I would consider any such encoding a bug. I can't see a rational for degrading from text to binary, unless you use a bit oriented operator.
Decoding from binary to text can happen, like if a transformation like lc or uc can't map otherwise.°
But that's documented and can be configured too, AFAIR.
And to be honest, treating a binary string as ASCII/Latin-1 is guesswork from Perl, it could be anything.
Strictly speaking lc and uc shouldn't be allowed then, the programmer should have properly decoded it first.
Cheers Rolf
(addicted to the 𐍀𐌴𐍂𐌻 Programming Language :)
Wikisyntax for the Monastery
°) or generally if only utf8-text characters are added.
In reply to Re^9: Seeking Perl docs about how UTF8 flag propagates
by LanX
in thread Seeking Perl docs about how UTF8 flag propagates
by raygun
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |