| Subject: | case not preserved |
Howdy and thank you for Net::IDN::Encode!
That case is not preserved may be part of the spec (if so perhaps a note in the POD?) but I noticed that at least one converter does preserve it:
e.g. using: I.♥.perl
1. Net::IDN::Encode and charset.org agree when punycoding something:
a. perl -MNet::IDN::Encode=:all -le 'print domain_to_ascii("I.\x{2665}.perl");' # I.xn--g6h.perl
b. http://www.charset.org/punycode.php?decoded=I.%E2%99%A5.perl&encode=Normal+text+to+Punycode
2. Net::IDN::Encode and charset.org do not agree when un-punycoding something, namely the case difference:
a. perl -MNet::IDN::Encode=:all -C -le 'print domain_to_unicode("I.xn--g6h.perl");' # i.♥.perl
b. http://www.charset.org/punycode.php?encoded=I.xn--g6h.perl&decode=Punycode+to+normal+text (I.♥.perl)
Problem can be seen by noting that in 2.a the I is lowercased but in 2.b the I remains uppercased.
So I guess the question is: 2.b or not 2.b (sorry couldn't resist ;p)?