On 2013-10-17, Axel Rau <axel.rau@???> wrote:
> The attached patch against 4.82.RC3 implements a new expansion operator, which
> replaces invalid utf-8 sequences by an replacement character. Testcase and
> documentaion update are included.
patch:
#define UTF8_REPLACEMENT_CHAR US"?"
It seems to me that "� " might be a better choice.
AIUI this is written as "\uFFFD" or "\xEF\xBF\xBD" in C source.
