https://bugs.exim.org/show_bug.cgi?id=2120
--- Comment #2 from Christian Persch (GNOME) <chpe@???> ---
>From my POV as a PCRE user, I think this would be rather surprising.
NO_UTF_CHECK is used when you've previously ensured that strings you pass to
PCRE are valid UTF, so the extra check inside PCRE would just be redundant. I
would not expect that NO_UTF_CHECK also lets through patterns that are valid
UTF as strings, but contain invalid codepoints such as surrogates via \x{...}
escapes. I would want the compile step to reject such patterns. So IMHO if this
proposed behaviour is desirable for some applications, it would be better to
add this as an option, independent of NO_UTF_CHECK.
--
You are receiving this mail because:
You are on the CC list for the bug.
