Creating a language interpreter

Question

I am trying to understand how a language interpreter works. Can you guys point me the general lines on how an interpreter works?

I mean, suppose I have some lines written like this

10  x = 200;
20  for r = x to 1000 step 1
25  z = r + 32;
30  print z;
40  next r;
50  end;

what's the best way to build an interpreter that could run something like that?

Having a large matrix containing all functions allowed and searching for a match? The first line, for example: it is assigning 200 to a variable x, but these are symbols that does not exist.

If you guys can give me the direction...

Thanks for any help.

Answer 1

Perhaps you are talking about creating a DSL. You might find this helpful (if you are ok with spending $$)

http://gilesbowkett.blogspot.com/2010/03/create-your-own-programming-language.html

Answer 2

Compiler creation is a complex topic (an interpreter can be seen as a special compiler).

You have to first parse it - try to understand the syntax and then create some internal representation (abstract syntax tree) and then create some execution logic.

Wikpedia suggests http://mcs.une.edu.au/~comp319/

Answer 3

Learn about tools such as lex/flex http://flex.sourceforge.net/ and yacc/bison http://www.gnu.org/software/bison/. These are most popular tools for building compilers in open software world. Many well known open source programs are written using them (including PHP, gcc, doxygen). You'll find a lot of free books and tutorials. They not only show how to use lex and yacc tools, but also explain general ideas behind compilers.

Answer 4

I'm interested in learning more about this as well. I found Douglas Crockford's JavaScript parser interesting, though from what I understand he's using a different method than is typical for parsing languages. It's not the full picture for interpreting and compiling, but I found it helpful to see some actual parsing implementation and the resulting restructuring of the code.