Bitwise AND on Double variables

The range of long integers in Access VBA is:
-2147483648 to 2147483647

Your number is outside the range and hence the overflow!

What is the type of the variable, is it declared as a VB Double?
Do you want to check the individual bits as they are stored in memory (however that is) or is it the logical values - i.e. powers of two?

Well, yes, that's what I'm saying. Bitwise AND is apparently limited to long variables, but I need to use it on doubles. So I'm looking for a way to either split a double into two long variables or an alternative AND function.

Hi MikeToole,

Yes, I need to check individual bits. Currently I'm using doubles, but I can change that if needed. However, I need at least 50 bits, so long variables are not enough.

Background:
I have a list of patients who are housed in different sections of the building (numbered 1 to 50). A caretaker is only allowed to consult the files of a patient in his/her section. However, some caretakers are assigned to several sections, so they should have access to all patients of those sections.
   Caretaker 1:  section 1, 4, 5
   Caretaker 2:  section 3, 4, 7
   etc.
The way I want to solve it is by filtering the list of patients using the section number as a bit in a mask. So in the above example the permissionsmask of caretaker 1 would be (binary) 10011. A bitwise AND on the section nr of the patient returns a true or false, which can be used in the queries.

hope this makes sence to you :-)

regards,
Michiel

>> permissionsmask of caretaker 1 would be (binary) 10011

This should be right-to-left: 11001

The limitation looks to be in the AND operator, it doesn't seem to want to return anything than can't go in a Long.
How are the patient records stored, what are you using to filter them?

Thanks for replying.
Heavily simplified it looks like this:

tbl_Patients:
ID | Name       | Section | pMask 
---------------------------------
1  | Andersson  | 1       | 1      (bit 1 set)
2  | Johnson    | 4       | 8      (bit 4 set)
3  | Parker     | 7       | 64     (bit 7 set)
 
tbl_Caretakers:
ID | Name       | cMask
-----------------------
1  | Joe        | 65     (bits 1 & 7 set)
2  | Ann        | 72     (bits 4 & 7 set)
 
To list the patients Joe can see, I would query:
SELECT * FROM tbl_Patients WHERE (pMASK AND 65) = pMASK

Select allOpen in new window

I see what you mean.
2.147483648e9# and 1  fails
2.147483647e9# and 1  returns 1 which it should

Clearly VBA is converting the double to a long to do the bitwise comparison. It cannot convert 2.147483648e9# so you get an overflow. I can understand this because I cannot get my head around bitwise operations on a float data type. How do you perform bitwise comparisons on the exponent component? I would expect bitwise comparisons would be disallowed on float variables although I have not been able to google an article on the subject.

I think you would need to use two longs to keep track of 50 bits.

Found on article. Refers to Basic:
http://support.microsoft.com/kb/61434

Thenelson,
Your link doesn't really solve the problem as it merely moves de values from a single to a long, not a double to two longs.

MikeToole,
Your solution is certainly an option. It might put a performance penalty on the query, so I'll have to run a few tests to see if it's workable. However, it also removes the limit of 64 sections, which definately is a plus.