The PgUtf8Codec is a simple modification of Utf8Codec to be able to use PostgreSQL 8.1 well.
PostgreSQL 8.1 refuses to store the unicode codepoint 0. The software reports that it is an invalid byte sequence and refers to htt://www.postgresql.org/docs/techdocs.50, but the real reason is that postgresql never was intended to store nulls in text, and versions up to 8.0 allowed it only by accident.
Since quite a few messages contain null bytes, we remap 0 to U+ED00 (a private-use codepoint, also used by Unknown8BitCodec) and back.
This class is not listed as a supported codec, since it's meant only for postgres use, not for any other purpose.
Constructs an empty PgUtf8Codec.
This web page based on source code belonging to The Archiveopteryx Developers. All rights reserved.