Merge pull request #103 from validates-email-format-of/idn-support

Add support for Internationalized domain names (IDN)

Internationalised domains are converted to Punycode by default.  You can opt-out by setting `idn: false`

Thanks to https://github.com/sbilharz for the PR many years ago....

README.md

@@ -60,6 +60,7 @@ ValidatesEmailFormatOf::validate_email_format("invalid@") # => ["does not appear
 | `:message` | String | A custom error message when the email format is invalid (default is: "does not appear to be a valid email address") |
 | `:check_mx` | Boolean | Check domain for a valid MX record (default is false) |
 | `:check_mx_timeout` | Integer | Timeout in seconds for checking MX records before a `ResolvTimeout` is raised (default is 3). |
+| `:idn` | Boolean | Allowed internationalized domain names like `test@exämple.com` and `test@пример.рф`. Otherwise only domains that have already been converted to [Punycode](https://en.wikipedia.org/wiki/Punycode) are supported. (default is true) |
 | `:mx_message` | String | A custom error message when the domain does not match a valid MX record (default is: "is not routable").  Ignored unless :check_mx option is true. |
 | `:local_length` |Integer | Maximum number of characters allowed in the local part (everything before the '@') (default is 64) |
 | `:domain_length` | Integer | Maximum number of characters allowed in the domain part (everything after the '@') (default is 255) |

lib/validates_email_format_of.rb

@@ -1,4 +1,5 @@
 require "validates_email_format_of/version"
+require "simpleidn"
 
 module ValidatesEmailFormatOf
   def self.load_i18n_locales
@@ -77,7 +78,7 @@ def self.load_i18n_locales
   # > restriction on the first character is relaxed to allow either a
   # > letter or a digit.  Host software MUST support this more liberal
   # > syntax.
-  DOMAIN_PART_LABEL = /\A[A-Za-z0-9][A-Za-z0-9\-]*[A-Za-z0-9]?\Z/
+  DOMAIN_PART_LABEL = /\A[A-Za-z0-9][A-Za-z0-9-]*[A-Za-z0-9]?\Z/
 
   # From https://tools.ietf.org/id/draft-liman-tld-names-00.html#rfc.section.2
   #
@@ -92,10 +93,12 @@ def self.load_i18n_locales
   # ld       = ALPHA / DIGIT
   # ALPHA    = %x41-5A / %x61-7A   ; A-Z / a-z
   # DIGIT    = %x30-39             ; 0-9
-  DOMAIN_PART_TLD = /\A[A-Za-z][A-Za-z0-9\-]*[A-Za-z0-9]\Z/
+  DOMAIN_PART_TLD = /\A[A-Za-z][A-Za-z0-9-]*[A-Za-z0-9]\Z/
 
-  def self.validate_email_domain(email, check_mx_timeout: 3)
+  def self.validate_email_domain(email, idn: true, check_mx_timeout: 3)
     domain = email.to_s.downcase.match(/@(.+)/)[1]
+    domain = SimpleIDN.to_ascii(domain) if idn
+
     Resolv::DNS.open do |dns|
       dns.timeouts = check_mx_timeout
       @mx = dns.getresources(domain, Resolv::DNS::Resource::IN::MX) + dns.getresources(domain, Resolv::DNS::Resource::IN::A)
@@ -119,6 +122,7 @@ def self.default_message
   # * <tt>message</tt> - A custom error message (default is: "does not appear to be valid")
   # * <tt>check_mx</tt> - Check for MX records (default is false)
   # * <tt>check_mx_timeout</tt> - Timeout in seconds for checking MX records before a `ResolvTimeout` is raised (default is 3)
+  # * <tt>idn</tt> - Enable or disable Internationalized Domain Names (default is true)
   # * <tt>mx_message</tt> - A custom error message when an MX record validation fails (default is: "is not routable.")
   # * <tt>local_length</tt> Maximum number of characters allowed in the local part (default is 64)
   # * <tt>domain_length</tt> Maximum number of characters allowed in the domain part (default is 255)
@@ -127,6 +131,7 @@ def self.validate_email_format(email, options = {})
     default_options = {message: options[:generate_message] ? ERROR_MESSAGE_I18N_KEY : default_message,
                        check_mx: false,
                        check_mx_timeout: 3,
+                       idn: true,
                        mx_message: if options[:generate_message]
                                      ERROR_MX_MESSAGE_I18N_KEY
                                    else
@@ -156,10 +161,10 @@ def self.validate_email_format(email, options = {})
       deprecation_warn(":with option is deprecated and will be removed in the next version")
       return [opts[:message]] unless email&.match?(opts[:with])
     else
-      return [opts[:message]] unless validate_local_part_syntax(local) && validate_domain_part_syntax(domain)
+      return [opts[:message]] unless validate_local_part_syntax(local) && validate_domain_part_syntax(domain, idn: opts[:idn])
     end
 
