I ported some of the code in Zend_Validator_Email:
@FacesValidator("emailValidator")
public class EmailAddressValidator implements Validator {
private String localPart;
private String hostName;
private boolean domain = true;
Locale locale;
ResourceBundle bundle;
private List<FacesMessage> messages = new ArrayList<FacesMessage>();
private HostnameValidator hostnameValidator;
@Override
public void validate(FacesContext context, UIComponent component, Object value) throws ValidatorException {
setOptions(component);
String email = (String) value;
boolean result = true;
Pattern pattern = Pattern.compile("^(.+)@([^@]+[^.])$");
Matcher matcher = pattern.matcher(email);
locale = context.getViewRoot().getLocale();
bundle = ResourceBundle.getBundle("com.myapp.resources.validationMessages", locale);
boolean length = true;
boolean local = true;
if (matcher.find()) {
localPart = matcher.group(1);
hostName = matcher.group(2);
if (localPart.length() > 64 || hostName.length() > 255) {
length = false;
addMessage("enterValidEmail", "email.AddressLengthExceeded");
}
if (domain == true) {
hostnameValidator = new HostnameValidator();
hostnameValidator.validate(context, component, hostName);
}
local = validateLocalPart();
if (local && length) {
result = true;
} else {
result = false;
}
} else {
result = false;
addMessage("enterValidEmail", "invalidEmailAddress");
}
if (result == false) {
throw new ValidatorException(messages);
}
}
private boolean validateLocalPart() {
// First try to match the local part on the common dot-atom format
boolean result = false;
// Dot-atom characters are: 1*atext *("." 1*atext)
// atext: ALPHA / DIGIT / and "!", "#", "$", "%", "&", "'", "*",
// "+", "-", "/", "=", "?", "^", "_", "`", "{", "|", "}", "~"
String atext = "a-zA-Z0-9\\u0021\\u0023\\u0024\\u0025\\u0026\\u0027\\u002a"
+ "\\u002b\\u002d\\u002f\\u003d\\u003f\\u005e\\u005f\\u0060\\u007b"
+ "\\u007c\\u007d\\u007e";
Pattern regex = Pattern.compile("^["+atext+"]+(\\u002e+["+atext+"]+)*$");
Matcher matcher = regex.matcher(localPart);
if (matcher.find()) {
result = true;
} else {
// Try quoted string format
// Quoted-string characters are: DQUOTE *([FWS] qtext/quoted-pair) [FWS] DQUOTE
// qtext: Non white space controls, and the rest of the US-ASCII characters not
// including "\" or the quote character
String noWsCtl = "\\u0001-\\u0008\\u000b\\u000c\\u000e-\\u001f\\u007f";
String qText = noWsCtl + "\\u0021\\u0023-\\u005b\\u005d-\\u007e";
String ws = "\\u0020\\u0009";
regex = Pattern.compile("^\\u0022(["+ws+qText+"])*["+ws+"]?\\u0022$");
matcher = regex.matcher(localPart);
if (matcher.find()) {
result = true;
} else {
addMessage("enterValidEmail", "email.AddressDotAtom");
addMessage("enterValidEmail", "email.AddressQuotedString");
addMessage("enterValidEmail", "email.AddressInvalidLocalPart");
}
}
return result;
}
private void addMessage(String detail, String summary) {
String detailMsg = bundle.getString(detail);
String summaryMsg = bundle.getString(summary);
messages.add(new FacesMessage(FacesMessage.SEVERITY_ERROR, summaryMsg, detailMsg));
}
private void setOptions(UIComponent component) {
Boolean domainOption = Boolean.valueOf((String) component.getAttributes().get("domain"));
//domain = (domainOption == null) ? true : domainOption.booleanValue();
}
}
With a hostname validator as follows:
