

Compiled languages compile the entire source code into machine code, much like assembly language but with a higher level of abstraction between the source and the machine code. As with assembly languages, the machine code can be saved and executed without any further translation. Being more abstract, compiled languages are much easier to port to other architectures than assembly languages. However they must still be compiled separately upon each architecture. The most popular examples of compiled languages are C and C++.

Interpreted languages do not translate the entire source code, they step through the source code one statement at a time, converting each statement to machine code and executing it before moving onto the next statement. This has the advantage over compiled languages in that you do not have to wait for the entire program to compile before executing the program. A compiled language program can take several minutes to compile, whereas an interpreted language can be executed immediately. However, the need to interpret every statement before executing it incurs a performance penalty. With compiled languages, you only pay the cost during compilation; once compiled, you do not need the compiler. Interpreted languages cannot execute without the interpreter, but any machine that has a suitable interpreter can execute the same source code regardless of which machine architecture it was written on. The most popular interpreted language is BASIC since it is the language most programmers learn before moving onto other languages.

Languages that are both compiled and interpreted use a compiler to produce an intermediate source known as byte code. Byte code is not machine code, it still has to be interpreted, however the byte code "statements" are much simpler to interpret than an interpreted language's statements thus it performs much better. Also, unlike compiled languages, the compiled byte code is highly portable. The best example of such a language is Java which compiles to Java byte code suitable for interpretation by the Java virtual machine (JVM). Given that virtually all platforms provide a JVM implementation, it's no surprise that Java is by far the most popular application programming language in use today. However, it's lack of low-level features limits its usefulness in non-application fields such as driver software, operating system kernels and embedded systems software. Its performance and memory consumption also leaves a lot to be desired, however it's ease-of-use makes it a good stepping stone to learning C++.