Introduction

This document is to describe the M code language and its functions; this document goes together with the MCode Language in explanation in which Andrés Moreno García documented for his thesis. Here are a few examples of the Mcode that is generated by the compiler for the use of the visualisation tool.

First of all we will look at some basic programming statements and see the output from them after that we aim to look at how we believe the compiler for the C version should output M Code and the new instructions that we believe are required for that to happen.

Note the separator for this in the MCode language is § this has been left out so that it is easier to read the statements, at the end of the document you will find all examples in there true form.

Details

1. int x;

26 x 0 ? int 1 6,9,6,17

26 is the number given to the instruction of VD, an index of which can be found at the end of this document.

VD (Variable Declaration) is described as

VD name NO_REFERENCE/ir value type FINAL/NOT_FINAL Loc

When declaring a variable the corresponding the MCode instruction needs to be omplemented with its name, value, type and the modifier (FINAL or NOT_FINAL). Instruction reference (ir) is given if the variable has an initializer otherwise a NO_REFERENCE value is written.

This can be seen form the values found up above.

26 is the reference for VD as found in the reference table.

X is the name of the variable

0 is the instruction reference, in this case 0 the first instruction

? is the value of the variable which if not know will be ?

int is the type of the variable others can be double, char etc

1 means that the value of the variable is not final and can be altered latter on in the program

6,9,6,17 is the location of the variable.

2. x = 3;

After setting up the variable we can begin to use the variable The MCode from the program 3 5 2 8,5,8,9 4 3 27 3 V ? int 0,0,0,0 28 4 3 Int 8,9,8,9 5 24 3 3 int 8,5,8,9

Translated

Begin A 2 8,5,8,9

To 3 QN 3 v ? int 0,0,0,0

L 4 3 int 8,9,8,9 A 2 4 3 3 int 8,5,8,9

The instructions are described as follows with the values shown

BEGIN ti ir loc

Instruction used to mark the beginning of those instructions that admit instructions to be encapsulated within them. Those instructions are Assignment, Return, Parameter, Array Access and all unary and binary operations; and they are referred through it. This is represented by the type of instruction (ti) The referred instructions get they ir assigned in the BEGIN instruction.

What we have here is an assignment operator classed as A

3 is the begin operator

5 is the A (Assignment operator) 2 is the instruction reference (ir) 6,9,6,17 is the location of the expression in the source code.

TO ir This instruction is used in assignments and reflects the movement of the value to the left hand side of the assignment. The ir points to the qualified name that will hold the value. As said before, this instruction is used in assignment and more concretely in Assignment, Variable Declaration (those with initializer) and Compound Assignment.

4 is the To operator 3 is the instruction reference from which it is related to.

QN ir name value type loc Qualified names are all the local variables. MCode instruction contains the reference (ir) of the instruction and the name value and type of the qualified name. If the variable is not initialized value will be UNKNOWN.

27 is Qualified Name (QN) 3 is the ir V is the variable name ? the value is unknown at the moment Int is the type 0,0,0,0 is the location

L ir value type loc

Literals are the constants values in the source code (e.g. 3 is one integer literal and “3” is one string literal). The MCode instruction contains all the information needed for the visualization as well as its reference (ir), value, type and loc.

28 is L (Literal) 4 is the ir 3 is the value Int is the type 8,9,8,9 is the location

A ir rir lir value type loc The assignment instruction is composed by its own reference (ir) and the references to the left and right hand sides (lir, rir). Furthermore it contains the assigned value and its type. It is worth to mention that compound assignments are decomposed into the operation and a simple assignment.

5 is A (from table) 2 is the ir 4 is the right instruction reference (rir) 3 is the left instruction referenct (lir) 3 is the value Int is the type 8,5,8,9 is the location

3.

int x = 0; int i = 0;

while (i <2){ x = i; i++; }

The declaration of the two variables is the same as they where shown up above in example 1 and 2, the steps are combined to assign 0 to each variable.

For the while statement, the following sample of the M Code is generated.