@FacesValidator("hostNameValidator")
public class HostnameValidator implements Validator {
private Locale locale;
private ResourceBundle bundle;
private List<FacesMessage> messages;
private boolean checkTld = true;
private boolean allowLocal = false;
private boolean allowDNS = true;
private String tld;
private String[] validTlds = {"ac", "ad", "ae", "aero", "af", "ag", "ai",
"al", "am", "an", "ao", "aq", "ar", "arpa", "as", "asia", "at", "au",
"aw", "ax", "az", "ba", "bb", "bd", "be", "bf", "bg", "bh", "bi", "biz",
"bj", "bm", "bn", "bo", "br", "bs", "bt", "bv", "bw", "by", "bz", "ca",
"cat", "cc", "cd", "cf", "cg", "ch", "ci", "ck", "cl", "cm", "cn", "co",
"com", "coop", "cr", "cu", "cv", "cx", "cy", "cz", "de", "dj", "dk",
"dm", "do", "dz", "ec", "edu", "ee", "eg", "er", "es", "et", "eu", "fi",
"fj", "fk", "fm", "fo", "fr", "ga", "gb", "gd", "ge", "gf", "gg", "gh",
"gi", "gl", "gm", "gn", "gov", "gp", "gq", "gr", "gs", "gt", "gu", "gw",
"gy", "hk", "hm", "hn", "hr", "ht", "hu", "id", "ie", "il", "im", "in",
"info", "int", "io", "iq", "ir", "is", "it", "je", "jm", "jo", "jobs",
"jp", "ke", "kg", "kh", "ki", "km", "kn", "kp", "kr", "kw", "ky", "kz",
"la", "lb", "lc", "li", "lk", "lr", "ls", "lt", "lu", "lv", "ly", "ma",
"mc", "md", "me", "mg", "mh", "mil", "mk", "ml", "mm", "mn", "mo",
"mobi", "mp", "mq", "mr", "ms", "mt", "mu", "museum", "mv", "mw", "mx",
"my", "mz", "na", "name", "nc", "ne", "net", "nf", "ng", "ni", "nl",
"no", "np", "nr", "nu", "nz", "om", "org", "pa", "pe", "pf", "pg", "ph",
"pk", "pl", "pm", "pn", "pr", "pro", "ps", "pt", "pw", "py", "qa", "re",
"ro", "rs", "ru", "rw", "sa", "sb", "sc", "sd", "se", "sg", "sh", "si",
"sj", "sk", "sl", "sm", "sn", "so", "sr", "st", "su", "sv", "sy", "sz",
"tc", "td", "tel", "tf", "tg", "th", "tj", "tk", "tl", "tm", "tn", "to",
"tp", "tr", "travel", "tt", "tv", "tw", "tz", "ua", "ug", "uk", "um",
"us", "uy", "uz", "va", "vc", "ve", "vg", "vi", "vn", "vu", "wf", "ws",
"ye", "yt", "yu", "za", "zm", "zw"};
private Map<String, Map<Integer, Integer>> idnLength;
private void init() {
Map<Integer, Integer> biz = new HashMap<Integer, Integer>();
biz.put(5, 17);
biz.put(11, 15);
biz.put(12, 20);
Map<Integer, Integer> cn = new HashMap<Integer, Integer>();
cn.put(1, 20);
Map<Integer, Integer> com = new HashMap<Integer, Integer>();
com.put(3, 17);
com.put(5, 20);
Map<Integer, Integer> hk = new HashMap<Integer, Integer>();
hk.put(1, 15);
Map<Integer, Integer> info = new HashMap<Integer, Integer>();
info.put(4, 17);
Map<Integer, Integer> kr = new HashMap<Integer, Integer>();
kr.put(1, 17);
Map<Integer, Integer> net = new HashMap<Integer, Integer>();
net.put(3, 17);
net.put(5, 20);
Map<Integer, Integer> org = new HashMap<Integer, Integer>();
org.put(6, 17);
Map<Integer, Integer> tw = new HashMap<Integer, Integer>();
tw.put(1, 20);
Map<Integer, Integer> idn1 = new HashMap<Integer, Integer>();
idn1.put(1, 20);
Map<Integer, Integer> idn2 = new HashMap<Integer, Integer>();
idn2.put(1, 20);
Map<Integer, Integer> idn3 = new HashMap<Integer, Integer>();
idn3.put(1, 20);
Map<Integer, Integer> idn4 = new HashMap<Integer, Integer>();
idn4.put(1, 20);
idnLength = new HashMap<String, Map<Integer, Integer>>();
idnLength.put("BIZ", biz);
idnLength.put("CN", cn);
idnLength.put("COM", com);
idnLength.put("HK", hk);
