2009/10/2 Rafael Garcia-Suarez <rgs@consttype.org>: > 2009/10/2 karl williamson <public@khwilliamson.com>: >> FFFE is a 'noncharacter'. It isn't an 'illegal character'. It is perfectly >> legal in Unicode to use this code point. It is, however, illegal to use it >> in "interchange", and hence would never be confused with a BOM from an input >> stream. I believe the message should be changed to be something more like >> Unicode noncharacter 0xfffe is illegal for interchange >> >> with a fuller explanation in perldiag.pod, and the ability to turn off just >> this warning. > > Like this ? > > =item Unicode non-character %s is illegal for interchange > > (W utf8) Certain codepoints, such as U+FFFE and U+FFFF, are defined by the > Unicode standard to be non-characters. Those are legal codepoints, but are > reserved for internal use; so, applications shouldn't attempt to exchange > them. If you know what you are doing you can turn off this warning by > C<no warnings 'utf8';>. (now committed as 6f6ac1dea8501596050bc974dc468632797d51eb)Thread Previous | Thread Next