Neil Conway writes:
> nconway=# select 1::bit;
> bit
> -----
> 0
> (1 row)
Oops. I've always thought that casting between int and bit should be
disallowed, but apparently it keeps sneaking back in.
> nconway=# select X'4'::bit varying;
> varbit
> --------
> 0100
> (1 row)
> -- why is that 4 bits, not 3?
SQL says so:
12) The declared type of a <hex string literal> is fixed-length
bit string. Each <hexit> appearing in the literal is equivalent
to a quartet of bits: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C,
D, E, and F are interpreted as 0000, 0001, 0010, 0011, 0100,
0101, 0110, 0111, 1000, 1001, 1010, 1011, 1100, 1101, 1110,
and 1111, respectively. The <hexit>s a, b, c, d, e, and f have
respectively the same values as the <hexit>s A, B, C, D, E, and
F.
> nconway=# select "bit"('14'::int);
> bit
> ----------------------------------
> 00000000000000000000000000001110
> (1 row)
> -- shouldn't bit be equivalent to bit(1), which should be
> right-truncated?
It is, but here you're calling a function, not referring to the type.
--
Peter Eisentraut [EMAIL PROTECTED]
---------------------------(end of broadcast)---------------------------
TIP 9: the planner will ignore your desire to choose an index scan if your
joining column's datatypes do not match
