Byte Ptr
BlitzMax Forums/BlitzMax Beginners Area/Byte Ptr
| ||
| To date the term "Byte Ptr" still has me confused. Afterall, its not making a variable point to a 'Byte-sized' variable. or is it? Can someone explain (in laymans terms) what goes on when you use Bte Ptr? I've used it in Win32 API functions like this:
Extern "Win32"
Function InvertRect(hdc%,tREC:Byte Ptr)
End Extern
...
Local r:RECT=New RECT
r.l=x ; r.t=y ; r.r=x+width ; r.b=y+height
InvertRect bufferDC,Byte Ptr(r)
It seems to work without error. So, is Byte Ptr the same as C's *pointers? What about Int Ptr, Float Ptr, Varptr, Var, etc .. Why, where, when, how, would you use these? |
| ||
| basicaly all it is, is the adress of an individual byte in memory, for any other type the data that stores the value must have an address, it is the first byte of this data that the byte ptr is pointing to, whether the data is a 1 byte or 1meg long is up to you.... |
| ||
| I think I get what your saying. Lets take the API InvertRect again as an example. Here is a description of how the function works: "InvertRect" (ByVal hdc As Long, lpRect As RECT) As Long · hDC Identifies the device context. · lprc Points to a RECT structure that contains the logical coordinates of the rectangle to be inverted. Ok, since lprc needs to point to a RECT structure we do this by using 'Byte Ptr' in BlitzMax:
Extern "Win32"
Function InvertRect(hdc%,tREC:Byte Ptr)
End Extern
Type RECT
Field l%, t%, r%, b% ' left top right bottom
End Type
Local r:RECT=New RECT
r.l=x ; r.t=y ; r.r=x+width ; r.b=y+height
InvertRect bufferDC,Byte Ptr(r)Therefore, InvertRects second parameter will receive the address of the first byte in the RECT structure: Sound about right to you? |
| ||
| yep, funny how a picture explains it better than by brain dead english if there was a function that needed 2 ints you could for instant send it the byte ptr of rect.r if you wanted to send it r,b have I just made things more confusing? |
| ||
| have I just made things more confusing? Actually thats pretty well clear ..Local myrect:RECT=New RECT myrect.r=123 ; myrect.b=252 InvertRect bufferDC,Byte Ptr(myrect.r) I think the confusion comes from passing the myrect instance rather that the first field in the myrect type. Afterall, whats the difference between these two? .. Type RECT Field l%, t%, r%, b% ' left top right bottom End Type Local myrect:RECT=New RECT myrect.l=10 ; myrect.t=20 ; myrect.r=30 ; myrect.b=40 ' is this InvertRect bufferDC,Byte Ptr(myrect) ' the same as this? InvertRect bufferDC,Byte Ptr(myrect.l) |
| ||
| Not the same. The first line is converting the `myrect ptr` into a byte pointer - giving you an address of the first byte of that instance of RECT. The second line is giving you the pointer to the first byte of the first field within that instance. You might assume they are the same, they're probably not, I think Max has some other data stored at the beginning of a type instance, so the variables do not begin at 0 offset from the type's location. Or i could be wrong. |
| ||
| I think it used to be that way, but they seem to be the same now, which is rather handy... |
| ||
Just tested this therory,Type RECT Field l%,r% End Type Local myrect:RECT=New RECT myrect.l=123 Local bp1:Byte Ptr=Byte Ptr(myrect) Local bp2:Byte Ptr=Byte Ptr(myrect.l) Print Int(bp1) Print Int(bp2) Results: bp1 = address of myrect structure bp2 = value stored at myrect.l |
| ||
| yeah, i'm not sure that behaviour makes sense! try getting varptr's for myrect and myrect.l .... They should be the same address ie bp1 |
| ||
Chris C is right. Check it out.Type RECT Field l:Int,r:Int End Type Local myrect:RECT=New RECT myrect.l=123 Local bp1:Int Ptr=Int Ptr(Byte Ptr(myrect)) bp1[0] = 3444 Print myrect.l |
   |