x = 5 * 3;
   y = x + 2;

   Program    ::= { Definition }
   Definition ::= { IDENT "=" Expression ";" }
   Expression ::= Term { ("+"|"-") Term }
   Term       ::= Factor { ("*"|"/") Factor }
   Factor     ::= IDENT | INTLIT | "(" Expression ")"

The abstract syntax is

   Program    = PROGRAM { Definition }
   Definition = DEF String Expression
   Expression = IDENT String
              | INTLIT int
              | BINOP Expression Expression char

Your job is to complete the pretty-printer visitor in Printer.java. If you have extra time you can also work on the interpreter visitor, but this will only work with identifiers after name analysis.

• Get the file exercice2.tar (Alternatively, if you have done the previous exercise, you can continue on that code).
• Unpack it with the command tar xof exercice2.tar (This will create a new directory exercice2)
• Go into the directory exercice2
• Run make to compile the given sources
• Check the given pretty-printer with make printerTest < examples/example1.expr (This will just print some empty lines)
• Now complete the source of the pretty-printer in source/expression/Printer.java by writing the methods caseDef(..), caseIdent(..), and caseBinOp(..).
• The tree data structure is given in source/expression/Tree.java
• You can look at the example in caseProgram(..)
• If you use your own parser solution, be aware that strange results or NullPointerExceptions may happen because of errors in the tree generation part of the parser.
• Recompile your pretty printer with make and check it with make printerTest < examples/example2.expr
• Now you can continue on the interpreter-visitor in source/expression/Interpreter.java. Here you can fix caseBinOp(..) to return the proper result. You can again look at the other cases for examples.
• Check the interpreter with make interpreterTest < examples/example1.expr
• We cannot yet do the caseIdent(..), because we have not yet done name analysis, so the second example will not yet work.