idnLength.put("INFO", info);
idnLength.put("KR", kr);
idnLength.put("NET", net);
idnLength.put("ORG", org);
idnLength.put("TW", tw);
idnLength.put("ایران", idn1);
idnLength.put("中国", idn2);
idnLength.put("公司", idn3);
idnLength.put("网络", idn4);
messages = new ArrayList<FacesMessage>();
}
public HostnameValidator() {
init();
}
@Override
public void validate(FacesContext context, UIComponent component, Object value) throws ValidatorException {
String hostName = (String) value;
locale = context.getViewRoot().getLocale();
bundle = ResourceBundle.getBundle("com.myapp.resources.validationMessages", locale);
Pattern ipPattern = Pattern.compile("^[0-9a-f:\\.]*$", Pattern.CASE_INSENSITIVE);
Matcher ipMatcher = ipPattern.matcher(hostName);
if (ipMatcher.find()) {
addMessage("hostname.IpAddressNotAllowed");
throw new ValidatorException(messages);
}
boolean result = false;
// removes last dot (.) from hostname
hostName = hostName.replaceAll("(\\.)+$", "");
String[] domainParts = hostName.split("\\.");
boolean status = false;
// Check input against DNS hostname schema
if ((domainParts.length > 1) && (hostName.length() > 4) && (hostName.length() < 255)) {
status = false;
dowhile:
do {
// First check TLD
int lastIndex = domainParts.length - 1;
String domainEnding = domainParts[lastIndex];
Pattern tldRegex = Pattern.compile("([^.]{2,10})", Pattern.CASE_INSENSITIVE);
Matcher tldMatcher = tldRegex.matcher(domainEnding);
if (tldMatcher.find() || domainEnding.equals("ایران")
|| domainEnding.equals("中国")
|| domainEnding.equals("公司")
|| domainEnding.equals("网络")) {
// Hostname characters are: *(label dot)(label dot label); max 254 chars
// label: id-prefix [*ldh{61} id-prefix]; max 63 chars
// id-prefix: alpha / digit
// ldh: alpha / digit / dash
// Match TLD against known list
tld = (String) tldMatcher.group(1).toLowerCase().trim();
if (checkTld == true) {
boolean foundTld = false;
for (int i = 0; i < validTlds.length; i++) {
if (tld.equals(validTlds[i])) {
foundTld = true;
}
}
if (foundTld == false) {
status = false;
addMessage("hostname.UnknownTld");
break dowhile;
}
}
/**
* Match against IDN hostnames
* Note: Keep label regex short to avoid issues with long patterns when matching IDN hostnames
*/
List<String> regexChars = getIdnRegexChars();
// Check each hostname part
int check = 0;
for (String domainPart : domainParts) {
// Decode Punycode domainnames to IDN
if (domainPart.indexOf("xn--") == 0) {
domainPart = decodePunycode(domainPart.substring(4));
}
// Check dash (-) does not start, end or appear in 3rd and 4th positions
if (domainPart.indexOf("-") == 0
|| (domainPart.length() > 2 && domainPart.indexOf("-", 2) == 2 && domainPart.indexOf("-", 3) == 3)
|| (domainPart.indexOf("-") == (domainPart.length() - 1))) {
status = false;
addMessage("hostname.DashCharacter");
break dowhile;
}
// Check each domain part
boolean checked = false;
for (int key = 0; key < regexChars.size(); key++) {
String regexChar = regexChars.get(key);
Pattern regex = Pattern.compile(regexChar);
Matcher regexMatcher = regex.matcher(domainPart);
status = regexMatcher.find();
if (status) {
int length = 63;
if (idnLength.containsKey(tld.toUpperCase())
&& idnLength.get(tld.toUpperCase()).containsKey(key)) {
length = idnLength.get(tld.toUpperCase()).get(key);
}
int utf8Length;
try {
utf8Length = domainPart.getBytes("UTF8").length;
