Make tests less istreambuf_iterator implementation dependent

[Jonathan Wakely]

On 04/10/17 18:21 +0200, François Dumont wrote:
>On 03/10/2017 16:20, Jonathan Wakely wrote:
>>Doesn't the modified test PASS anyway, even without changing how _M_c
>>is used?
>
>No it doesn't because the first check for eof capture the 'a' char 
>which is never reseted so after string construction it is still 
>pointing to this char and is not an eof iterator. Without the _M_c 
>assignment the iterator is still "live" and so keeps on reflecting the 
>streambuf state.

Ah yes, I must have not adjusted the test properly when I tried to
check that. Your new test does indeed fail with the old
implementation.

>Attached patch committed with:

Thanks.