This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
toke.c: Consistently upgrade under encoding
authorKarl Williamson <khw@cpan.org>
Wed, 19 Nov 2014 05:02:21 +0000 (22:02 -0700)
committerKarl Williamson <khw@cpan.org>
Fri, 21 Nov 2014 04:45:18 +0000 (21:45 -0700)
commit5eef447b12283016e9c57a31362edb829884a7ed
treed1bd82d014afa163f7c5aaaac3ee9bd7bff779f8
parentd5b4785c1592d741608f5cff697058cc284db4eb
toke.c: Consistently upgrade under encoding

The documentation says that intermixing above-Latin1 code points with
ones that would be otherwise encoded to something else, like Greek,
causes the encoding to be foregone.  Until this commit, this only
happened when the above-latin1 code point came first in the string
constant being scanned; meaning string-order was important.  This
changes things to match the documentation
cpan/Encode/t/encoding.t
toke.c