Fix labels length check (#278)

* Fix label length check

* Test label too long on single word hostname

.github/FUNDING.yml

@@ -0,0 +1,3 @@
+# Thanks!
+
+github: [egulias]

.gitignore

@@ -1,4 +1,6 @@
+.phpunit.result.cache
 .idea
+.vscode
 composer.lock
 report/
 vendor/

.travis.yml

@@ -1,5 +1,4 @@
 language: php
-
 matrix:
   include:
     - php: 5.5.9
@@ -28,6 +27,29 @@ matrix:
       env:
         - psalm=yes
       dist: bionic
+#ppc64le support code
+    - php: 5.6
+      arch: ppc64le
+      env:
+      dist: xenial
+    - php: 7.0
+      arch: ppc64le
+      dist: xenial
+    - php: 7.1
+      arch: ppc64le
+      env:
+        - psalm=yes
+      dist: bionic
+    - php: 7.2
+      arch: ppc64le
+      env:
+        - psalm=yes
+      dist: bionic
+    - php: 7.4
+      arch: ppc64le
+      env:
+        - psalm=yes
+      dist: bionic
 
 install:
   - if [ "$deps" = "low" ]; then composer update --prefer-lowest; else composer install; fi

README.md

@@ -42,16 +42,14 @@ $validator->isValid("example@example.com", new RFCValidation()); //true
 
 ### Available validations ###
 
-1. [RFCValidation](/src/Validation/RFCValidation.php)
+1. [RFCValidation](/src/Validation/RFCValidation.php): Standard RFC-like email validation.
-2. [NoRFCWarningsValidation](/src/Validation/NoRFCWarningsValidation.php)
+2. [NoRFCWarningsValidation](/src/Validation/NoRFCWarningsValidation.php): RFC-like validation that will fail when warnings* are found.
-3. [DNSCheckValidation](/src/Validation/DNSCheckValidation.php)
+3. [DNSCheckValidation](/src/Validation/DNSCheckValidation.php): Will check if there are DNS records that signal that the server accepts emails. This does not entails that the email exists.
-4. [SpoofCheckValidation](/src/Validation/SpoofCheckValidation.php)
+4. [SpoofCheckValidation](/src/Validation/SpoofCheckValidation.php): Will check for multi-utf-8 chars that can signal an erroneous email name.
-5. [MultipleValidationWithAnd](/src/Validation/MultipleValidationWithAnd.php)
+5. [MultipleValidationWithAnd](/src/Validation/MultipleValidationWithAnd.php): It is a validation that operates over other validations performing a logical and (&&) over the result of each validation.
-6. [Your own validation](#how-to-extend)
+6. [Your own validation](#how-to-extend): You can extend the library behaviour by implementing your own validations.
 
-`MultipleValidationWithAnd`
+*warnings: Warnings are deviations from the RFC that in a broader interpretation are acceptded.
-
-It is a validation that operates over other validations performing a logical and (&&) over the result of each validation.
 
 ```php
 <?php
@@ -66,7 +64,8 @@ $multipleValidations = new MultipleValidationWithAnd([
     new RFCValidation(),
     new DNSCheckValidation()
 ]);
-$validator->isValid("example@example.com", $multipleValidations); //true
+//ietf.org has MX records signaling a server with email capabilites
+$validator->isValid("example@ietf.org", $multipleValidations); //true
 ```
 