-    if opts[:check_mx] && !validate_email_domain(email, check_mx_timeout: opts[:check_mx_timeout])
+    if opts[:check_mx] && !validate_email_domain(email, check_mx_timeout: opts[:check_mx_timeout], idn: opts[:idn])
       return [opts[:mx_message]]
     end
 
@@ -263,7 +268,8 @@ def self.validate_local_part_syntax(local)
     true
   end
 
-  def self.validate_domain_part_syntax(domain)
+  def self.validate_domain_part_syntax(domain, idn: true)
+    domain = SimpleIDN.to_ascii(domain) if idn
     parts = domain.downcase.split(".", -1)
 
     return false if parts.length <= 1 # Only one domain part

spec/validates_email_format_of_spec.rb

@@ -58,9 +58,15 @@ def self.model_name
       "_somename@example.com",
       # apostrophes
       "test'test@example.com",
-      # international domain names
+      # punycode domain names
       "test@xn--bcher-kva.ch",
       "test@example.xn--0zwm56d",
+
+      # IDN domains,
+      "test@exämple.com",
+      "test@пример.рф",
+      "test@почта.бел",
+
       "test@192.192.192.1",
       # Allow quoted characters.  Valid according to http://www.rfc-editor.org/errata_search.php?rfc=3696
       '"Abc\@def"@example.com',
@@ -185,6 +191,26 @@ def self.model_name
       end
     end
 
+    describe "when idn support is disabled" do
+      before(:each) do
+        allow(SimpleIDN).to receive(:to_ascii).never
+      end
+      let(:options) { {idn: false} }
+      describe "test@exämple.com" do
+        it { should have_errors_on_email.because("does not appear to be a valid email address") }
+      end
+    end
+
+    describe "when idn support is enabled" do
+      before(:each) do
+        allow(SimpleIDN).to receive(:to_ascii).once.with("exämple.com").and_return("xn--exmple-cua.com")
+      end
+      let(:options) { {idn: true} }
+      describe "test@exämple.com" do
+        it { should_not have_errors_on_email }
+      end
+    end
+
     describe "mx record" do
       domain = "example.com"
       email = "valid@#{domain}"
@@ -242,6 +268,7 @@ def self.model_name
           end
         end
       end
+
       describe "when not testing" do
         before(:each) { allow(Resolv::DNS).to receive(:open).never }
         describe "by default" do
@@ -264,6 +291,36 @@ def self.model_name
       end
     end
 
+    describe "mx record for internationalized domain" do
+      domain = "пример.рф"
+      email = "valid@#{domain}"
+
+      describe "when idn support is enabled" do
+        let(:dns) { double(Resolv::DNS) }
+        let(:options) { {check_mx: true, idn: true} }
+
+        before(:each) do
+          allow(Resolv::DNS).to receive(:open).and_yield(dns)
+          allow(dns).to receive(:"timeouts=").with(3).once
+          allow(dns).to receive(:getresources).with(SimpleIDN.to_ascii(domain), Resolv::DNS::Resource::IN::A).once.and_return([double])
+          allow(dns).to receive(:getresources).with(SimpleIDN.to_ascii(domain), Resolv::DNS::Resource::IN::MX).once.and_return([double])
+        end
+
+        describe email do
+          it { should_not have_errors_on_email }
+        end
+      end
+
+      describe "when idn support is disabled" do
+        let(:options) { {check_mx: true, idn: false} }
+
+        describe "test@пример.рф" do
+          let(:domain) { "exämple.com" }
+          it { should have_errors_on_email.because("does not appear to be a valid email address") }
+        end
+      end
+    end
+
     describe "custom regex" do
       let(:options) { {with: /[0-9]+@[0-9]+/} }
       describe "012345@789" do

validates_email_format_of.gemspec

@@ -19,6 +19,7 @@ Gem::Specification.new do |s|
     s.add_dependency "i18n", ">= 0.8.0"
   end
 
+  s.add_dependency "simpleidn"
   s.add_development_dependency "activemodel"
   s.add_development_dependency "bundler"
   s.add_development_dependency "rspec"