;======================================================================================
;Defines
;======================================================================================
!R0			= $3000		;default source/destination register
!R1			= $3002		;PLOT X position
!R2			= $3004		;PLOT Y position
!R3			= $3006
!R4			= $3008
!R5			= $300A		;
!R6			= $300C		;multiplier for FMULT and LMULT
!R7			= $300E		;used in MERGE
!R8			= $3010		;used in MERGE
!R9			= $3012		;
!R10			= $3014		;
!R11			= $3016		;Return adress set by LINK
!R12			= $3018		;loop counter
!R13			= $301A		;loop adress
!R14			= $301C		;ROM buffer
!R15			= $301E		;Program counter
!SFR			= $3030		;SuperFx status flag
!PBR			= $3034		;Program bank
!ROMBR			= $3036		;
!RAMBR			= $303C		;
!CBR			= $303E		;
!SCBR			= $3038		;
!SCMR			= $303A		;
!BRAMR			= $3033		;
!VCR			= $303B		;
!CFGR			= $3037		;
!CLSR			= $3039		;
;======================================================================================

print "INIT ",pc
	RTL

print "MAIN ",pc
	PHB
	PHK ; Always change the data bank.
	PLB
	JSR SprRoutine
	PLB
	RTL

SprRoutine:
	JSR Graphics
	STX !R6
	REP #$20
	LDY.b #(SpriteCustom>>16)&$3F	;\ Load the sprite code
	LDA.w #SpriteCustom				;/
	JSR $1E80
	JSL $01802A			; Apply speed.
	RTS

arch superfx
SpriteCustom:
	IBT R0,#(SpriteCustom>>16)&$3F	;\ Set ROM Bank
	ROMB							;/
	IBT R0,#$19	;\	This code loads the table below
	LDB (R0)	; | Based on player powerup (1-bit)
	AND #1		; | If Player is Small, it walks only
	MULT #2		;/	Otherwise, it jumps.

	LEA R10,SpriteTable	; Load the table
	FROM R10			;\	Load table value...
	TO R14				; |
	ADD R0				; | Add with the values we set with R0
	GETB				; |
	INC R14				; |
	TO R10				; |
	GETBH				; |
	JMP R10				;/	Jump according to the result
	CACHE				; Jump and load code to be on CACHE RAM
	;The whole routine above, sets the ROM bank and loads the Sprite table below
	;After that, it gets the RAM value set in R0 and then does everything SNES does
	;Multiplies the value by 2 and add to the current table value which THEN is buffered
	;Transferred to R10 and properly jumps at the address location

SpriteTable:
	dw Walk
	dw Jump

	NOP
Walk:
	IBT R0,#$08
	LEA R1,$B6
	WITH R1
	ADD R6
	STB (R1)
	STOP
	NOP

	NOP
Jump:
	SUB R0
	LEA R1,$B6
	WITH R1
	ADD R6
	STB (R1)

	IBT R0,#$F8
	LEA R1,$AA
	WITH R1
	ADD R6
	STB (R1)
	
	STOP
	NOP
arch 65816

Graphics:
	REP #$20
	LDY.b #(GDI>>16)&$3F	;\ Load the sprite code
	LDA.w #GDI				;/
	JSR $1E80
	LDA $75C4,x				;\	This is a RAM related to off-screen...
	BNE	++					;/	If this isn't zero, pull A and return to new routine?
	LDY !R1					;RAM_SprOAMIndex   ; Y = Index into sprite OAM
	BRA	+
++	PLA	;\	I don't understand much, from what I know...
	PLA	; | This pulls A two times and returns to original routine.
		;/	Without OAM Index. This saves a BIT of time, the original didn't had this.
	; after: Y = index to sprite OAM ($300)
	;	$00 = sprite x position relative to screen boarder 
	;	$01 = sprite y position relative to screen boarder  

+	LDA $00                 ; set x position of the tile
	STA $6300,y

	LDA $01                 ; set y position of the tile
	STA $6301,y

	LDA #$24
	STA $6302,y

	LDA $75F6,x             ; get sprite palette info
	ORA $64                 ; add in the priority bits from the level settings
	STA $6303,y             ; set properties

	LDY #$02                ; #$02 means the tiles are 16x16
	LDA #$00                ; This means we drew one tile
	JSL $01B7B3
	RTS

