antonpirker/validate-email.html

## validate-email.html

<html>
<head>
<script>
    // Based on RFC 2822 section 3.2.4 / RFC 5322 section 3.2.3, these
    // characters are permitted in email addresses (not taking into
    //  account internationalization):
    var ATEXT = "a-zA-Z0-9_!#\\$%&\\'\\*\\+\\-\/=\\?\\^`\\{\\|\\}~";

    // A "dot atom text", per RFC 2822 3.2.4:
    var DOT_ATOM_TEXT = '[' + ATEXT + ']+(?:\\.[' + ATEXT + ']+)*';

    // RFC 6531 section 3.3 extends the allowed characters in internationalized
    // addresses to also include three specific ranges of UTF8 defined in
    // RFC3629 section 4, which appear to be the Unicode code points from
    // U+0080 to U+10FFFF.
    var ATEXT_UTF8 = ATEXT + '\u{0080}-\u{0010FFFF}';
    var DOT_ATOM_TEXT_UTF8 = '[' + ATEXT_UTF8 + ']+(?:\\.[' + ATEXT_UTF8 + ']+)*';

    // The domain part of the email address, after IDNA (ASCII) encoding,
    // must also satisfy the requirements of RFC 952/RFC 1123 which restrict
    // the allowed characters of hostnames further. The hyphen cannot be at
    // the beginning or end of a *dot-atom component* of a hostname either.
    var ATEXT_HOSTNAME = '(?:(?:[a-zA-Z0-9][a-zA-Z0-9\\-]*)?[a-zA-Z0-9])';


    function validateEmailLocalPart(localPart) {
        if(!localPart || localPart.length == 0) {
            throw 'There must be something before the @-sign.';
        }

        // RFC 5321 4.5.3.1.1
        if(localPart.length > 64) {
            throw 'The email address is too long before the @-sign.';
        }

        // Check the local part against the regular expression for the older ASCII requirements.
        var re = new RegExp('^' + DOT_ATOM_TEXT + '$');
        var found = null;
        found = localPart.match(re);

        if(!found) {
            var re = new RegExp('^' + DOT_ATOM_TEXT_UTF8 + '$');
            var found1 = null;
            found1 = localPart.match(re);

            if(!found1) {
                var badChars = '';
                var re1 = new RegExp('^['+ATEXT_UTF8+']');
                for(var i=0; i<localPart.length; i++) {
                    var c = localPart[i];
                    var found2 = null;
                    found2 = c.match(re1);
                    if(!found2) {
                        badChars += c;
                    }
                }

                throw 'The email address contains invalid characters before the @-sign: "' + badChars + '"';
            }
        }

        return localPart;
    }


    function validateEmailDomainPart(domainPart) {
        if(!domainPart || domainPart.length == 0) {
            throw 'There must be something after the @-sign.';
        }

        if(domainPart.endsWith('.')) {
            throw 'An email address cannot end with a period.';
        }

        if(domainPart.startsWith('.')) {
            throw 'An email address cannot have a period immediately after the @-sign.';
        }

        if(domainPart.indexOf('..') != -1) {
            throw 'An email address cannot have two periods in a row.';
        }

        if(domainPart.length > 255) {
            throw 'The email address is too long after the @-sign.';
        }

        // A "dot atom text", per RFC 2822 3.2.4, but using the restricted
        // characters allowed in a hostname (see ATEXT_HOSTNAME above).
        DOT_ATOM_TEXT_2 = ATEXT_HOSTNAME + '(?:\.' + ATEXT_HOSTNAME + ')*'

        // Check the regular expression. This is probably entirely redundant
        // with idna.decode, which also checks this format.
        var re = new RegExp('^' + DOT_ATOM_TEXT_2 + '$');
        var found = domainPart.match(re);

        if(!found) {
            throw 'The email address contains invalid characters after the @-sign.';
        }

        // All publicly deliverable addresses have domain named with at least
        // one period. We also know that all TLDs end with a letter.
        if(domainPart.indexOf('.') == -1) {
            throw 'The domain name ' + domainPart + ' is not valid. It should have a period.';
        }

        var re = new RegExp('[A-Za-z]$');
        var found = domainPart.search(re);
        if(!found) {
            throw 'The domain name ' + domainPart + ' is not valid. It is not within a valid top-level domain.';
        }

        return domainPart;
    }


    function isValidEmail(email) {
        if(!email) {
            return false;
        }
        var isValid = true;
        var error = '';

        var parts = email.split('@');

        if(parts.length != 2) {
            isValid = false;
            error = 'The email address is not valid. It must have exactly one @-sign.';
            return isValid;
        }

        var ret = [];
        ret['local'] = validateEmailLocalPart(parts[0]);
        ret['domain'] = validateEmailDomainPart(parts[1]);

        return isValid;
    }

    function testEmails() {
        var emails = [
            'anton pirker@bla.com',
            'nijole.s@hotmail.de',
            'anton@ignaz.at',
            'ööö@bbb.com',
            'bbb@ööö.com',
            'anton pirker@bla.com',
            'anton@@anton.com',
            'bla.@bla.com',
            'adsf',
            '.@.',
            '.@.com',
            'bla@.com',
            '..@bla.com',
            'bla@bla..com',
            '',
            ' ',
            '_!#$%&\'*+-/=?^`{|}~@bla.com',
            '@bla.com',
        ]

        for(var i=0; i<emails.length; i++) {
            console.log('"' + emails[i] + '"');
            var res = '';
            try {
                res = isValidEmail(emails[i]);
                if(res) {
                    res = '[true]';
                }
            } catch(ex) {
                res = '[false] ' + ex;
            }
            console.log('---> ', res);
        }
    }

    testEmails();
</script>
</head>

<body>
</body>
</html>

	<html>
	<head>
	<script>
	// Based on RFC 2822 section 3.2.4 / RFC 5322 section 3.2.3, these
	// characters are permitted in email addresses (not taking into
	// account internationalization):
	var ATEXT = "a-zA-Z0-9_!#\\$%&\\'\\*\\+\\-\/=\\?\\^`\\{\\\|\\}~";