if (utf8Length > length) {
addMessage("hostname.InvalidHostname");
} else {
checked = true;
break;
}
} catch (UnsupportedEncodingException ex) {
Logger.getLogger(HostnameValidator.class.getName()).log(Level.SEVERE, null, ex);
}
}
}
if (checked) {
++check;
}
}
// If one of the labels doesn't match, the hostname is invalid
if (check != domainParts.length) {
status = false;
addMessage("hostname.InvalidHostnameSchema");
}
} else {
// Hostname not long enough
status = false;
addMessage("hostname.UndecipherableTld");
}
} while (false);
if (status == true && allowDNS) {
result = true;
}
} else if (allowDNS == true) {
addMessage("hostname.InvalidHostname");
throw new ValidatorException(messages);
}
// Check input against local network name schema;
Pattern regexLocal = Pattern.compile("^(([a-zA-Z0-9\\x2d]{1,63}\\x2e)*[a-zA-Z0-9\\x2d]{1,63}){1,254}$", Pattern.CASE_INSENSITIVE);
boolean checkLocal = regexLocal.matcher(hostName).find();
if (allowLocal && !status) {
if (checkLocal) {
result = true;
} else {
// If the input does not pass as a local network name, add a message
result = false;
addMessage("hostname.InvalidLocalName");
}
}
// If local network names are not allowed, add a message
if (checkLocal && !allowLocal && !status) {
result = false;
addMessage("hostname.LocalNameNotAllowed");
}
if (result == false) {
throw new ValidatorException(messages);
}
}
private void addMessage(String msg) {
String bundlMsg = bundle.getString(msg);
messages.add(new FacesMessage(FacesMessage.SEVERITY_ERROR, bundlMsg, bundlMsg));
}
/**
* Returns a list of regex patterns for the matched TLD
* @param tld
* @return
*/
private List<String> getIdnRegexChars() {
List<String> regexChars = new ArrayList<String>();
regexChars.add("^[a-z0-9\\x2d]{1,63}$");
Document doc = null;
DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
factory.setNamespaceAware(true);
try {
InputStream validIdns = getClass().getClassLoader().getResourceAsStream("com/myapp/resources/validIDNs_1.xml");
DocumentBuilder builder = factory.newDocumentBuilder();
doc = builder.parse(validIdns);
doc.getDocumentElement().normalize();
} catch (SAXException ex) {
Logger.getLogger(HostnameValidator.class.getName()).log(Level.SEVERE, null, ex);
} catch (IOException ex) {
Logger.getLogger(HostnameValidator.class.getName()).log(Level.SEVERE, null, ex);
} catch (ParserConfigurationException ex) {
Logger.getLogger(HostnameValidator.class.getName()).log(Level.SEVERE, null, ex);
}
// prepare XPath
XPath xpath = XPathFactory.newInstance().newXPath();
NodeList nodes = null;
String xpathRoute = "//idn[tld=\'" + tld.toUpperCase() + "\']/pattern/text()";
try {
XPathExpression expr;
expr = xpath.compile(xpathRoute);
Object res = expr.evaluate(doc, XPathConstants.NODESET);
nodes = (NodeList) res;
} catch (XPathExpressionException ex) {
Logger.getLogger(HostnameValidator.class.getName()).log(Level.SEVERE, null, ex);
}
for (int i = 0; i < nodes.getLength(); i++) {
regexChars.add(nodes.item(i).getNodeValue());
}
return regexChars;
}
/**
* Decode Punycode string
* @param encoded
* @return
*/
private String decodePunycode(String encoded) {
Pattern regex = Pattern.compile("([^a-z0-9\\x2d]{1,10})", Pattern.CASE_INSENSITIVE);
Matcher matcher = regex.matcher(encoded);
boolean found = matcher.find();
if (encoded.isEmpty() || found) {
