Umlauts in email patterns

[lesomue]

Hello,

email patterns within eForms are currently very strict, which means that no umlauts or similar are permitted. However, we are receiving a number of enquiries from our contracting authorities as to why they cannot use their email addresses containing umlauts such as ä, ö or ü.
We are wondering whether it is possible to adapt the email pattern so that these email addresses can also be taken into account?
Background: In Germany, domains and email addresses with umlauts have been commonplace for several years now.
See also https://www.denic.de/en/faqs/faqs-about-idns-ss or IDNA2008 or RFC 5890 (and following).
Please let us know if we can expect an adaptation.
Many thanks!

[bertrand-lorentz]

Thanks for your feedback.

We plan to address this in SDK 1.14.0, by changing the regular expression used to validate e-mail addresses.
It currently allows alphanumerical characters from the ASCII set: A-Z, a-z, 0-9.
We will use instead the "\w" shorthand character class.
In XPath this matches all "letter-like" Unicode characters, not only ASCII (see https://www.w3.org/TR/xmlschema-2/#nt-MultiCharEsc), but in for some other implementations/flavors of regular expressions "\w" only matches ASCII characters.
So if you use the regular expression with an implementation for which "\w" corresponds only to ASCII characters, then this will not have the intended effect.
We cannot list all possible acceptable characters, so this is the best solution we could find.

The new regular expression we plan to use is:
^[\w!#$%&''*+/=?_-]+(\.[\w!#$%&''*+/=?_-]+)*@([\w]([\w_-]*[\w])?\.)+([\w]{2,})$

[lesomue]

Thank you for your answer!