	// A "dot atom text", per RFC 2822 3.2.4:
	var DOT_ATOM_TEXT = '[' + ATEXT + ']+(?:\\.[' + ATEXT + ']+)*';

	// RFC 6531 section 3.3 extends the allowed characters in internationalized
	// addresses to also include three specific ranges of UTF8 defined in
	// RFC3629 section 4, which appear to be the Unicode code points from
	// U+0080 to U+10FFFF.
	var ATEXT_UTF8 = ATEXT + '\u{0080}-\u{0010FFFF}';
	var DOT_ATOM_TEXT_UTF8 = '[' + ATEXT_UTF8 + ']+(?:\\.[' + ATEXT_UTF8 + ']+)*';

	// The domain part of the email address, after IDNA (ASCII) encoding,
	// must also satisfy the requirements of RFC 952/RFC 1123 which restrict
	// the allowed characters of hostnames further. The hyphen cannot be at
	// the beginning or end of a dot-atom component of a hostname either.
	var ATEXT_HOSTNAME = '(?:(?:[a-zA-Z0-9][a-zA-Z0-9\\-]*)?[a-zA-Z0-9])';


	function validateEmailLocalPart(localPart) {
	if(!localPart \|\| localPart.length == 0) {
	throw 'There must be something before the @-sign.';
	}

	// RFC 5321 4.5.3.1.1
	if(localPart.length > 64) {
	throw 'The email address is too long before the @-sign.';
	}

	// Check the local part against the regular expression for the older ASCII requirements.
	var re = new RegExp('^' + DOT_ATOM_TEXT + '$');
	var found = null;
	found = localPart.match(re);

	if(!found) {
	var re = new RegExp('^' + DOT_ATOM_TEXT_UTF8 + '$');
	var found1 = null;
	found1 = localPart.match(re);

	if(!found1) {
	var badChars = '';
	var re1 = new RegExp('^['+ATEXT_UTF8+']');
	for(var i=0; i<localPart.length; i++) {
	var c = localPart[i];
	var found2 = null;
	found2 = c.match(re1);
	if(!found2) {
	badChars += c;
	}
	}

	throw 'The email address contains invalid characters before the @-sign: "' + badChars + '"';
	}
	}

	return localPart;
	}


	function validateEmailDomainPart(domainPart) {
	if(!domainPart \|\| domainPart.length == 0) {
	throw 'There must be something after the @-sign.';
	}

	if(domainPart.endsWith('.')) {
	throw 'An email address cannot end with a period.';
	}

	if(domainPart.startsWith('.')) {
	throw 'An email address cannot have a period immediately after the @-sign.';
	}

	if(domainPart.indexOf('..') != -1) {
	throw 'An email address cannot have two periods in a row.';
	}

	if(domainPart.length > 255) {
	throw 'The email address is too long after the @-sign.';
	}

	// A "dot atom text", per RFC 2822 3.2.4, but using the restricted
	// characters allowed in a hostname (see ATEXT_HOSTNAME above).
	DOT_ATOM_TEXT_2 = ATEXT_HOSTNAME + '(?:\.' + ATEXT_HOSTNAME + ')*'

	// Check the regular expression. This is probably entirely redundant
	// with idna.decode, which also checks this format.
	var re = new RegExp('^' + DOT_ATOM_TEXT_2 + '$');
	var found = domainPart.match(re);

	if(!found) {
	throw 'The email address contains invalid characters after the @-sign.';
	}

	// All publicly deliverable addresses have domain named with at least
	// one period. We also know that all TLDs end with a letter.
	if(domainPart.indexOf('.') == -1) {
	throw 'The domain name ' + domainPart + ' is not valid. It should have a period.';
	}

	var re = new RegExp('[A-Za-z]$');
	var found = domainPart.search(re);
	if(!found) {
	throw 'The domain name ' + domainPart + ' is not valid. It is not within a valid top-level domain.';
	}

	return domainPart;
	}


	function isValidEmail(email) {
	if(!email) {
	return false;
	}
	var isValid = true;
	var error = '';

	var parts = email.split('@');

	if(parts.length != 2) {
	isValid = false;
	error = 'The email address is not valid. It must have exactly one @-sign.';
	return isValid;
	}

	var ret = [];
	ret['local'] = validateEmailLocalPart(parts[0]);
	ret['domain'] = validateEmailDomainPart(parts[1]);

	return isValid;
	}

	function testEmails() {
	var emails = [
	'anton pirker@bla.com',
	'nijole.s@hotmail.de',
	'anton@ignaz.at',
	'ööö@bbb.com',
	'bbb@ööö.com',
	'anton pirker@bla.com',
	'anton@@anton.com',
	'bla.@bla.com',
	'adsf',
	'.@.',
	'.@.com',
	'bla@.com',
	'..@bla.com',
	'bla@bla..com',
	'',
	' ',
	'_!#$%&\'*+-/=?^`{\|}~@bla.com',
	'@bla.com',
	]

	for(var i=0; i<emails.length; i++) {
	console.log('"' + emails[i] + '"');
	var res = '';
	try {
	res = isValidEmail(emails[i]);
	if(res) {
	res = '[true]';
	}
	} catch(ex) {
	res = '[false] ' + ex;
	}
	console.log('---> ', res);
	}
	}

	testEmails();
	</script>
	</head>

	<body>
	</body>
	</html>