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

Commit db990751 authored by Shimeng (Simon) Wang's avatar Shimeng (Simon) Wang
Browse files

Updated URL pattern based on http://data.iana.org/TLD/tlds-alpha-by-domain.txt 

Note: this change doesn't solve the Arabic country code top level domains, which
was turned on today.  For that, more changes need to be done; probably we need a way
to convert the BiDi url to punycode first; then apply this URL pattern.

issue: 2663885
Change-Id: I0597fe620c4d44259bff88ecfcbcca1565928c6f
parent e1acf111

api/9.xml

+2 −2
Original line number Diff line number Diff line
@@ -164722,7 +164722,7 @@ 
 type="java.lang.String"

 transient="false"

 volatile="false"

 value=""((aero|arpa|asia|a[cdefgilmnoqrstuwxz])|(biz|b[abdefghijmnorstvwyz])|(cat|com|coop|c[acdfghiklmnoruvxyz])|d[ejkmoz]|(edu|e[cegrstu])|f[ijkmor]|(gov|g[abdefghilmnpqrstuwy])|h[kmnrtu]|(info|int|i[delmnoqrst])|(jobs|j[emop])|k[eghimnprwyz]|l[abcikrstuvy]|(mil|mobi|museum|m[acdeghklmnopqrstuvwxyz])|(name|net|n[acefgilopruz])|(org|om)|(pro|p[aefghklmnrstwy])|qa|r[eosuw]|s[abcdeghijklmnortuvyz]|(tel|travel|t[cdfghjklmnoprtvwz])|u[agksyz]|v[aceginu]|w[fs]|(xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-zckzah)|y[etu]|z[amw])""

 value=""((aero|arpa|asia|a[cdefgilmnoqrstuwxz])|(biz|b[abdefghijmnorstvwyz])|(cat|com|coop|c[acdfghiklmnoruvxyz])|d[ejkmoz]|(edu|e[cegrstu])|f[ijkmor]|(gov|g[abdefghilmnpqrstuwy])|h[kmnrtu]|(info|int|i[delmnoqrst])|(jobs|j[emop])|k[eghimnprwyz]|l[abcikrstuvy]|(mil|mobi|museum|m[acdeghklmnopqrstuvwxyz])|(name|net|n[acefgilopruz])|(org|om)|(pro|p[aefghklmnrstwy])|qa|r[eosuw]|s[abcdeghijklmnortuvyz]|(tel|travel|t[cdfghjklmnoprtvwz])|u[agksyz]|v[aceginu]|w[fs]|(xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-mgbaam7a8h|xn\\-\\-mgberp4a5d4ar|xn\\-\\-wgbh1c|xn\\-\\-zckzah)|y[et]|z[amw])""

 static="true"

 final="true"

 deprecated="not deprecated"
@@ -164733,7 +164733,7 @@ 
 type="java.lang.String"

 transient="false"

 volatile="false"

 value=""(?:(?:aero|arpa|asia|a[cdefgilmnoqrstuwxz])|(?:biz|b[abdefghijmnorstvwyz])|(?:cat|com|coop|c[acdfghiklmnoruvxyz])|d[ejkmoz]|(?:edu|e[cegrstu])|f[ijkmor]|(?:gov|g[abdefghilmnpqrstuwy])|h[kmnrtu]|(?:info|int|i[delmnoqrst])|(?:jobs|j[emop])|k[eghimnprwyz]|l[abcikrstuvy]|(?:mil|mobi|museum|m[acdeghklmnopqrstuvwxyz])|(?:name|net|n[acefgilopruz])|(?:org|om)|(?:pro|p[aefghklmnrstwy])|qa|r[eosuw]|s[abcdeghijklmnortuvyz]|(?:tel|travel|t[cdfghjklmnoprtvwz])|u[agksyz]|v[aceginu]|w[fs]|(?:xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-zckzah)|y[etu]|z[amw]))""

 value=""(?:(?:aero|arpa|asia|a[cdefgilmnoqrstuwxz])|(?:biz|b[abdefghijmnorstvwyz])|(?:cat|com|coop|c[acdfghiklmnoruvxyz])|d[ejkmoz]|(?:edu|e[cegrstu])|f[ijkmor]|(?:gov|g[abdefghilmnpqrstuwy])|h[kmnrtu]|(?:info|int|i[delmnoqrst])|(?:jobs|j[emop])|k[eghimnprwyz]|l[abcikrstuvy]|(?:mil|mobi|museum|m[acdeghklmnopqrstuvwxyz])|(?:name|net|n[acefgilopruz])|(?:org|om)|(?:pro|p[aefghklmnrstwy])|qa|r[eosuw]|s[abcdeghijklmnortuvyz]|(?:tel|travel|t[cdfghjklmnoprtvwz])|u[agksyz]|v[aceginu]|w[fs]|(?:xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-mgbaam7a8h|xn\\-\\-mgberp4a5d4ar|xn\\-\\-wgbh1c|xn\\-\\-zckzah)|y[et]|z[amw]))""

