Implement allowUnicode option (defaults to true)

README.md

@@ -44,6 +44,8 @@ The `tldBlacklist` option can be either an object lookup table or an array of in
 
 The `tldWhitelist` option can be either an object lookup table or an array of valid top-level domains. If the email address has a top-level domain that is not in the whitelist, the email will be marked as invalid.
 
+The `allowUnicode` option governs whether non-ASCII characters are allowed. Defaults to `true` per RFC 6530.
+
 Only one of `tldBlacklist` and `tldWhitelist` will be consulted for TLD validity.
 
 The `minDomainAtoms` option is an optional positive integer that specifies the minimum number of domain atoms that must be included for the email address to be considered valid. Be careful with the option, as some top-level domains, like `io`, directly support email addresses.

lib/index.js

@@ -40,6 +40,11 @@ const internals = {
         cfwsComment: 17,
         cfwsFWS: 18,
 
+        // Address contains non-ASCII when the allowUnicode option is false
+        // Has to be > internals.defaultThreshold so that it's rejected
+        // without an explicit errorLevel:
+        undesiredNonAscii: 25,
+
         // Address contains deprecated elements, but may still be valid in some contexts
 
         deprecatedLocalPart: 33,
@@ -214,6 +219,7 @@ internals.validDomain = function (tldAtom, options) {
  *     addresses.
  *   {*} tldBlacklist The set of domains to consider invalid.
  *   {*} tldWhitelist The set of domains to consider valid.
+ *   {*} allowUnicode Whether to allow non-ASCII characters, defaults to true.
  *   {*} minDomainAtoms The minimum number of domain atoms which must be present
  *     for the address to be valid.
  * @param {function(number|boolean)} callback The (optional) callback handler.
@@ -276,6 +282,11 @@ exports.validate = internals.validate = function (email, options, callback) {
         }
     };
 
+    const allowUnicode = options.allowUnicode === undefined || !!options.allowUnicode;
+    if (!allowUnicode && /[^\x00-\x7f]/.test(email)) {
+        updateResult(internals.diagnoses.undesiredNonAscii);
+    }
+
     const context = {
         now: internals.components.localpart,
         prev: internals.components.localpart,

test/index.js

@@ -96,6 +96,41 @@ describe('validate()', () => {
         done();
     });
 
+    describe('with options.allowUnicode', () => {
+
+        it('should accept a pure ASCII email address when false', (done) => {
+
+            expect(Isemail.validate('[email protected]', {
+                allowUnicode: false
+            })).to.equal(true);
+            done();
+        });
+
+        it('should reject email addresses containing unicode when false', (done) => {
+
+            expect(Isemail.validate('üñïçø∂é@example.com', {
+                allowUnicode: false
+            })).to.equal(false);
+
+            expect(Isemail.validate('unicode@exãmple.com', {
+                allowUnicode: false
+            })).to.equal(false);
+            done();
+        });
+
+        describe('in combination with errorLevel', () => {
+
+            it('should return the right diagnosis when allowUnicode is false', (done) => {
+
+                expect(Isemail.validate('üñïçø∂é@example.com', {
+                    allowUnicode: false,
+                    errorLevel: 8
+                })).to.equal(25);
+                done();
+            });
+        });
+    });
+
     it('should check options.minDomainAtoms', (done) => {
 
         expect(() => {