• (nodebb)

    The user must do something? I don't think so. The user always has the option of leaving your website, turning the computer off, and going to the beach.

  • (nodebb)

    Oh great, another one who wants to punish poor john@uk...

  • Grammar Nazi (unregistered)

    Remy, the parenthetical comment contains a "they're"/"their" mistake.

  • JanVP (unregistered)

    "in all they're forms" -> their

  • Peter L (unregistered)

    I love the check that the address shouldn't end with a dot, since that kicks out the very valid [email protected]. variant.

  • Rob (unregistered) in reply to Mr. TA

    It says they must do theze things "for a response." If they don't want a response then, yes, they are free to leave. I don't think the statements are inaccurate.

  • (nodebb)

    Apparently the programmer heard that if you solve a problem with regex, you now have two problems, so they solved the problem with more problems.

  • (nodebb) in reply to Peter L

    Wait, an email address can end with a dot? Is [email protected]. just another way to write [email protected] or is it a different address?

    Addendum 2025-03-03 09:58: Looks like Cloudflare is protecting those sample email addresses. The question is whether the domain domain.tld. is different from domain.tld.

  • (nodebb) in reply to jkshapiro

    Looks like Cloudflare is protecting those sample email addresses

    Well, duh. You should have used [email protected]. and [email protected], as God and RFC 2606 intended.

  • (nodebb) in reply to jkshapiro

    I'm pretty sure they're equivalent. In DNS, ending a name with a dot makes it absolute instead of relative. But I don't think the hostname in mail is ever treated as relative, so a trailing dot is redundant.

  • Argle (unregistered)

    Beginner: email.contains('@');

    Intermediate: (?:[a-z0-9!#$%&'*+/=?^_`{|}~-]+(?:\.[a-z0-9!#$%&'*+/=?^_`{|}~-]+)*|"(?:[\x01-\x08\x0b\x0c\x0e-\x1f\x21\x23-\x5b\x5d-\x7f]|\\[\x01-\x09\x0b\x0c\x0e-\x7f])*")@(?:(?:[a-z0-9](?:[a-z0-9-]*[a-z0-9])?\.)+[a-z0-9](?:[a-z0-9-]*[a-z0-9])?|\[(?:(?:(2(5[0-5]|[0-4][0-9])|1[0-9][0-9]|[1-9]?[0-9]))\.){3}(?:(2(5[0-5]|[0-4][0-9])|1[0-9][0-9]|[1-9]?[0-9])|[a-z0-9-]*[a-z0-9]:(?:[\x01-\x08\x0b\x0c\x0e-\x1f\x21-\x5a\x53-\x7f]|\\[\x01-\x09\x0b\x0c\x0e-\x7f])+

    Advanced: email.contains('@');

  • (nodebb)

    If I just saw "@uk..." or "@example.tld." in an email field, I'd say that's invalid, and 99% of the time I'd be right.

    These are not RFC 5322 compliant.

  • MRAB (unregistered)

    What does str.indexOf(at)==lstr do?

    If the character is present, indexOf returns its position, which will be in the range 0 to length - 1.

    lstr is the length of the string.

    Therefore, str.indexOf(at)==lstr will always be false.

    Similarly, str.indexOf(dot)==lstr will always be false.

  • validationAlert (unregistered) in reply to Ralf

    Validation is like crack to some programmers. If it's 99.9999999%, still don't. Even a syntactically valid address may not actually exist. And what's the chance that any user will ever accidentally omit the TLD and be saved by your cleverness? A typo that leave steh address syntactically valid is infinitely more likely, so design your app to be able to easily recover from that state and the benefit of obsessive validation becomes even smaller. Make inner peace with the fact you can't truly validate email and stop trying. Those 5 users will feel enough relief for a million.

  • Klimax (unregistered)

    I love how every attempt at showing some weird email variant ends up in failure…

  • Vilx- (unregistered) in reply to MRAB

    like an icing on a cake, this is the off-by-one error, hiding in plain sight.