 ### How to extend ###

src/EmailLexer.php

@@ -14,6 +14,7 @@ class EmailLexer extends AbstractLexer
     const S_DOT              = 46;
     const S_DQUOTE           = 34;
     const S_SQUOTE           = 39;
+    const S_BACKTICK         = 96;
     const S_OPENPARENTHESIS  = 49;
     const S_CLOSEPARENTHESIS = 261;
     const S_OPENBRACKET      = 262;
@@ -60,6 +61,7 @@ class EmailLexer extends AbstractLexer
         ','    => self::S_COMMA,
         '.'    => self::S_DOT,
         "'"    => self::S_SQUOTE,
+        "`"    => self::S_BACKTICK,
         '"'    => self::S_DQUOTE,
         '-'    => self::S_HYPHEN,
         '::'   => self::S_DOUBLECOLON,

src/Parser/DomainPart.php

@@ -35,6 +35,7 @@ use Egulias\EmailValidator\Warning\TLD;
 class DomainPart extends Parser
 {
     const DOMAIN_MAX_LENGTH = 254;
+    const LABEL_MAX_LENGTH = 63;
 
     /**
      * @var string
@@ -160,6 +161,7 @@ class DomainPart extends Parser
     protected function doParseDomainPart()
     {
         $domain = '';
+        $label = '';
         $openedParenthesis = 0;
         do {
             $prev = $this->lexer->getPrevious();
@@ -190,7 +192,12 @@ class DomainPart extends Parser
                 $this->parseDomainLiteral();
             }
 
-            $this->checkLabelLength($prev);
+            if ($this->lexer->token['type'] === EmailLexer::S_DOT) {
+                $this->checkLabelLength($label);
+                $label = '';
+            } else {
+                $label .= $this->lexer->token['value'];
+            }
 
             if ($this->isFWS()) {
                 $this->parseFWS();
@@ -203,6 +210,8 @@ class DomainPart extends Parser
             }
         } while (null !== $this->lexer->token['type']);
 
+        $this->checkLabelLength($label);
+
         return $domain;
     }
 
@@ -336,6 +345,7 @@ class DomainPart extends Parser
         $invalidDomainTokens = array(
             EmailLexer::S_DQUOTE => true,
             EmailLexer::S_SQUOTE => true,
+            EmailLexer::S_BACKTICK => true,
             EmailLexer::S_SEMICOLON => true,
             EmailLexer::S_GREATERTHAN => true,
             EmailLexer::S_LOWERTHAN => true,
@@ -385,16 +395,31 @@ class DomainPart extends Parser
         return true;
     }
 
-    protected function checkLabelLength(array $prev)
+    /**
+     * @param string $label
+     */
+    protected function checkLabelLength($label)
     {
-        if ($this->lexer->token['type'] === EmailLexer::S_DOT &&
+        if ($this->isLabelTooLong($label)) {
-            $prev['type'] === EmailLexer::GENERIC &&
-            strlen($prev['value']) > 63
-        ) {
             $this->warnings[LabelTooLong::CODE] = new LabelTooLong();
         }
     }
 
+    /**
+     * @param string $label
+     * @return bool
+     */
+    private function isLabelTooLong($label)
+    {
+        if (preg_match('/[^\x00-\x7F]/', $label)) {
+            idn_to_ascii($label, IDNA_DEFAULT, INTL_IDNA_VARIANT_UTS46, $idnaInfo);
+
+            return (bool) ($idnaInfo['errors'] & IDNA_ERROR_LABEL_TOO_LONG);
+        }
+
+        return strlen($label) > self::LABEL_MAX_LENGTH;
+    }
+
     protected function parseDomainComments()
     {
         $this->isUnclosedComment();

src/Warning/Warning.php

@@ -29,7 +29,7 @@ abstract class Warning
      */
     public function code()
     {
-        return self::CODE;
+        return static::CODE;
     }
 
     /**

tests/EmailValidator/EmailLexerTest.php

@@ -139,6 +139,7 @@ class EmailLexerTests extends TestCase
             array(".", EmailLexer::S_DOT),
             array("\"", EmailLexer::S_DQUOTE),
             array("'", EmailLexer::S_SQUOTE),
+            array("`", EmailLexer::S_BACKTICK),
             array("-", EmailLexer::S_HYPHEN),
             array("\\", EmailLexer::S_BACKSLASH),
             array("/", EmailLexer::S_SLASH),

tests/EmailValidator/Validation/NoRFCWarningsValidationTest.php

@@ -26,11 +26,22 @@ class NoRFCWarningsValidationTest extends TestCase
         $this->assertInstanceOf(RFCWarnings::class, $validation->getError());
     }
 
