This assembler interpreter supports a number of basic instructions. They are common to most of the different assembler dialects. It also supports labels, user comments and creating an output message. The entire list of supported instructions is shown below:
mov x, y- copy y (either an integer or the value of a register) into register x.inc x- increase the content of register x by one.dec x- decrease the content of register x by one.add x, y- add the content of the register x with y (either an integer or the value of a register) and stores the result in x (i.e.register[x] += y).sub x, y- subtract y (either an integer or the value of a register) from the register x and stores the result in x (i.e.register[x] -= y).mul x, y- same with multiply (i.e.register[x] *= y).div x, y- same with integer division (i.e.register[x] /= y).label:- define a label position (label = identifier + ":", an identifier being a string that does not match any other command). Jump commands and call are aimed to these labels positions in the program.jmp lbl- jumps to the labellbl.cmp x, y- compares x (either an integer or the value of a register) and y (either an integer or the value of a register). The result is used in the conditional jumps (jne,je,jge,jg,jleandjl)jne lbl- jump to the labellblif the values of the previouscmpcommand were not equal.je lbl- jump to the labellblif the values of the previouscmpcommand were equal.jge lbl- jump to the labellblif x was greater or equal than y in the previouscmpcommand.jg lbl- jump to the labellblif x was greater than y in the previouscmpcommand.jle lbl- jump to the labellblif x was less or equal than y in the previouscmpcommand.jl lbl- jump to the labellblif x was less than y in the previouscmpcommand.call lbl- call to the subroutine identified bylbl. When aretis found in a subroutine, the instruction pointer should return to the instruction next to thiscallcommand.ret- when aretis found in a subroutine, the instruction pointer should return to the instruction that called the current function.msg 'Register: ', x- this instruction stores the output of the program. It may contain text strings (delimited by single quotes) and registers. The number of arguments isn't limited and will vary, depending on the program.end- this instruction indicates that the program ends correctly, so the stored output is returned (if the program terminates without this instruction it should return the default output: see below).; comment- comments should not be taken in consideration during the execution of the program.
The output format is a string. But if the program doesn't reach the end instruction, the number -1 will be returned.
asmint --helpusage: asmint [-h] [-l] [-p] [-r] path
Assembler interpreter
positional arguments:
path the path to the program to be executed
optional arguments:
-h, --help show this help message and exit
-l, --labels show labels
-p, --program show prepared program
-r, --registers show register values
There is a directory examples containing a bunch of example programs implemented using this assembly language.
You can add your own interesting use case by creating a pull request on GitHub. All PRs are welcome :)
An example program named factorial.txt:
mov a, 5
mov b, a
mov c, a
call proc_fact
call print
end
proc_fact:
dec b
mul c, b
cmp b, 1
jne proc_fact
ret
print:
msg a, '! = ', c ; output text
ret
Calling the utility:
asmint factorial.txtProgram output:
Output: 5! = 120
Another example of use - power.txt:
mov a, 2 ; value1
mov b, 10 ; value2
mov c, a ; temp1
mov d, b ; temp2
call proc_func
call print
add a, -1
end
proc_func:
cmp d, 1
je continue
mul c, a
dec d
call proc_func
continue:
ret
print:
msg a, '^', b, ' = ', c
ret
Output:
Output: 2^10 = 1024
This interpreter can also handle multiple tabs and spaces in program code and ignore all comment text. Let's use a "mangled" version of the power.txt program. Let's call it power_mangled.txt. Here is the full code:
mov a, 2 ; value1
mov b, 10 ; value2
mov c, a ; temp1
mov d, b ; temp2
call proc_func
call print
add a, -1
end
; comment1
; comment2
; comment3
; yet another comment
proc_func:
cmp d, 1
je continue
mul c, a
dec d ; here is a comment message
call proc_func
continue:
ret
print:
; comment1
;comment2
;comment3
; yet another comment
msg a, '^', b, ' = ', c
ret
As you can see, the result is the same as in the previous example:
Output: 2^10 = 1024
This project was inspired by this CodeWars kata