[Pypgsql-users] possible problem with unicode_results = 1

[Karsten H. <Kar...@gm...>]

Hello community,

I am not sure I understand how unicode_results is supposed
to work right:

Assume this:

- client_encoding is set to 'iso8859-1' at the connect() level
- hence "set client_encoding" must also be set to 'iso8859-1' at the SQL level
- thereby, strings *returning* from the database are encoded
  as 'iso8859-1', too, when they enter pyPgSQL from libpq

- now I want to have results in unicode
- hence I set unicode_results=1 at the connect() level
- thereby string-types are passed to unicode() in pyPgSQL
  without further arguments as per the code
- hence unicode() uses the default encoding as per Python docs
- which in most cases is "ascii"
- bad things happen

Really, the client_encoding value should be passed in to
unicode() as well, shouldn't it ?

Or is my logic flawed ?

No, it is not always possible to set the Python default
string encoding to something sane as would seem prudent to
do. It is not recommended to do so, either.

Karsten
-- 
GPG key ID E4071346 @ wwwkeys.pgp.net
E167 67FD A291 2BEA 73BD  4537 78B9 A9F9 E407 1346