-    public function testEmailWithoutWarningsIsValid()
+    /**
+     * @dataProvider getValidEmailsWithoutWarnings
+     */
+    public function testEmailWithoutWarningsIsValid($email)
     {
         $validation = new NoRFCWarningsValidation();
 
-        $this->assertTrue($validation->isValid('example@example.com', new EmailLexer()));
+        $this->assertTrue($validation->isValid($email, new EmailLexer()));
         $this->assertNull($validation->getError());
     }
+
+    public function getValidEmailsWithoutWarnings()
+    {
+        return [
+            ['example@example.com',],
+            [sprintf('example@%s.com', str_repeat('ъ', 40)),],
+        ];
+    }
 }

tests/EmailValidator/Validation/RFCValidationTest.php

@@ -24,6 +24,7 @@ use Egulias\EmailValidator\Warning\CFWSWithFWS;
 use Egulias\EmailValidator\Warning\Comment;
 use Egulias\EmailValidator\Warning\DomainLiteral;
 use Egulias\EmailValidator\Warning\DomainTooLong;
+use Egulias\EmailValidator\Warning\EmailTooLong;
 use Egulias\EmailValidator\Warning\IPV6BadChar;
 use Egulias\EmailValidator\Warning\IPV6ColonEnd;
 use Egulias\EmailValidator\Warning\IPV6ColonStart;
@@ -169,6 +170,7 @@ class RFCValidationTest extends TestCase
             ['test@ '],
             ['test@example.com []'],
             ["test@example.com'"],
+            ["test@exam`ple.com"],
         ];
     }
 
@@ -225,12 +227,12 @@ class RFCValidationTest extends TestCase
         $this->assertTrue($this->validator->isValid($email, $this->lexer));
         $warnings = $this->validator->getWarnings();
         $this->assertCount(
-            count($warnings), $expectedWarnings,
+            count($expectedWarnings), $warnings,
             "Expected: " . implode(",", $expectedWarnings) . " and got " . implode(",", $warnings)
         );
 
         foreach ($warnings as $warning) {
-            $this->assertArrayHasKey($warning->code(), $expectedWarnings);
+            $this->assertContains($warning->code(), $expectedWarnings);
         }
     }
 
@@ -277,17 +279,29 @@ class RFCValidationTest extends TestCase
                 'example@toolonglocalparttoolonglocalparttoolonglocalparttoolonglocalpart.co.uk'
             ],
             [
-                [DomainTooLong::CODE, LabelTooLong::CODE,],
+                [DomainTooLong::CODE, LabelTooLong::CODE, EmailTooLong::CODE],
                 'example2@toolonglocalparttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocal'.
                 'parttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocalpart'.
                 'toolonglocalparttoolonglocalparttoolonglocalparttoolonglocalpart'
             ],
             [
-                [DomainTooLong::CODE, LabelTooLong::CODE,],
+                [DomainTooLong::CODE, LabelTooLong::CODE, EmailTooLong::CODE],
                 'example@toolonglocalparttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocal'.
                 'parttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocalparttoolonglocalpart'.
                 'toolonglocalparttoolonglocalparttoolonglocalparttoolonglocalpar'
             ],
+            [
+                [LabelTooLong::CODE,],
+                sprintf('example@%s.com', str_repeat('ъ', 60)),
+            ],
+            [
+                [LabelTooLong::CODE,],
+                sprintf('example@%s.com', str_repeat('a4t', 22)),
+            ],
+            [
+                [LabelTooLong::CODE,],
+                sprintf('example@%s', str_repeat('a4t', 22)),
+            ],
         ];
     }
 }