From 63e3a1ea7eecef3ea6a212f3855cb26ba33fc621 Mon Sep 17 00:00:00 2001
From: andromeda <andromeda@lenovo>
Date: Sat, 7 Mar 2026 21:33:28 +0100
Subject: [PATCH] add some tokens, length1 token support

---
 twasm/README.md    | 23 +++++++++++++
 twasm/asm/main.asm | 86 ++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 109 insertions(+)

diff --git a/twasm/README.md b/twasm/README.md
index 4607b01..91a7eb3 100644
--- a/twasm/README.md
+++ b/twasm/README.md
@@ -129,4 +129,27 @@ supported tokens are listed below
 | cr3   | 0x004C | |
 | cr4   | 0x004D | |
 | cr8   | 0x004E | |
+| hlt   | 0x004F | |
+| int3  | 0x0050 | |
+| [     | 0x0051 | |
+| ]     | 0x0052 | |
+| xor   | 0x0053 | |
+| inc   | 0x0054 | |
+| dec   | 0x0055 | |
+| mov   | 0x0056 | |
+| add   | 0x0057 | |
+| sub   | 0x0058 | |
+| call  | 0x0059 | |
+| ret   | 0x005A | |
+| cmp   | 0x005B | |
+| je    | 0x005C | |
+| jne   | 0x005D | |
+| jge   | 0x005E | |
+| jg    | 0x005F | |
+| jle   | 0x0060 | |
+| jl    | 0x0061 | |
+| +     | 0x0062 | |
+| -     | 0x0063 | |
+| *     | 0x0064 | |
+| /     | 0x0065 | |
 |       | 0xFFFF | unrecognised token |
diff --git a/twasm/asm/main.asm b/twasm/asm/main.asm
index e498f91..a3dea27 100644
--- a/twasm/asm/main.asm
+++ b/twasm/asm/main.asm
@@ -47,6 +47,9 @@ start:
 ; ------------------------------------------------------------------------------
 
 identify_token:
+  cmp rsi, 1        ; if the token has length 1
+  je .start_length1 ; then enter the length 1 loop
+
   cmp rsi, 2        ; if the token has length 2
   je .start_length2 ; then enter the length 2 loop
 
@@ -58,6 +61,26 @@ identify_token:
 
   jmp .unrecognised ; else unrecognised
 
+  .start_length1
+    mov rcx, tokens.length1 ; rcx -> list of known tokens
+
+  .loop_length1
+    cmp rcx, tokens.length2 ; check if rcx still in the bounds of length1 tokens
+    jge .unrecognised       ; if not, unrecognised
+
+    mov r10b, [rcx] ; known token
+    mov r11b, [rdi] ; token
+    cmp r10b, r11b    ; if known token matches token
+    je .found_length1 ; exit loop
+
+    add rcx, 3 ; length of token + length of id
+    jmp .loop_length1
+
+  .found_length1
+    xor eax, eax      ; make sure rest of rax is zeroed
+    mov ax, [rcx + 1] ; return id of token
+    ret
+
   .start_length2
     mov rcx, tokens.length2 ; rcx -> list of known tokens
 
@@ -496,6 +519,22 @@ test_identify_token:
   mov rsi, .msg
   call print
 
+  ; length1 token that exists
+  mov byte [TEST_ARENA_ADDR], "*"
+  mov rdi, TEST_ARENA_ADDR
+  mov rsi, 1
+  call identify_token
+  cmp ax, 0x0064
+  jne .fail
+
+  ; length1 token that doesn't exist
+  mov byte [TEST_ARENA_ADDR], " "
+  mov rdi, TEST_ARENA_ADDR
+  mov rsi, 1
+  call identify_token
+  cmp ax, 0xFFFF
+  jne .fail
+
   ; length2 token that exists
   mov word [TEST_ARENA_ADDR], "sp"
   mov rdi, TEST_ARENA_ADDR
@@ -575,6 +614,19 @@ test_identify_token:
 ; ------------------------------------------------------------------------------
 
 tokens:
+  .length1
+    db "["
+    dw 0x0051
+    db "]"
+    dw 0x0052
+    db "+"
+    dw 0x0062
+    db "-"
+    dw 0x0063
+    db "*"
+    dw 0x0064
+    db "/"
+    dw 0x0065
   .length2
     db "r8"
     dw 0x0008
@@ -624,6 +676,12 @@ tokens:
     dw 0x0048
     db "ss"
     dw 0x0049
+    db "je"
+    dw 0x005C
+    db "jg"
+    dw 0x005F
+    db "jl"
+    dw 0x0061
   .length3
     db "rax"
     dw 0x0000
@@ -699,6 +757,30 @@ tokens:
     dw 0x004D
     db "cr8"
     dw 0x004E
+    db "hlt"
+    dw 0x004F
+    db "xor"
+    dw 0x0053
+    db "inc"
+    dw 0x0054
+    db "dec"
+    dw 0x0055
+    db "mov"
+    dw 0x0056
+    db "add"
+    dw 0x0057
+    db "sub"
+    dw 0x0058
+    db "ret"
+    dw 0x005A
+    db "cmp"
+    dw 0x005B
+    db "jne"
+    dw 0x005D
+    db "jge"
+    dw 0x005E
+    db "jle"
+    dw 0x0060
   .length4
     db "r10d"
     dw 0x001A
@@ -736,6 +818,10 @@ tokens:
     dw 0x003E
     db "r15b"
     dw 0x003F
+    db "int3"
+    dw 0x0050
+    db "call"
+    dw 0x0059
   .length5
   .end