array questions . . .

Don Pomplun

can I:

1. have multidimensional arrays? e.g. VAR byte XX[noparse][[/noparse]20,2]

2. have arrays of constants? e.g. CON XX = 1,2,3

3. initialize VAR arrays simply? e.g. VAR byte XX = 1,2,3

syntaces I've tried don't seem to take. TIA,


suggested alternatives or refrences welcome.

-- Don

Don Pomplun

I saw a reference to defining a zero-element array. I winder if this will work:

var long xx[noparse][[/noparse]0]
con a=10, b=20, c=30

I wonder if that will that make XX[noparse][[/noparse]0] be 10, XX be 20 and XX be 30 ??

-- Don

Mike Green

Don,
1) No, although it's really easy to either do multiple dimension subscripting with one dimensional arrays or implement an array object where you have
a function for producing a value (XXvalue(20,2)) or for storing it (XXsave(20,2,value)).
2) You can have constant tables in a DAT section and the compiler takes the element size from the declaration of the first item in the table
(XX long 1,2,3) and will let you use subscript notation to access the elements.
3) Easiest is to use a constant table in a DAT section, just change the values from your Spin code with the usual subscript notation. The only place
where there's a significant difference between these two approaches is when there are multiple instances of the same object. There's a unique VAR
section set up for each instance, but only one shared DAT section.
Mike

Mike Green

Regarding your second posting: Named constants (CON section) don't exist in the compiled program. They're known only to the compiler. There's no memory allocated to them. I could be wrong, but I don't think it's legal to declare an array with zero size. I think the reference was to the notion of the zero'th element. Array indices start at zero so there's a zero'th element which is actually the first element, but with a subscript of zero.

Paul Baker

My previous post (eaten by my expired cookie) suggested exact what Mike has.

To expand on Mike's #1, to put a 2D array in a 1D array use [noparse][[/noparse]i+j*xdim] where i,j is your index and xdim is the number columns, the total size is [noparse][[/noparse]xdim*ydim].

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

Post Edited (Paul Baker (Parallax)) : 1/23/2007 10:45:39 AM GMT

Mike Green

A practical example for a 2x3 array:

CON aa = 2 ' x dimension
       bb = 3 ' y dimension
VAR xx[noparse][[/noparse]aa * bb]
PRI main
   repeat i from 0 to aa
      repeat j from 0 to bb
         xx[noparse][[/noparse]i * bb + j] = 0

Don Pomplun

I think I saw gthe zero-length reference in the Tricks & Traps section.
e.g.
VAR
byte xx[noparse][[/noparse]0]
byte A, B, C

XX[noparse][[/noparse]0] will be the same as A, XX as B, etc.
(unless I read it wrong -- still didn't help initialize the arrys, though)
Thanx for the hints on DAT. I'll work that for a while.
-- Don

Bill Henning

Umm... I do believe you mean...

myarray[noparse][[/noparse]y*xdim+x]

[noparse]:)[/noparse]

Paul Baker (Parallax) said...
My previous post (eaten by my expired cookie) suggested exact what Mike has.

To expand on Mike's #1, to put a 2D array in a 1D array use [noparse][[/noparse]i*j*xdim] where i,j is your index and xdim is the number columns, the total size is [noparse][[/noparse]xdim*ydim].

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.mikronauts.com - a new blog about microcontrollers

Paul Baker

You are correct, thank you for pointing that out.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.