3 22 8 8,12,8,15 1 9 27 9 i 0 int 0,0,0,0 2 10 28 10 2 Int 8,15,8,15 22 8 9 10 TRUE boolean 8,12,8,15 40 8 TRUE 0 8,17,11,9 35 1 3 5 11 9,12,9,16 4 12 27 12 x 0 int 0,0,0,0 27 13 i 0 int 0,0,0,0 5 11 13 12 0 int 9,12,9,16 3 13 14 10,12,10,14 2 15 27 15 i 0 int 0,0,0,0 13 14 15 1 int 10,12,10,14 35 0

This then translate to the following using the table in the reference.

Begin LE 8 8,12,8,15 Left 9 QN 9 i 0 int 0,0,0,0 Right 10 L 10 2 int 8,15,8,15 LE 8 9 10 TRUE Boolean 8,12,8,15 WHI 8 TRUE 0 8,17,11,9 SCOPE 1 Begin A 11 9,12,9,16 To 12 QN 12 x 0 int 0,0,0,0 QN 13 i 0 int 0,0,0,0 A 11 13 12 0 int 9,12,9,16 Begin Pie 14 10,12,10,14 Right 15 QN 15 I 0 int 0,0,0,0 Pie 14 15 1 int 10,12,10,14 Scope 0

LEFT/RIGHT ir These instructions are similar to BEGIN. Both of them are used to mark the beginning of the left/right side of a binary operation. The ir obeys the same reason than in BEGIN, “reserves” the reference number for the following instruction.

1 9

1 is the number given to the instruction set Left 9 is the ir

binaryCode ir rir lir value type loc

Binary instructions are composed by its binaryCode, its own reference (ir) and the references to the left and right sides of the expression (lir, rir). Furthermore, it contains the assigned value and its type. The binaryCode can take any of the values shown in Table A.3

22 8 9 10 TRUE Boolean 8,12,8,15

22 is the Less than operator. 8 is the ir 9 is the lir 10 is the rir TRUE is the value of the operation Boolean is the type of answer the valus is 8,12,8,15 is the location WHI/FOR/DO condition value round loc These statements produce a similar to the previous one. They only differ in the round token. This token holds the number of iterations the loop has made.

40 8 TRUE 0 8,17,11,9

40 is the while number on the reference 8 is the condition TRUE is the value 0 is the round 8,17,11,9 is the location

SCOPE 1/0 New blocks of Java code are delimited in MCode through SCOPE. This instruction second token indicates whether it is opening a new one (1) or closing one (0).

35 1

35 is the Scope number

1 indicates that the scope is a new block of code.

unaryCode ir reference value type loc

As with binary operations the unary instructions (unaryCode) take the similar tokens. The only difference is that there is only one reference to another instruction. The value, type and loc maintain the same meaning. As before there are several unary operators that can be assigned to the unaryCode, all of them are listed in Table A.4.

13 14 15 1 int 10,12,10,14

13 is the PIE reference number 14 is the ir 15 is the reference 1 is the new value of the variable Int is the type of variable 10,12,10,14 is the location

These instructions continue until the condition is not meet i.e the LE comes back as False and the WHI instruction stops going into the loop. 4.

int x = 0;

int i = 5;

if (i < 3){

x = 10;

} else {

x = 2;

}

The beginning of this instruction is the same as the other 3 examples above.

The code above is very similar to that of the while statement in that the M code generated is the same except instead of the Whi command we get an IFTE command

The values of the M Code are displayed below

3 22 8 8,9,8,13 1 9 27 9 i 5 int 0,0,0,0 2 10 28 10 3 int 8,13,8,13 22 8 9 10 FALSE boolean 8,9,8,13 34 8 FALSE 11,10,13,5 35 1 3 5 11 12,9,12,13 4 12 27 12 x 0 int 0,0,0,0 28 13 2 int 12,13,12,13 5 11 13 12 2 int 12,9,12,13 35 0

BEGIN LE 8 8,9,8,13

LEFT 9 QN 9 i 5 int 0,0,0,0 RIGHT 10 L 10 3 int 8,13,8,13

LE 8 9 10 FALSE boolean 8,9,8,13 IFTE 8 FALSE 11,10,13,5 SCOPE 1 BEGIN A 11 12,9,12,13

TO 12 QN 12 x 0 int 0,0,0,0 L 13 2 int 12,13,12,13

A 11 13 12 2 int 12,9,12,13

The IFTE command is described as follows

IFT/IFTE condition value loc There are two possible instructions for an “If” statement. IFT is printed if there is not an else statement or IFTE if there is an else statement. The composition, however, is similar. The condition is the reference to the instruction that evaluate the condition. The value holds the result of the evaluated condition and will tell which branch execution is following. The loc as usual contains the code location.

