Does the local part of an e-mail address have a minimum length requirement?

Question

I am using this expression to validate e-mail addresses:

"^[_a-z0-9-]+(.[a-z0-9-]+)@[a-z0-9-]+(.[a-z0-9-]+)*(.[a-z]{2,4})$"

I noticed that in order for the e-mail address to validate, I need to put in at least 3 characters before the @ symbol. Is this a requirement? What if I just want to have an e-mail address with 1 or 2 characters before the @ symbol?

I found this here. I am not sure how to change it to allow a minimum of 1 character. — Xaisoft, Jul 19 '11 at 17:31
where did you find this on SO? This REGEX you provide is wrong, for many reasons. — Brad, Jul 19 '11 at 17:32
@Brad - I found it here: http://stackoverflow.com/questions/369543/validating-e-mail-with-regular-expression-vb-net/369554#369554 — Xaisoft, Jul 19 '11 at 17:35
Basically I just need to check if the user is entering something like a@a.com or a@a.au or a@a.com2 — Xaisoft, Jul 19 '11 at 17:35
I find all regex solutions to the problem of validating the structure of an e-mail address to be woefully inadequate. I have quite a number of examples of what some might argue are "edge cases" that are 100% valid. Some forms of notation (comments within `()` and target folder selection via `+` notation) are incredibly useful, but too often artificially restricted because of use of bad regexen such as several demonstrated here. Instead, split and validate the parts. It's a structured value. (Edited to note: but w/o arbitrary limits such as TLD length restrictions.) — amcgregor, Nov 17 '21 at 17:12

score 8 · Accepted Answer · answered Jul 19 '11 at 17:29

8

It is entirely possible to have an e-mail address with just one character. All of the guidelines are defined in RFC822: http://www.ietf.org/rfc/rfc0822.txt?number=822

The necessary REGEX: http://www.ex-parrot.com/pdw/Mail-RFC822-Address.html

answered Jul 19 '11 at 17:29

Brad

159,648
54
349
530

How would I change my regular expression to reflect that? – Xaisoft Jul 19 '11 at 17:29
Yes, that regex is way too long. The one I have is fine, I just want to allow a minimum of 1 character before the @. – Xaisoft Jul 19 '11 at 17:31
2

@Xaisoft, the one you have is **not** fine, it is incorrect. Suppose I want to use `#` in my e-mail address. – Brad Jul 19 '11 at 17:33
I found it here: http://stackoverflow.com/questions/369543/validating-e-mail-with-regular-expression-vb-net/369554#369554. I don't need to cover all bases. – Xaisoft Jul 19 '11 at 17:34
1

@Brad: it isn't _perfect_, but I think it's fine--I doubt that a non-negligible amount of people uses a "#" in their email address. – Tikhon Jelvis Jul 19 '11 at 17:35
2

@Tikhon, your doubts are wrong. I know several people who use `#` to designate different tags and such in their mailbox. It is a basic feature that is easy to allow via regex. There is little excuse for not doing so. – Brad Jul 19 '11 at 17:37
@Brad: good point. I have never heard of using `#` to do that, but I have seen people use `+`. – Tikhon Jelvis Jul 19 '11 at 18:07
Apparently, it's even possible to have a email address with 0 characters in front of it. Something like `""@example.com`. This is equivalent to having no local part and is valid. See: http://tools.ietf.org/html/rfc5321#section-4.5.3.1.3 – conradkleinespel Aug 16 '14 at 09:31

score 2 · Answer 2 · edited Oct 07 '21 at 11:04

2

0, if you don’t count quoting double-quotes. If you do, the minimum’s 1.

""@example.com

a@example.com

RFC 5321

   Mailbox        = Local-part "@" ( Domain / address-literal )

   Local-part     = Dot-string / Quoted-string
                  ; MAY be case-sensitive


   Dot-string     = Atom *("."  Atom)

   Atom           = 1*atext

   Quoted-string  = DQUOTE *QcontentSMTP DQUOTE

edited Oct 07 '21 at 11:04

Community

1
1

answered Nov 29 '19 at 05:57

Константин Ван

12,550
7
61
73

score 1 · Answer 3 · answered Jul 19 '11 at 18:17

1

Your regex is bad. You really should change it. Even Microsoft offers more flexible variant:

"^(?("")("".+?""@)|(([0-9a-zA-Z]((\.(?!\.))|[-!#\$%&'\*\+/=\?\^`\{\}\|~\w])*)(?<=[0-9a-zA-Z])@))(?(\[)(\[(\d{1,3}\.){3}\d{1,3}\])|(([0-9a-zA-Z][-\w]*[0-9a-zA-Z]\.)+[a-zA-Z]{2,6}))$"

Yes, it is long. But you should write good programs, not fast developed programs.

answered Jul 19 '11 at 18:17

VMAtm

27,943
17
79
125

VMAtm, I get your point, but you must realize that I just found regular expression online and gave it a shot. I never committed to it. – Xaisoft Jul 19 '11 at 19:14
Should definitely write fast developed programs – Dmitry Minkovsky Dec 02 '15 at 23:16

score -2 · Answer 4 · answered Jul 19 '11 at 18:14

-2

I went ahead and decided to use this one:

"^([0-9a-zA-Z]([-.\w]*[0-9a-zA-Z])*@([0-9a-zA-Z][-\w]*[0-9a-zA-Z]\.)+[a-zA-Z]{2,9})$"

It allows me to enter 1 or more characters for the local part.

answered Jul 19 '11 at 18:14

Xaisoft

45,655
87
279
432

An explanation of why this was downvoted would be nice. This is one I found on here, I believe it is from regular-expressions-info. – Xaisoft Jul 19 '11 at 19:13
What do you mean by quality in this case? If it is because it does not match certain addresses, I would not define that as quality. I would say that is more of a quantitative aspect to it. The regular expression I am looking for does not need to validate every possible address in the world. If it just had a @ symbol that would be fine. – Xaisoft Jul 19 '11 at 19:33
1

This criteria is bad criteria. That's why downvotes, I think. – VMAtm Jul 19 '11 at 19:51

Does the local part of an e-mail address have a minimum length requirement?

4 Answers4

RFC 5321