On Wed, Feb 23, 2011 at 3:16 PM, Andrew Dunstan <andrew@dunslane.net> wrote:
> On 02/23/2011 10:09 AM, Peter Geoghegan wrote:
>>
>> On 23 February 2011 04:36, Greg Stark<gsstark@mit.edu> wrote:
>>>
>>> This is only true for server encodings. In a client library I think
>>> you lose on this and do have to deal with it. I'm not sure what client
>>> encodings we do support that aren't ascii-supersets though, it's
>>> possible none of them generate quote characters this way.
>>
>> I'm pretty sure all of the client encodings Tatsuo mentions are ASCII
>> supersets. The absence of by far the most popular non-ASCII superset
>> encoding, UTF-16, as a client encoding indicated that to me. It isn't
>> byte oriented, and Postgres is.
>
>
> They are not. It's precisely because they are not that they are not allowed
> as server encodings.
>
Well just as an example, BIG5 is a 16-bit encoding where the first
byte always has the high bit set. The second byte can't be a ' or "
because those aren't in the allowable range for the second byte. So it
might be safe anyways. However \ is in the allowable range so I'm not
sure.
In the case of BIG5 ascii characters are included inline so any byte
with the high bit *not* set that isn't the second byte of a two-byte
sequence is assumed to be ascii. So an ascii parser would work fine
modulo the problem above with backslashes.
But this is just a special case. Wikipedia implies it's also true for
shift-JIS but there's no guarantee it would work for other client
encodings.
--
greg