Hello,
In a postgresql-9.3.1 database with UTF8 encoding I can do:
select convert_from (E'\\x68656c6c6f', 'LATIN1');
convert_from
--------------
hello
But when I explicitly give the "to" encoding:
select convert (E'\\x68656c6c6f', 'LATIN1', 'UTF8');
convert
--------------
\x68656c6c6f
Why does that second one give different results from the first?
Second question: why is that none of the convert* functions are
marked as immutable (thus preventing me from creating a functional
index using them). Surely if I convert \x68 to utf-8 the result
will *always* be "h", won't it?