You are correct, and I was surprised, too.
Fortunately, I could easily adjust pcre2test so that is passes the
original LF newline to pcre2_compile().
For regcomp() this is not possible, hence testinput18 fails.
Ralf
On 05.09.2015 17:34, ph10@??? wrote:
> On Sat, 5 Sep 2015, Ralf Junker wrote:
>
>> >For me, newline == CRLF causes just testinput18 to fail.
> I get many more failures, including a lot in testinput1. For example:
>
> /(?m)^.*B/
> abc\nB
> - 0: B
> + 0: abc\x0aB
>
> I am surprised you don't see those.