  • (nodebb) in reply to Klimax

    I love how every attempt at showing some weird email variant ends up in failure…

    Similar to date processing or "legal name" processing. It's just a thorny problem with very few good worked solutions that nobody seems to discover without going through the pain of doing it the wrong way a few times.

  • (nodebb) in reply to validationAlert

    Well, there's one stunningly simple way to validate email addresses: send an email message to that address with a confirmation link and wait for the user to click the link.

    ... though I have seen some convincing arguments to put up an "are you sure" message up if the entered domain is close to one of the big email providers (eg., "@gamil.com") or if the local part is close to the name (eg, if I entered "[email protected]").

  • Officer Johnny Holzkopf (unregistered)

    The code presented here is future-proof for upcoming "Ém@il 2.Ø", where you can simply write 'var at="$";' and 'var dot="::";' and nothing else needs to be changed...

  • (nodebb)

    RFC compliant email validator:

    function isValidEmail(value)
{
	let length = value.length;
	if(length < 3) return false;
	let index = value.indexOf('@');
	return index > 0 && index < length - 1;
}
  • (nodebb) in reply to Robert Fensterman

    Well, there's one stunningly simple way to validate email addresses: send an email message to that address with a confirmation link and wait for the user to click the link.

    It feels like there is some confusion out there between the difference of validation and verification. I have always known validation to be checking for well-formed-ness, to know if it is properly following the governing spec.

    Verification on the other hands, is where one checks if something exists. Pretty much every time I've made an account on a web site (message board, social media sites, blog comments sections, etc) there is always a verification email sent with a verification link used to verify that email address entered at the site does indeed exist.

    It can be compared to something like XML. A validator (like xmllint and others) makes sure it follows the spec/schema, it does not check that the data itself is correct or what was expected in a given XML document.

    So doing some basic validation can make sense (for email it's often better to be more lax about it, like the examples above of just checking for a '@' character), but it's the verification that actually makes sure it can be reached.

  • (nodebb) in reply to Steve_The_Cynic

    Oh great, another one who wants to punish poor john@uk

    He has bigger problems. The uk top level domain has no MX records.

    % dig -t mx uk 

; <<>> DiG 9.10.6 <<>> -t mx uk
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 16349
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 512
;; QUESTION SECTION:
;uk.				IN	MX

;; AUTHORITY SECTION:
uk.			950	IN	SOA	dns1.nic.uk. hostmaster.nic.uk. 1741179401 7200 900 2419200 10800

;; Query time: 13 msec
;; SERVER: fe80::8e83:94ff:fe91:6fee%14#53(fe80::8e83:94ff:fe91:6fee%14)
;; WHEN: Wed Mar 05 12:57:02 GMT 2025
;; MSG SIZE  rcvd: 87

    It's probably safe to assume user@tld is invalid

  • (nodebb)

    It's probably safe to assume user@tld is invalid

    Depends what you mean by "safe". At least at one time, the ai TLD had an MX record and was assigning email addresses. But yes, this is a very unusual case and I think it's safe to assume that everyone who had one of those addresses also had at least one other email address.

  • (nodebb) in reply to Vilx-

    like an icing on a cake, this is the off-by-one error, hiding in plain sight.

    In the programmer's defense, an e-mail address where str.indexOf(at)==lstr would be invalid.

  • TVJohn (unregistered)

    It's surprising that even big company websites can still get email validation wrong. I have a hyphenated name and my own domain. I was recently putting my email address on one of those big company sites (I think it was SRAM, but don't quote me on that) and it wouldn't accept the hyphen. Obviously, I have other email addresses I can use, but those tend to be for more throw-away purposes. I prefer to use my domain one when I think I'm going to be wanting that link for the foreseeable future.

  • Ishura (unregistered) in reply to jkshapiro

    For what it's worth, I just tried emailing myself with an extra dot. iOS Mail warned me that the address appeared to be invalid, but I told it to send anyway, and I successfully received the message.

Leave a comment on “An Alerting Validation”

Log In or post as a guest

Replying to comment #:

Log Out

« Return to Article