 static="true"

 final="true"

 deprecated="not deprecated"

api/current.xml

+2 −2
Original line number Diff line number Diff line
@@ -169810,7 +169810,7 @@ 
 type="java.lang.String"

 transient="false"

 volatile="false"

 value=""((aero|arpa|asia|a[cdefgilmnoqrstuwxz])|(biz|b[abdefghijmnorstvwyz])|(cat|com|coop|c[acdfghiklmnoruvxyz])|d[ejkmoz]|(edu|e[cegrstu])|f[ijkmor]|(gov|g[abdefghilmnpqrstuwy])|h[kmnrtu]|(info|int|i[delmnoqrst])|(jobs|j[emop])|k[eghimnprwyz]|l[abcikrstuvy]|(mil|mobi|museum|m[acdeghklmnopqrstuvwxyz])|(name|net|n[acefgilopruz])|(org|om)|(pro|p[aefghklmnrstwy])|qa|r[eosuw]|s[abcdeghijklmnortuvyz]|(tel|travel|t[cdfghjklmnoprtvwz])|u[agksyz]|v[aceginu]|w[fs]|(xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-zckzah)|y[etu]|z[amw])""

 value=""((aero|arpa|asia|a[cdefgilmnoqrstuwxz])|(biz|b[abdefghijmnorstvwyz])|(cat|com|coop|c[acdfghiklmnoruvxyz])|d[ejkmoz]|(edu|e[cegrstu])|f[ijkmor]|(gov|g[abdefghilmnpqrstuwy])|h[kmnrtu]|(info|int|i[delmnoqrst])|(jobs|j[emop])|k[eghimnprwyz]|l[abcikrstuvy]|(mil|mobi|museum|m[acdeghklmnopqrstuvwxyz])|(name|net|n[acefgilopruz])|(org|om)|(pro|p[aefghklmnrstwy])|qa|r[eosuw]|s[abcdeghijklmnortuvyz]|(tel|travel|t[cdfghjklmnoprtvwz])|u[agksyz]|v[aceginu]|w[fs]|(xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-mgbaam7a8h|xn\\-\\-mgberp4a5d4ar|xn\\-\\-wgbh1c|xn\\-\\-zckzah)|y[et]|z[amw])""

 static="true"

 final="true"

 deprecated="not deprecated"
@@ -169821,7 +169821,7 @@ 
 type="java.lang.String"

 transient="false"

 volatile="false"

 value=""(?:(?:aero|arpa|asia|a[cdefgilmnoqrstuwxz])|(?:biz|b[abdefghijmnorstvwyz])|(?:cat|com|coop|c[acdfghiklmnoruvxyz])|d[ejkmoz]|(?:edu|e[cegrstu])|f[ijkmor]|(?:gov|g[abdefghilmnpqrstuwy])|h[kmnrtu]|(?:info|int|i[delmnoqrst])|(?:jobs|j[emop])|k[eghimnprwyz]|l[abcikrstuvy]|(?:mil|mobi|museum|m[acdeghklmnopqrstuvwxyz])|(?:name|net|n[acefgilopruz])|(?:org|om)|(?:pro|p[aefghklmnrstwy])|qa|r[eosuw]|s[abcdeghijklmnortuvyz]|(?:tel|travel|t[cdfghjklmnoprtvwz])|u[agksyz]|v[aceginu]|w[fs]|(?:xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-zckzah)|y[etu]|z[amw]))""

 value=""(?:(?:aero|arpa|asia|a[cdefgilmnoqrstuwxz])|(?:biz|b[abdefghijmnorstvwyz])|(?:cat|com|coop|c[acdfghiklmnoruvxyz])|d[ejkmoz]|(?:edu|e[cegrstu])|f[ijkmor]|(?:gov|g[abdefghilmnpqrstuwy])|h[kmnrtu]|(?:info|int|i[delmnoqrst])|(?:jobs|j[emop])|k[eghimnprwyz]|l[abcikrstuvy]|(?:mil|mobi|museum|m[acdeghklmnopqrstuvwxyz])|(?:name|net|n[acefgilopruz])|(?:org|om)|(?:pro|p[aefghklmnrstwy])|qa|r[eosuw]|s[abcdeghijklmnortuvyz]|(?:tel|travel|t[cdfghjklmnoprtvwz])|u[agksyz]|v[aceginu]|w[fs]|(?:xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-mgbaam7a8h|xn\\-\\-mgberp4a5d4ar|xn\\-\\-wgbh1c|xn\\-\\-zckzah)|y[et]|z[amw]))""

 static="true"

 final="true"

 deprecated="not deprecated"

core/java/android/util/Patterns.java

+6 −6
Original line number Diff line number Diff line
@@ -25,7 +25,7 @@ import java.util.regex.Pattern; 
public class Patterns {

    /**

     *  Regular expression to match all IANA top-level domains.

