On Fri, 4 Sep 2015, I wrote:
> The test can be fixed by inserting "(*LF)", but what is more mysterious
> is why this was not picked up by the many tests that get run before a
> release - one of which configures with newline=crlf.
>
> I will investigate. Thanks for tracking down this issue.
Aha! It turns out that quite a few tests fail when a single LF is not a
valid newline, and the script for running lots and lots of tests knows
this, and does not run them. Doh! There is this comment:
# Some of the tests are automatically skipped when PCRE2 is built with non-Unix
# newline recognition because they don't work. I am hoping to reduce this as
# much as possible in due course.
I have clearly forgotten about this and done nothing about it. I will
work on it.