If the value is true it will go into the first scope other wise it will go into the second scope. In this case the value returned is false and therefore the command will go into the second scope. In this section we will look at how the C language will change the different instructions in the MCode language by adding certain information into the instruction. These differences will be detailed below.

Please note that these examples where done out by hand and not through the use of a MCode compiler.

1.

int ptr; int x = 3; ptr = &x;

Since the MCode will be representing the C language, which includes pointers (a before the name of the variable and after the type of variable) and pointers to pointers ( before the name of the variable), it is necessary to include the address of the variable (which is assigned by the compiler) at the end of the Mcode statement when the variable is being declared.

The new variable declaration will be

VD name NO_REFERENCE/ir value type FINAL/NOT_FINAL/GLOBAL/NOT_GLOBAL/STATIC/NOT_STATIC Loc Address

When declaring a variable the corresponding the MCode instruction needs to be omplemented with its name, value, type and the modifier. Instruction reference (ir) is given if the variable has an initializer otherwise a NO_REFERENCE value is written. The modifier can be Static, Not static, global, not global, Final and Not Final.

The code above will be transferred into the following MCode.

Any new or different instructions will be described below the MCode.

VD ptr 0 ? int 0 location address VD x 1 3 int 0 location address Begin A 2 location

To 3 QN 3 ptr address?? int location address Begin AO 4 location

Right 4 QN 5 x 3 int location address

AO 3 5 address int location

A 2 3 4 value(address) int location

int is a new type of variable for the MCode, it represents the pointer variable of type int. Another pointer variable is int, this represents a pointer to a pointer.

QN ir name value type loc address

Qualified names are all the local variables. MCode instruction contains the reference (ir) of the instruction and the name value and type of the qualified name. If the variable is not initialized value will be UNKNOWN. The address of the variable is also included in this for the use of the C language.

Address Of (AO)

AO ir reference value type location

Address of is an instruction which will allow the address of a variable to be placed into a pointer variable. The MCode instruction contains the ir the reference the value, type and location.

2.

ptr = 10

Begin A 6 location

To 7 Begin DR 7 location

QN 8 ptr value int location address

DR 7 8 address 3 int location Literal 9 10 int location

A 6 9 7 10 int location

DR

De-Reference ir reference address value type location

The de-reference instruction will allow a pointer variable to be de-referenced i.e. the variable will go back along it’s path to the previous variable and find outs its value. The DR consists of the ir the reference the address, the value the type and the location.

3.

X = (ptr) +5

Begin A 10 location

To 11 QN 11 x 10 int location address Begin AE 12 location

L 13 Begin DR 13 location

Right 14 QN 14 ptr 0x0013(address) int location address

DR 13 14 0x0013(address) 10 int location R 15 L 15 5 int location

AE 12 13 14 14 int location

A 10 13 11 15 int location

4.

Ptr2 = &ptr

VD ptr2 1 ? int 0 location address Begin A 2 location

To 3 QN 3 ptr2 2 int location address Begin AO 4 location

Right 5 QN 5 ptr address int location address

AO 4 5 address int location

A 2 4 3 address int location

5.

Function(20)

SMC function ? 1 location modifier Begin P 1 location

Literal 1 20 int location

P 1 20 int address location Parameters i MD Location

SMC name declaringClassName numArgs loc The SMC instruction consists of the actual name of the method, its declaring class name and the number of arguments this method call consists of. It also consists of the modifier, whether it is global, final or static.

SMCC

This instruction indicates the end of a call to a static method.

PARAMETERS parameterArray The PARAMETERS instruction will provide the visualization engine a list of the types of the parameters used in the method declaration.

P ir value type address loc This instruction declares a new parameters in a method call. It requires a ir the value of the parameters the type the address and its location in the program.

Method Declaration

MD ir loc A method declaration instruction will indicate the location of the called method code and the beginning of its evaluation. 6.

