Re: [xml] Character encodng cleanup

["Peter Jacobi" <pj walter-graphtek com>]

Hi Daniel,

I've found (unfortunately) another use of the deemed-to-be-obsolete
charset field:

In parserInternals.c, HTMLparser.c and DOCBparser.c, when an
invalid UTF-8 encoding is detected, ctxt->charset is set to
XML_CHAR_ENCODING_8859_1.

Given that this would indicate 8859-1 encoding of the in-memory
strings (as leading researchers tend to believe), this doesn't seem
to be that useful.

Do you have any test case where this does anything good?

Regards,
Peter Jacobi