Re: My bug or else regarding Visibility Rules

[Simon Wright <simon@pushface.org>]

Anh Vo <anhvofrcaus@gmail.com> writes:

> On Thursday, April 4, 2013 1:19:54 AM UTC-7, Simon Wright wrote:
>> Anh Vo <anhvofrcaus@gmail.com> writes: > On Tuesday, April 2, 2013
>> 1:16:34 PM UTC-7, Simon Wright wrote:

>>> I do not understand this. Why isn't Buffer(1) valid if my index
>>> starts at 1?
>
>> Because after the index has wrapped round Buffer'Old(1) will not be the 
>> same as Buffer(1). 
>
> In this case the loop will drop out because loop lower end is greater
> than loop upper end.

Consider a queue of max length 5 containing letters. 5 letters (a, b, c,
d, e) have been Put, and one (a) Got.

   I
   a b c d e
     O

After the next letter is inserted, you'd have

     I
   f b c d e
     O

Your check is

   (for all I in 1 .. Queue_Length'Old =>
                 Buffer(I) = Buffer'Old (I))

which is going to fail.

>> No. You wrote 
>>   (Index mod Length) + 1 
>> but it should be 
>>   (Index + 1) mod Length 
>
> (In_Index + 1) mod length will raise Constraint_Error when the Queue
> is just full. In fact, (10 + 1) > 10 when the queue has the lenght of
> 10.

OK, I've never seen this way of managing an index but I can see it
working. Still think that 0-based is easier.

That said, (In_Index + 1) won't raise CE, because Index is a subtype and
inherits the operations of Integer. CE on assignment, sure, but the 'mod
Length' would have been applied by then.