Re: [regext] Internationalized Email Addresses and EPP

> On 23 Nov 2020, at 22:07, Hollenbeck, Scott <shollenbeck=40verisign.com@dmarc.ietf.org> wrote:
> 
>> -----Original Message-----
>> From: Taras Heichenko <tasic@academ.kiev.ua>
>> Sent: Monday, November 23, 2020 1:46 PM
>> To: Dmitry Belyavsky <beldmit@gmail.com>
>> Cc: Hollenbeck, Scott <shollenbeck@verisign.com>;
>> Klaus.Malorny@knipp.de; regext@ietf.org; Gould, James
>> <jgould@verisign.com>; alex.mayrhofer.ietf@gmail.com
>> Subject: [EXTERNAL] Re: [regext] Internationalized Email Addresses and EPP
>> 
>> Caution: This email originated from outside the organization. Do not click links
>> or open attachments unless you recognize the sender and know the content
>> is safe.
>> 
>> Hi!
>> 
>>> On 23 Nov 2020, at 17:00, Dmitry Belyavsky <beldmit@gmail.com> wrote:
>>> 
>>> Dear Scott,
>>> 
>> 
>> [skip]
>> 
>>> 
>>> This may be the path of least resistance. I'm still trying to think through hat
>> would happen if a registry returns an internationalized email address to a
>> registrar that doesn't expect one. This could happen after a domain transfer,
>> for example. Is this a problem? If not, maybe we could just get by without
>> any other protocol changes or extensions.
>>> 
>>> From my point of view, if the registry has implemented EAI support, all the
>> registrars will have to do it. They should deal with the clients with such emails
>> _somehow_.
>>> E.g., they hardly can reject the transfer relying on this reason.
>> 
>> There are cases when registry cannot enforce registrars to implement some
>> features. But I see no problem in the registry response with EAI. EPP works
>> with UTF-8 encoding so EAI should not cause the EPP interface crash. Such
>> email can cause strange symbols in the web interface but as for me it should
>> not bring serious problems.
> 
> That's only true if you're not also doing syntax checking of the local part outside of what your XML parser is doing with the schema. A regular expression to match an ascii-only local part would NOT match an internationalized email address.

Usually, software developers use libraries (libxml2 usually) to parse XML. If the XML
contains encoding=utf-8 in its header it would be instruction for the parser to use utf-8
encoding. Of course, further checking will give an error if we try to parse data returned
from a registry. But I see no reason to parse this data – we should show what we got 
from the registry. (Of course, it does not mean that this is impossible.)

> 
> Scott
> 
> _______________________________________________
> regext mailing list
> regext@ietf.org
> https://www.ietf.org/mailman/listinfo/regext

--
Taras Heichenko
tasic@academ.kiev.ua