PSEUDO ASSEMBLER

This is the program that will copmlile Player's code and translate it from pseudo-assembly language into "byte-code". Bytecode is a machine code, which will be directly interpreted by the virtual machine.

The program will get a file written in pseudo-assembly language as an argument and generate a file that will be understood by VM.

It will read the code processed from the file .s given as argument, and write the resulting bytecode in a file named same as argument by replacing the extension .s by .cor.

WHAT IS PSEUDO-ASSEMBLY LANGUAGE?

Code structure:

1. it is composed of one instruction per line;
2. instruction is usually composed of three elements: a label (optional) composed with a chain of characters followed be ":", an opcode (certain operation) and its parameters;
3. comments are allowed but ignored ( character '#' starts a comment ).

Parameters can be of three different types:

1. registry (REG) : r1 <-> r16 ;
2. direct (DIR) : the character '%' followed by a nymeric value ir a label which represents a direct value;
3. indirect (IND) : a value or a label, which represents a value located at the address of the parameter, relative to the Program Counter of the current process.

Name and description of a Player

A Player always has a name and a description at the beggining. Player's name must be on the first line (otherwise it'll be invalid) following the marker ".name" . Requirements for a description are almost the same besides the fact that it must be on the second line following the marker ".comment". All other markers will be treated as invalid and may occure an error.

Operations

NAME	INDEX	FIRST ARG	SECOND ARG	THIRD ARG
live	1	DIR	-	-
ld	2	DIR or IND	REG	-
st	3	REG	IND or REG	-
add	4	REG	REG	REG
sub	5	REG	REG	REG
and	6	REG or DIR or IND	REG or DIR or IND	REG
or	7	REG or DIR or IND	REG or DIR or IND	REG
xor	8	REG or DIR or IND	REG or DIR or IND	REG
zjmp	9	DIR	-	-
ldi	10	REG or DIR or IND	REG or DIR	REG
sti	11	REG	REG or DIR or IND	DIR or REG
fork	12	DIR	-	-
lld	13	DIR or IND	REG
lldi	14	REG or DIR or IND	REG or DIR	REG
lfork	15	DIR	-	-
aff	16	REG	-	-

Encoding of arguments

It can be seen that there are operations that can use different types of arguments on different places of the operation call. So there's a problem how to differ and store what types of arguments were given. The thing is that arguments' types are encoded with one byte:

1. We devide byte into 4 pairs of bits: each pair corresponds to a particular argument.
2. Maximum number of arguments is 3, but we have 4 pairs of bits, that's why the last pair will always be filled with zeros.

TYPE OF ARGUMENT	PAIR OF BITS
REGISTRY	01
INDIRECT	10
DIRECT	11

4. As a result we have a byte that encodes all types of arguments given to a particular option. Convert this result from binary to hexadecimal.