;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; GET_DRAW_INFO - Optimized version
; This is a helper for the graphics routine.  It sets off screen flags, and sets up
; variables.  It will return with the following:
;
;		Y = index to sprite OAM ($300)
;		$00 = sprite x position relative to screen boarder
;		$01 = sprite y position relative to screen boarder  
;
; It is adapted from the subroutine at $03B760
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; GetDrawInfo - Optimized version
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
arch superfx
GDI:
	LEA R1,$14E1	; LDA $14E0,x : XBA : LDA $E4,x
	WITH R1			; $14E0 is using an odd number because
	ADD R6			; In Super FX, odd addresses are loaded like this
	TO R7			; High byte = Address-1
	LDW (R1)		; Thus, it'll effectively load $14E0, only at high byte though
	LEA R1,$E4		;
	WITH R1			;
	ADD R6			;
	TO R8			;
	LDW (R1)		; The MERGE operation grabs high byte (R7) and low byte (R8)
	MERGE			; and obviously MERGEs it

	LMS R1,($1A)	; REP #$20 : SEC : SBC $1A : STA $00
	SUB R1			; CLC : ADC #$0040 : CMP #$0180 : SEP #$20
	SMS ($00),R0	;
	MOVE R12,R0		; This is used in order to avoid another buffering
	IBT R1,#$40		;
	IWT R2,#$0180	;
	ADD R1			;
	CMP R2			;
	FROM R12		; LDA $01 : BEQ +
	HIB				; Bearing in mind, the earlier operation stored R0 on $00
	BEQ +			; Thus, another loading is not necessary
	NOP
	IBT R0,#$01		; LDA #$01 : STA $15A0,x
+	LEA R2,$15A0	; or
	WITH R2			; [...] : LDA #$01 : + STA $15A0,x
	ADD R6			;
	STB (R2)		;
	XOR R0			; TDC (Technically Zero)
	ROL				; ROL A : STA $15C4 : BNE
	LEA R2,$15C4	;
	WITH R2			;
	ADD R6			;
	ALT1			; This ALT and STW is a smart way of saving cycles
	BNE .Invalid	; AND being efficient
	STW (R2)		;

	LEA R1,$14D5	; LDA $14D4,x
	WITH R1			; Same case as above, unlike above routine
	ADD R6			; It will be used several times, so it is "ready for usage"
	TO R7			; High byte = Address-1
	LDW (R1)		;
	IBT R1,#$20		; LDA $1662,x : AND #$20 : BEQ
	LEA R2,$1662	;
	WITH R2			;
	ADD R6			;
	LDW (R2)		;
	AND R1			;
	BEQ .CheckOnce	;
	NOP
.CheckTwice
	LEA R1,$D8		; $14D4 is already at high byte...
	WITH R1			; LDA $D8,x : REP #$21 : ADC #$001C
	ADD R6			; SEC : SBC $1C : SEP $20
	TO R8			;
	LDW (R1)		;
	MERGE			;
	IBT R1,#$1C		;
	ADD R1			;
	FROM R1			;
	LDW (R1)		;
	SUB R1			;
	BEQ .CheckOnce
	NOP
	IBT R0,#$02
.CheckOnce
	LEA R1,$186C	;LDA #$02 : STA $186C,x
	WITH R1			;or
	ADD R6			;[...] : LDA #$02 : + STA $186C,x
	STB (R1)

	LEA R1,$D8		; $14D4 is already at high byte...
	WITH R1			; LDA $D8,x : REP #$21 : ADC #$000C
	ADD R6			; SEC : SBC $1C : SEP $21 : SBC #$0C
	TO R8			; STA $01 : XBA : BEQ : INC $186C,x
	LDW (R1)		; LDY $15EA,x
	MERGE			;
	ADD #$C			;
	LMS R1,($1C)	;
	SUB R1
	MOVE R12,R0
	SUB #$0C
	IBT R1,#$01
	STB (R1)
	FROM R12
	HIB
	BEQ .OnScreenY
	NOP
	LEA R1,$186C
	WITH R1
	ADD R6
	LDB (R1)
	INC R0
	STB (R1)
.OnScreenY
	LEA R1,$15EA
	WITH R1
	ADD R6
	LDW (R1)
	STOP

.Invalid
	STOP
	PLA
	PLA
	RTS
arch 65816