From 0eb8b9538c68048bc24520719ea4fa69b1509030 Mon Sep 17 00:00:00 2001
From: kitty <nepeta@canaglie.net>
Date: Fri, 30 Jan 2026 21:25:04 +1100
Subject: more primitives, notably PARSE, untested

---
 jefs.s    | 69 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 readme.md |  5 ++++-
 2 files changed, 73 insertions(+), 1 deletion(-)

diff --git a/jefs.s b/jefs.s
index a53b8d4..c0b1b82 100644
--- a/jefs.s
+++ b/jefs.s
@@ -68,18 +68,87 @@ _start:
 	mov rax, __NR_exit
 	syscall
 
+defword "bye", bye, 0
+	mov rdi, 0
+	mov rax, __NR_exit
+	syscall
+	ret ; will not be reached
+
 defword "@", fetch, 0
 	pspop r11
 	mov r12, qword [r11]
 	pspush r12
 	ret
 
+defword "c@", cfetch, 0
+	pspop r11
+	xor r12, r12
+	mov r12b, [r11]
+	pspush r12
+	ret
+
 defword "!", store, 0
 	pspop r11
 	pspop r12
 	mov qword [r11], r12
 	ret
 
+defword "c!", cstore, 0
+	pspop r11
+	pspop r12
+	mov [r11], r12b
+	ret
+
+; stage 1 parser. very rudimentary, does not recognise numbers,
+; only recognises newlines and spaces as whitespace.
+; since it will only parse a little bit of the init file
+; there won't be any error checking either.
+defword "parse", parse, 0
+	mov r13, qword [to_in]
+	add r13, initfile
+	mov r12b, byte [r13]
+
+.wsloop: ; skip initial ws
+	cmp r12b, 0x20
+	je .wsloop_cont
+	cmp r12b, 0x0a
+	jne .wordloop_start
+.wsloop_cont:
+	inc r13
+	mov r12b, byte [r13]
+	jmp .wsloop
+
+.wordloop_start:
+	push r13 ; keep start address of this word for later
+	mov r11, 1 ; W: word length count
+.wordloop:
+	cmp r12b, 0x20
+	je .wordloop_end
+	cmp r12b, 0x0a
+	je .wordloop_end
+
+	inc r11
+	inc r13
+	mov r12b, byte [r13]
+	jmp .wordloop
+
+.wordloop_end:
+	sub r13, initfile
+	mov qword [to_in], r13
+	pop r13
+	pspush r13 ; c-addr
+	pspush r11 ; u
+	ret
+
+defword "find", find, 0
+	ret
+
+; stage 1 interpreter, just reads from initfile
+defword "interpret", interpret, 0
+	ret
+
+defvar ">in", to_in, 0, initfile
+
 initfile:
 incbin "jefs.fs"
 initlen equ $ - initfile
diff --git a/readme.md b/readme.md
index e01353b..97ed56a 100644
--- a/readme.md
+++ b/readme.md
@@ -28,7 +28,7 @@ there are a set of 'virtual registers' (see moving forth part 1)
 
 - W, the working register, is r11.
 - X, the secondary working register, is r12.
-- IP, the instruction pointer, is r13. (this will only be needed if we switch to DTC, otherwise this can be Y, a third general purpose register)
+- Y, the tertiary working register, is r13.
 - SP, the working stack pointer, is r14.
 - RSP, the return stack pointer, is rsp.
 
@@ -57,8 +57,11 @@ but obviously we need *something* to bootstrap from.
 - `:` (` -- `): compile a word from the current input source
 - `;` (` -- `): stop compiling the current word
 - `parse` (` -- c-addr u`): parse a word from the *forth init* (see footnote)
+- `find` (`c-addr u -- ?lfa flag`)
+- `interpret`: the stage 1 interpreter
 - `latest`: a variable, the latest defined word
 - `state`: a variable, determines current mode (interp./compiling)
+- `>in`: a variable, offset from start of forth init
 
 on `parse`: i don't want to deal with file access and stuff from the kernel.
 so this word will be Upgraded in the forth init to read from a file.
-- 
cgit v1.2.3