Q1:
1. struct declaration: 
		struct name becomes class name. 
		each struct field declaration follows the rules for variable declaration given below.
2. declaration of a variable of struct type (say, something like struct A x):
		- Translate it to "A x = new A()".
		- For each struct field fi of A (say of type struct Ti) --> 
				- x.fi = new Ti(); // use the rule recursively.
		- for each int / pointer to struct field gi -- do nothing.

3. Declaring a variable of struct * type: struct A * x --> A x.

4. passing a struct variable (say x, of type struct T) as an argument:
		- Copy x to a new variable (say y, of type class T).
			- y = new T();
			- y.gi = x.gi // for each non struct field gi.
			- for each struct field fi: recursively call this
			  routine.
		- Pass y to the function.
5. Passing of a variable of struct * type: just call the function with the variable name.

6. Dereferencing struct fields: x.fi --> x.fi (remains unchanged)

7. Declaring a variable of int type: --> remains unchanged.

7. sizeof(x) --> 
	If type of x is int/struct * --> 4 bytes (assuming sizeof(int) and sizeof(struct *X) is 4 bytes.

	Else --> Compute the sizeof the struct and emit the exact number.


Q2:
	Lattice: \bot -> {Boolean, Int} -> \top
	Initialize: for all variables F(x) = \top
	Transfer functions:
		- variable declaration T1 x: F(x) =  T1 \meet F(x)
		- copy: x = y // F(x) = F(x) \meet F(y)
		- copy: x = c // F(x) = F(x) \meet typeOf(c) 
		- if-stmt and loop-stmts mainly introduce a join stmt. 
	A use of a variable x is undefined - if F(x) = \bot.

	Initialize: for all variables F(x) = \top
	Transfer functions:
		- variable declaration T1 x: F(x) =  T1 \meet F(x)
		- copy: x = y // If F(x) = \top F(x) = F(y); if F(x) != \bot if F(x) != F(y): Type mismatch 
		- copy: x = c // If F(x) = \top F(x) = typeOf(c); if F(x) != \bot if F(x) != F(y): Type mismatch 
		- if-stmt and loop-stmts mainly introduce a join stmt. 
	A use of a variable x is undefined - if F(x) = \bot.