// no punycode encoded string, return as is
addMessage("hostname.CannotDecodePunycode");
throw new ValidatorException(messages);
}
int separator = encoded.lastIndexOf("-");
List<Integer> decoded = new ArrayList<Integer>();
if (separator > 0) {
for (int x = 0; x < separator; ++x) {
decoded.add((int) encoded.charAt(x));
}
} else {
addMessage("hostname.CannotDecodePunycode");
throw new ValidatorException(messages);
}
int lengthd = decoded.size();
int lengthe = encoded.length();
// decoding
boolean init = true;
int base = 72;
int index = 0;
int ch = 0x80;
int indexeStart = (separator == 1) ? (separator + 1) : 0;
for (int indexe = indexeStart; indexe < lengthe; ++lengthd) {
int oldIndex = index;
int pos = 1;
for (int key = 36; true; key += 36) {
int hex = (int) encoded.charAt(indexe++);
int digit = (hex - 48 < 10) ? hex - 22
: ((hex - 65 < 26) ? hex - 65
: ((hex - 97 < 26) ? hex - 97
: 36));
index += digit * pos;
int tag = (key <= base) ? 1 : ((key >= base + 26) ? 26 : (key - base));
if (digit < tag) {
break;
}
pos = (int) (pos * (36 - tag));
}
int delta = (int) (init ? ((index - oldIndex) / 700) : ((index - oldIndex) / 2));
delta += (int) (delta / (lengthd + 1));
int key;
for (key = 0; delta > 910; key += 36) {
delta = (int) (delta / 35);
}
base = (int) (key + 36 * delta / (delta + 38));
init = false;
ch += (int) (index / (lengthd + 1));
index %= (lengthd + 1);
if (lengthd > 0) {
for (int i = lengthd; i > index; i--) {
decoded.set(i, decoded.get(i - 1));
}
}
decoded.set(index++, ch);
}
// convert decoded ucs4 to utf8 string
StringBuilder sb = new StringBuilder();
for (int i = 0; i < decoded.size(); i++) {
int value = decoded.get(i);
if (value < 128) {
sb.append((char) value);
} else if (value < (1 << 11)) {
sb.append((char) (192 + (value >> 6)));
sb.append((char) (128 + (value & 63)));
} else if (value < (1 << 16)) {
sb.append((char) (224 + (value >> 12)));
sb.append((char) (128 + ((value >> 6) & 63)));
sb.append((char) (128 + (value & 63)));
} else if (value < (1 << 21)) {
sb.append((char) (240 + (value >> 18)));
sb.append((char) (128 + ((value >> 12) & 63)));
sb.append((char) (128 + ((value >> 6) & 63)));
sb.append((char) (128 + (value & 63)));
} else {
addMessage("hostname.CannotDecodePunycode");
throw new ValidatorException(messages);
}
}
return sb.toString();
}
/**
* Eliminates empty values from input array
* @param data
* @return
*/
private String[] verifyArray(String[] data) {
List<String> result = new ArrayList<String>();
for (String s : data) {
if (!s.equals("")) {
result.add(s);
}
}
return result.toArray(new String[result.size()]);
}
}
And a validIDNs.xml with regex patterns for the different tlds (too big to include:)
<idnlist>
<idn>
<tld>AC</tld>
<pattern>^[\u002d0-9a-zà-öø-ÿāăąćĉċčďđēėęěĝġģĥħīįĵķĺļľŀłńņňŋőœŕŗřśŝşšţťŧūŭůűųŵŷźżž]{1,63}$</pattern>
</idn>
<idn>
<tld>AR</tld>
<pattern>^[\u002d0-9a-zà-ãç-êìíñ-õü]{1,63}$</pattern>
</idn>
<idn>
<tld>AS</tld>
<pattern>/^[\u002d0-9a-zà-öø-ÿāăąćĉċčďđēĕėęěĝğġģĥħĩīĭįıĵķĸĺļľłńņňŋōŏőœŕŗřśŝşšţťŧũūŭůűųŵŷźż]{1,63}$</pattern>
</idn>
<idn>
<tld>AT</tld>
<pattern>/^[\u002d0-9a-zà-öø-ÿœšž]{1,63}$</pattern>
</idn>
<idn>
<tld>BIZ</tld>
<pattern>^[\u002d0-9a-zäåæéöøü]{1,63}$</pattern>
<pattern>^[\u002d0-9a-záéíñóúü]{1,63}$</pattern>
<pattern>^[\u002d0-9a-záéíóöúüőű]{1,63}$</pattern>
</id>
</idlist>