     *  List accurate as of 2010/02/05.  List taken from:

     *  List accurate as of 2010/05/06.  List taken from:

     *  http://data.iana.org/TLD/tlds-alpha-by-domain.txt

     *  This pattern is auto-generated by frameworks/base/common/tools/make-iana-tld-pattern.py

     */
@@ -53,8 +53,8 @@ public class Patterns { 
        + "|u[agksyz]"

        + "|v[aceginu]"

        + "|w[fs]"

        + "|(xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-zckzah)"

        + "|y[etu]"

        + "|(xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-mgbaam7a8h|xn\\-\\-mgberp4a5d4ar|xn\\-\\-wgbh1c|xn\\-\\-zckzah)"

        + "|y[et]"

        + "|z[amw])";



    /**
@@ -65,7 +65,7 @@ public class Patterns { 


    /**

     *  Regular expression to match all IANA top-level domains for WEB_URL.

     *  List accurate as of 2010/02/05.  List taken from:

     *  List accurate as of 2010/05/06.  List taken from:

     *  http://data.iana.org/TLD/tlds-alpha-by-domain.txt

     *  This pattern is auto-generated by frameworks/base/common/tools/make-iana-tld-pattern.py

     */
@@ -94,8 +94,8 @@ public class Patterns { 
        + "|u[agksyz]"

        + "|v[aceginu]"

        + "|w[fs]"

        + "|(?:xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-zckzah)"

        + "|y[etu]"

        + "|(?:xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-80akhbyknj4f|xn\\-\\-9t4b11yi5a|xn\\-\\-deba0ad|xn\\-\\-g6w251d|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-mgbaam7a8h|xn\\-\\-mgberp4a5d4ar|xn\\-\\-wgbh1c|xn\\-\\-zckzah)"

        + "|y[et]"

        + "|z[amw]))";



    /**

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

+16 −1
Original line number Diff line number Diff line
@@ -68,6 +68,13 @@ public class PatternsTest extends TestCase { 
        t = Patterns.WEB_URL.matcher("xn--fsqu00a.xn--0zwm56d").matches();

        assertTrue("Valid URL", t);



        // Url for testing top level Arabic country code domain in Punycode:

        //   http://xn--4gbrim.xn----rmckbbajlc6dj7bxne2c.xn--wgbh1c/ar/default.aspx

        t = Patterns.WEB_URL.matcher("http://xn--4gbrim.xn----rmckbbajlc6dj7bxne2c.xn--wgbh1c/ar/default.aspx").matches();

        assertTrue("Valid URL", t);

        t = Patterns.WEB_URL.matcher("xn--4gbrim.xn----rmckbbajlc6dj7bxne2c.xn--wgbh1c/ar/default.aspx").matches();

        assertTrue("Valid URL", t);



        // Internationalized URL.

        t = Patterns.WEB_URL.matcher("http://\uD604\uAE08\uC601\uC218\uC99D.kr").matches();

        assertTrue("Valid URL", t);
@@ -109,7 +116,7 @@ public class PatternsTest extends TestCase { 
        t = Patterns.DOMAIN_NAME.matcher("mail.example.com").matches();

        assertTrue("Valid domain", t);



        t = Patterns.WEB_URL.matcher("google.me").matches();

        t = Patterns.DOMAIN_NAME.matcher("google.me").matches();

        assertTrue("Valid domain", t);



        // Internationalized domains.
@@ -118,6 +125,14 @@ public class PatternsTest extends TestCase { 


        t = Patterns.DOMAIN_NAME.matcher("__+&42.xer").matches();

        assertFalse("Invalid domain", t);



        // Obsolete domain .yu

        t = Patterns.DOMAIN_NAME.matcher("test.yu").matches();

        assertFalse("Obsolete country code top level domain", t);



        // Testing top level Arabic country code domain in Punycode:

        t = Patterns.DOMAIN_NAME.matcher("xn--4gbrim.xn----rmckbbajlc6dj7bxne2c.xn--wgbh1c").matches();

        assertTrue("Valid domain", t);

    }



    @SmallTest