Top-Level Domain (TLD) validation function and email validation/sanitization

[stephdau]

The following patch proposes a new global function, does_value_end_in_valid_tld, to allow for a string value to be tested for as ending in a valid TLD.

TLD list as per ​http://data.iana.org/TLD/tlds-alpha-by-domain.txt

Regex inspired by ​http://gitorious.org/statusnet/mainline/blobs/master/lib/util.php#line724

The patch also implements using does_value_end_in_valid_tld() in is_email() and sanitize_email(), to start with.

How about naming it ends_with_valid_tld() instead? It's shorter.

Also, ICANN recently announced that TLDs will be a lot easier to register so I'm not sure if this function is a good idea anymore.

ends_with_valid_tld(): sure, why not

TLD: Ideally, I'd say we should have the function, and once the TLDs are easier to register, find a way to make it dynamic by accessing some kind of API (or at least let people do via plugin with a hook). regardless, the existing list won;t become obsolete when they do.

Plus, you know ICANN: "soon" is a relative term, W3C-style. ;)

Revised the patch with new suggested name: ends_with_valid_tld()

Found the place where I read that:

​http://arstechnica.com/web/news/2010/11/icann-to-open-top-level-domain-floodgates.ars

Replying to scribu:

Found the place where I read that:
​http://arstechnica.com/web/news/2010/11/icann-to-open-top-level-domain-floodgates.ars

Oh cool, thanks for the ref! Okay, it does seem to be happening faster than I thought, but I guess we still have until 2012:

ICANN plans to release its guidebook for more public comments on Nov. 9, and to
approve it at its next board meeting in Colombia in the first half of December. If
all goes smoothly, they will begin accepting applications in May 2011 for new
TLDs that would start functioning sometime in 2012. ICANN has said they will
draw the line at 1,000 new TLDs per year.

FTR: this TLD check implementation was implemented on WordPress.com for signup and such, as a response to users mistyping their email addresses and being left in the dark. TLD validation was the sanest route for us.

Not sure about this, due to the pending changes. Would rather just make sure a hook is in place.

Hook: agreed, definitely a good way to go.

So can this be entirely latched on to the various is_email hooks in is_email()?