Oops; looks like I forgot to attach my version last night; good thing, found a
couple of bugs tonight.
Looks like it could be speeded up a bit; no error checking.
I'd forgotten what a PITA it was to program on the real hardware.
1 DEFSTRH:H="0123456789ABCDEF":GOTO100
5 A=INT(X/4096):X=X-A*4096:B=INT(X/256)
6 PRINTMID$(H$,A+1,1)+MID$(H$,B+1,1);
7 C=INT((XMOD256)/16):D=XMOD16:Y=C*16+D
8 PRINTMID$(H,C+1,1)+MID$(H,D+1,1)+" ";
9 RETURN
20 X=0:X$=RIGHT$("0000"+X$,4)
30 FORI=1TO4:Y=ASC(MID$(X$,I,1))
40 Y$=CHR$(Y+(Y>95)*32)
50 L=INSTR(1,H,Y$)-1:X=X+L*(16^(4-I))
60 NEXT:RETURN
100 INPUT"From";S$:INPUT"to";E$
110 X$=S$:GOSUB20:S=X
120 X$=E$:GOSUB20:E=X
200 B$=TIME$:FORI=STOESTEP8:X=I:GOSUB5
210 L$="":FORJ=0TO7:X=PEEK(I+J):GOSUB7
220 Y=C*16+D
230 Y$=".":IFY>31ANDY<127THENY$=CHR$(Y)
240 L$=L$+Y$:NEXT:PRINTL$:NEXT
250 PRINTB$+" to "+TIME$:END
Constructive criticism welcome.
m
----- Original Message -----
From: "runrin" <[email protected]>
To: <[email protected]>
Sent: Wednesday, October 26, 2022 10:24 PM
Subject: Re: [M100] Notoriously S.L.O.W BASIC posted - help speeding it up
appreciated
> this seemed fun so i gave it a quick try. here's what i came up with.
>
> it's a shame you can't do bitshift operations because i suspect
> ``((V AND 240) >> 4)'' would be faster than using integer division.
>
> i'm a total neophyte when it comes to basic so i don't think this code
> is actually very fast. at least it's faster than the screen scrolls :P.
> in both examples i'm just peeking 0-59 since it's what fits on the
> screen nicely. that's also the reason for the " ".
>
> i like this version because it makes the print statement easy to read,
> it takes a lot more memory to store each character as it's own string
> though.
>
> 10 DATA "0","1","2","3","4","5","6","7","8","9","A","B","C","D","E","F"
> 20 DIM H$(15)
> 30 FOR I = 0 TO 15
> 40 READ H$(I)
> 50 NEXT
> 60 FOR I = 0 to 59
> 70 V = PEEK(I)
> 80 PRINT H$(V \ 16); H$(V AND 15); " ";
> 90 NEXT
>
> ---
>
> i made another version that uses string pointers because i felt bad
> using 16 strings. the PEEKS seem to be pretty expensive though, so i
> think it's actually slower than the previous version, but the code is a
> bit clearer. using MID$ is probably faster, but i mostly just did this
> for fun. this is more along the lines of how i would do something like
> this in C, which is the language i'm most familiar with.
>
> string pointers are weird in basic. the pointer returned by VARPTR(S$)
> points to the following:
>
> ptr + 0 = string length
> ptr + 1 = low byte of pointer to string
> ptr + 2 = high byte of pointer to string
>
> i had to look in the basic language lab (pp. 182) to find that
> information. it's not even listed on
> https://help.ayra.ch/trs80-reference . i think i would have preferred
> if it just returned the string pointer with null terminator lol. not
> sure why they did it this way. anyway, here's the code:
>
> 10 H$ = "0123456789ABCDEF"
> 20 P = VARPTR(H$)
> 30 SP = PEEK(P + 1) + PEEK(P + 2) * 256
> 40 FOR I = 0 TO 59
> 50 V = PEEK(I)
> 60 PRINT CHR$(PEEK(SP + (V \ 16))); CHR$(PEEK(SP + (V AND 15))); " ";
> 70 NEXT
>
> i know those extra spaces make my code slower, i'm just formatting it
> that way here so its actually legible.
>
> -runrin