Return (33);

Begin return 0 location

Literal 2 33 int location

Return 0(stack pop) 2 33 int location SMCC

Return R callIr loc

This first return instruction is just used for the "void" return, those that does not return anything. The callIr is the reference for the return value of the method and thus connect the return command to the specific method.

R callIr valueIr value type loc

This second instruction is used when a value is returned. Thus, we need a reference to the evaluation of the expression corresponding to the returned value (valueIr) and the result of this expression, its value and type.

In the instance above

0 is callIr 2 is valueIr 33 is the value int is the type And location is the loc

7.

Funtion (int i)

SMC function ? 1 location modifier Begin P 1 location

QN 1 ptr 0x1234 int location 0x0100 P 1 0x1234 int 0x0200

Parameters i MD location

8.

Function (&x)

SMC function ? 1 location modifier Begin P 1 location Begin AO 1 location

Right 2 QN 2 x ?? location address

AO 1 2 0x1234 int 0x0300

Parameters 0x0300 MD location 9.

Struct MyStruct {

Char c Int i;

} Typedef struct mystruct MY_NEW_TYPE; Struct mystruct S1; MY_NEW_TYPE s2 ps; Ps = &s1;

S1.c = ‘a’;

Class mystruct (modifier, FGS) Field c char Field i int ENDCLASS

VD S1 1 ? mystruct (modifier)0 location address (0x5678) Begin A 2 location

To QN 4 S1 ? mystruct location address OFA 3 4 c ? char modifier location address LIT 5 ‘a’ char location

A 2 5 3 ‘a’ char location

Object Field Access (OFA) OFA ir objectReference name value type modifier loc address

This instruction accesses to the value of an object field, thus it references to the instruction that evaluate the object (objectReference) (normally a Qualified Name. It contains the name, the type and the value of the accessed field, also the address of the object for use with C.

3 = ir 4 object Reference c name ? value char type modifier is the modifier location is the loc address is the address 10.

Struct MyStruct {

Char c Int i;

}

Typedef struct mystruct MY_NEW_TYPE;

Struct mystruct S1; MY_NEW_TYPE S2;

S1.c = ‘a’; S1.i = 47;

S2 = S1;

Class mystruct (moderator, FGS) Field c char Field i int ENDCLASS

VD S1 1 ? mystruct (modifier)0 location address (0x5678) VD S2 2 ? mystruct (modifier)0 location address (0x5678) Begin A 3 location

To QN 4 S1 ? mystruct location address OFA 5 4 c ? char modifier location address LIT 6 ‘a’ char location

Assign 2 7 3 ‘a’ char location Begin A 8 location

To 9 QN 9 S1 ? mystruct location address OFA 10 9 i ? int modifier location address LIT 11 47 int location

A 2 11 8 47 int location Begin A 12 location

To 13 QN 13 S2 ? mystruct location address QN 14 S1 ? mystruct location address

A 3 14 13 12 mystruct location

Int x = 15; //global variable

VD x 1 3 int 2 (modifier) location address

Reference

Value Name 1 LEFT 2 RIGHT 3 BEGIN 4 TO 5 A 6 AE 7 SE 8 ME 9 DE 10 RE 11 PLUS 12 MINUS 13 PIE 14 PRIE 15 PDE 16 PRDE 17 AND 18 OR 19 EE 20 NE 21 GT 22 LE 23 LQE 24 GQT 25 NO 26 VD 27 QN 28 L 29 SMC 30 P 31 PARAMETERS 32 R 33 IFT 34 IFTE 35 SCOPE 36 END 37 MD 38 SMCC 39 BR 40 WHI 41 FOR 42 CONT 43 DO 44 OUTPUT 45 COMP 46 BITOR 47 BITXOR 48 BITAND 49 LSHIFT 50 RSHIFT 51 URSHIFT 52 XOR 53 ERROR 54 INPUT 55 INPUTTED 56 AA 57 AAC 58 CLASS 59 END_CLASS 60 CONSTRUCTOR 61 METHOD 62 FIELD 63 SA 64 SAC 65 OFA 66 OMC 67 OMCC 68 SWIBF 69 SWITCHB 70 SWITCH 71 AL 72 CONSCN 73 CAST 74 SFA 75 AIBEGIN 76 AIE 77 AI