I just got a copy of Virtual Acorn and it prompted me to get back into a little programming again. Last I did was years ago on an A3000. I don't have access to the PRMs or much documentation any more.

I'd appreciate some help with the following. I'm trying to write an app in BASIC that generates new BASIC programs and saves them into a file. I thought this would be reasonably straightforward until I remembered that BBC BASIC files are tokenised. I've since been struggling to work out the format of the tokenised files. What I'm stuck on is what the first three bytes on each line do. I guess they're something to do with the line numbers, but any informed information on how to manipulate them (or how to write to BASIC files in general) would be much appreciated.

If this is a stupid or oft-repeated query I apologise: as I say, I've only got back into programming recently. I did look around on the net for info, though, without much luck.

Cheers,
Chris

Or are you trying to read IN basic files as well?

Probably a look at the StrongEd or Zap source code would help.

BASIC programs don't have to be be tokenised. You could write them as plain ASCII, and with the right filetype, they should work.

I think you actually have to get BASIC to load the text and re-save it before you can run it normally. I may be wrong though.

You could have a look at Brandy, a cross-platform BASIC interpreter:

http://www.argonet.co.uk/users/dave_daniels/

The source will be in C though, so might not be the easiest to understand if you only know BASIC.

All I can offer you is what's in an old BBC machine code book, which features a couple of BASIC fiddling programs:

The first byte is &D, the new line character.
The next two are the line number (high byte first)
The next is the line length; adding this to the address of the &D should find the next &D
Then you have the line itself.

The end of a file should be terminated by any top-bit set byte, after the last &D. This means the line number limit is 32767.

There are a couple of tokens given: &F1 is PRINT, and &E0 is END. Tokens range from &80 to &FF; anything else can be taken as plain text

Cheers.

Loading plain text into BASIC (a feature implemented in BASIC V, I think) does work quite well, although it does have to tokenise it and store it in memory before running it. Bear in mind that it is possible to write a tokenised program that can't be generated from any plain text - not all programs will survive being converted to plain text and back.

If anyone's interested, I could post the list of tokens somewhere. Perhaps on the Iconbar BASIC pages?

If anyone's interested, I could post the list of tokens somewhere. Perhaps on the Iconbar BASIC pages?

If you like; just email them to me (apologies again for lack of progress with programming section; and now Finals looming...). But I've done a search on Google (for BBC Basic tokens) and found:

http://www.geocities.com/SoHo/6413/BBC-Ch47.html
http://www.geffers.demon.co.uk/files/chap2.htm
http://www.cix.co.uk/~rrussell/products/bbcwin/manual/bbcwinb.html

Are any of them any good?

ATB,
Matthew