Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 0716d5f4 authored by Siyamed Sinir's avatar Siyamed Sinir
Browse files

Email address autolink regex updates. 

This CL updates the email address pattern with:
* Local-part should be max 64 characters, and domain should be max 255
  characters (it was the opposite before).
* Labels in the domain name should be at most 63 characters.
* Support non-ascii unicode characters.
* Add tests for email address pattern.

Bug:9585450
Change-Id: I983f269904ef014ef625417dd08b6509084e6879
parent f0febdcf

core/java/android/text/util/Linkify.java

+1 −1
Original line number Diff line number Diff line
@@ -224,7 +224,7 @@ public class Linkify { 
        }



        if ((mask & EMAIL_ADDRESSES) != 0) {

            gatherLinks(links, text, Patterns.EMAIL_ADDRESS,

            gatherLinks(links, text, Patterns.AUTOLINK_EMAIL_ADDRESS,

                new String[] { "mailto:" },

                null, null);

        }

core/java/android/util/Patterns.java

+30 −0
Original line number Diff line number Diff line
@@ -394,6 +394,36 @@ public class Patterns { 
    public static final Pattern AUTOLINK_WEB_URL = Pattern.compile(

            "(" + WEB_URL_WITH_PROTOCOL + "|" + WEB_URL_WITHOUT_PROTOCOL + ")");



    /**

     * Regular expression for valid email characters. Does not include some of the valid characters

     * defined in RFC5321: #&~!^`{}/=$*?|

     */

    private static final String EMAIL_CHAR = LABEL_CHAR + "\\+\\-_%'";



    /**

     * Regular expression for local part of an email address. RFC5321 section 4.5.3.1.1 limits

     * the local part to be at most 64 octets.

     */

    private static final String EMAIL_ADDRESS_LOCAL_PART =

            "[" + EMAIL_CHAR + "]" + "(?:[" + EMAIL_CHAR + "\\.]{1,62}[" + EMAIL_CHAR + "])?";



    /**

     * Regular expression for the domain part of an email address. RFC5321 section 4.5.3.1.2 limits

     * the domain to be at most 255 octets.

     */

    private static final String EMAIL_ADDRESS_DOMAIN =

            "(?=.{1,255}(?:\\s|$|^))" + HOST_NAME;



    /**

     * Regular expression pattern to match email addresses. It excludes double quoted local parts

     * and the special characters #&~!^`{}/=$*?| that are included in RFC5321.

     * @hide

     */

    public static final Pattern AUTOLINK_EMAIL_ADDRESS = Pattern.compile("(" + WORD_BOUNDARY +

            "(?:" + EMAIL_ADDRESS_LOCAL_PART + "@" + EMAIL_ADDRESS_DOMAIN + ")" +

            WORD_BOUNDARY + ")"

    );



    public static final Pattern EMAIL_ADDRESS

        = Pattern.compile(

            "[a-zA-Z0-9\\+\\.\\_\\%\\-\\+]{1,256}" +

core/tests/coretests/src/android/util/PatternsTest.java

+227 −7
Original line number Diff line number Diff line
@@ -508,6 +508,226 @@ public class PatternsTest extends TestCase { 
                Patterns.DOMAIN_NAME.matcher(domain).matches());

    }



    // Tests for Patterns.AUTOLINK_EMAIL_ADDRESS



    public void testAutoLinkEmailAddress_matchesShortValidEmail() throws Exception {

        String email = "a@a.co";

        assertTrue("Should match short valid email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesRegularEmail() throws Exception {

        String email = "email@android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesEmailWithMultipleSubdomains() throws Exception {

        String email = "email@e.somelongdomainnameforandroid.abc.uk";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartWithDot() throws Exception {

        String email = "e.mail@android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartWithPlus() throws Exception {

        String email = "e+mail@android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartWithUnderscore() throws Exception {

        String email = "e_mail@android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartWithDash() throws Exception {

        String email = "e-mail@android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartWithApostrophe() throws Exception {

        String email = "e'mail@android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartWithDigits() throws Exception {

        String email = "123@android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesUnicodeLocalPart() throws Exception {

        String email = "\uD604\uAE08\uC601\uC218\uC99D@android.kr";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartWithEmoji() throws Exception {

        String email = "smiley\u263A@android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartWithSurrogatePairs() throws Exception {

        String email = "\uD83C\uDF38@android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesDomainWithDash() throws Exception {

        String email = "email@an-droid.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesUnicodeDomain() throws Exception {

        String email = "email@\uD604\uAE08\uC601\uC218\uC99D.kr";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesUnicodeLocalPartAndDomain() throws Exception {

        String email = "\uD604\uAE08\uC601\uC218\uC99D@\uD604\uAE08\uC601\uC218\uC99D.kr";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesDomainWithEmoji() throws Exception {

        String email = "smiley@\u263Aandroid.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesDomainWithSurrogatePairs() throws Exception {

        String email = "email@\uD83C\uDF38android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartAndDomainWithSurrogatePairs()

            throws Exception {

        String email = "\uD83C\uDF38@\uD83C\uDF38android.com";

        assertTrue("Should match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchStringWithoutAtSign() throws Exception {

        String email = "android.com";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchPlainString() throws Exception {

        String email = "email";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchStringWithMultipleAtSigns() throws Exception {

        String email = "email@android@android.com";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchEmailWithoutTld() throws Exception {

        String email = "email@android";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchLocalPartEndingWithDot() throws Exception {

        String email = "email.@android.com";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchLocalPartStartingWithDot() throws Exception {

        String email = ".email@android.com";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchDomainStartingWithDash() throws Exception {

        String email = "email@-android.com";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchDomainWithConsecutiveDots() throws Exception {

        String email = "email@android..com";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchEmailWithIpAsDomain() throws Exception {

        String email = "email@127.0.0.1";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_doesNotMatchEmailWithInvalidTld() throws Exception {

        String email = "email@android.c";

        assertFalse("Should not match email: " + email,

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesLocalPartUpTo64Chars() throws Exception {

        String localPart = "";

        for (int i = 0; i < 64; i++) {

            localPart += "a";

        }

        String email = localPart + "@android.com";



        assertTrue("Should match local part of length: " + localPart.length(),

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());



        email = localPart + "a@android.com";

        assertFalse("Should not match local part of length: " + localPart.length(),

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesSubdomainUpTo63Chars() throws Exception {

        String subdomain = "";

        for (int i = 0; i < 63; i++) {

            subdomain += "a";

        }

        String email = "email@" + subdomain + ".com";



        assertTrue("Should match subdomain of length: " + subdomain.length(),

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());



        subdomain += "a";

        email = "email@" + subdomain + ".com";

        assertFalse("Should not match local part of length: " + subdomain.length(),

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    public void testAutoLinkEmailAddress_matchesDomainUpTo255Chars() throws Exception {

        String longDomain = "";

        while (longDomain.length() <= 250) {

            longDomain += "d.";

        }

        longDomain += "com";

        assertEquals(255, longDomain.length());

        String email = "a@" + longDomain;



        assertTrue("Should match domain of length: " + longDomain.length(),

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());



        email = email + "m";

        assertEquals(258, email.length());

        assertFalse("Should not match domain of length: " + longDomain.length(),

                Patterns.AUTOLINK_EMAIL_ADDRESS.matcher(email).matches());

    }



    // Tests for Patterns.PHONE



    @SmallTest