Massive refactoring to simplify this whole setup. The Token List will be modified in place and reused in the Parser to normalize the list and generate a symbols table. Normalizing the token list will make sure the syntax is valid and the symbols table will have the relative offsets and length of symbol values. At least this is all the plan but one step at a time.

misc/test.asm

@@ -1,4 +1,4 @@
-.include "/another_test.asm"
+.include "./another_test.asm"
 
 .db video_start 0xF37F
 .db msg "Hello, world!", 0
@@ -13,12 +13,13 @@ load r1, msg ; Because the lod* instructions can't load an address
 ;Out: Length in R2
 
 string_length:
-loadb r3, [r1] ; Load the byte from the address in R1, into R3
+load byte r3, [r1] ; Load the byte from the address in R1, into R3
 load r2, 0 ; String length
 cmp r3, null_byte ; Is R3 a null byte?
 je end
 inc r2
-
+start:
+    inc r1
 start:
     inc r1 ; next char
     loadb r3, [r1] ; Load the next character byte into R3