From db9c4cb3cfbf52a79f89e801f461ad17fbdb1975 Mon Sep 17 00:00:00 2001
From: Brad Jorsch <bjorsch@wikimedia.org>
Date: Wed, 23 Oct 2013 14:06:19 -0400
Subject: [PATCH] Remove linkprefix message, add $linkPrefixCharset
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

The existing "linkprefix" message is unlikely to be accurately
customized by message translators (as shown by the fact that, of the 10
distinct customizations prior to Iaa7eaa44 (which made them even more
complicated), 3 were broken or entirely ineffective, 1 was half
ineffective, and 2 more seem to have included the Latin-1 Supplement by
accident) or by local wiki admins. So, like linktrail before it, let's
move it out of the system messages and into a separate language
variable.

At the same time, let's make it a simple character set (like
$wgLegalTitleChars) rather than a complicated regular expression. The
complicated regex now lives in the parser.

This also adjusts the output of the API's action=query&meta=siteinfo and
adds an accessor parallel to the linkTrail accessor to Language.

Note the following changes that are not simply extracting the existing
charset from the linkprefix message for $linkPrefixCharset:
* The En message matched all non-ASCII UTF-8 characters by matching the
  component bytes (\\x80-\\xff). The new character set is equivalent.
* Various languages were identical to En and so have no $linkPrefixCharset
  set. These are: Ary Az Ce Ga Id Ka Kiu Km Ltg Mk Ms Ne Nn Ro Roa_tara Sc Si
  Sr_ec Sr_el Tl Tt_cyrl Tt_latn Ug_arab War
* Cu, Uk, and Udm are changed to match any number of „ or « in the prefix.
* Cv tried to include "«" that was redundant to the range \\x80-\\xff
  (see En comment). This was removed.
* Diq was entirely bogus, and so was removed.
* Gu included many additional UTF-8 characters that are redundant to the
  range \\x80-\\xff (see En comment). These were removed, and the
  resulting character set is equivalent to En.
* Mt has been broken since it was introduced in r37242. The charset used is
  equivalent to the broken regex.

Bug: 56031
Change-Id: I3369851b33113fc118a1bace38f3ac310cdd9725
---
 includes/api/ApiQuerySiteinfo.php       |  6 +++++-
 includes/cache/LocalisationCache.php    |  2 +-
 includes/parser/Parser.php              |  4 +++-
 languages/Language.php                  | 10 ++++++++++
 languages/messages/MessagesAry.php      |  2 --
 languages/messages/MessagesAz.php       |  2 --
 languages/messages/MessagesCe.php       |  2 --
 languages/messages/MessagesCrh_cyrl.php |  3 +--
 languages/messages/MessagesCrh_latn.php |  3 +--
 languages/messages/MessagesCu.php       |  3 +--
 languages/messages/MessagesCv.php       |  3 +--
 languages/messages/MessagesDiq.php      |  2 --
 languages/messages/MessagesEn.php       |  8 ++++++--
 languages/messages/MessagesGa.php       |  2 --
 languages/messages/MessagesGu.php       |  2 --
 languages/messages/MessagesId.php       |  2 --
 languages/messages/MessagesIs.php       |  3 +--
 languages/messages/MessagesKa.php       |  2 --
 languages/messages/MessagesKaa.php      |  3 +--
 languages/messages/MessagesKiu.php      |  2 --
 languages/messages/MessagesKm.php       |  2 --
 languages/messages/MessagesLtg.php      |  2 --
 languages/messages/MessagesMk.php       |  2 --
 languages/messages/MessagesMs.php       |  2 --
 languages/messages/MessagesMt.php       |  4 ++--
 languages/messages/MessagesNe.php       |  2 --
 languages/messages/MessagesNn.php       |  2 --
 languages/messages/MessagesQqq.php      |  2 --
 languages/messages/MessagesRo.php       |  2 --
 languages/messages/MessagesRoa_tara.php |  2 --
 languages/messages/MessagesSc.php       |  2 --
 languages/messages/MessagesSi.php       |  2 --
 languages/messages/MessagesSr_ec.php    |  2 --
 languages/messages/MessagesSr_el.php    |  2 --
 languages/messages/MessagesTl.php       |  2 --
 languages/messages/MessagesTt_cyrl.php  |  2 --
 languages/messages/MessagesTt_latn.php  |  2 --
 languages/messages/MessagesUdm.php      |  3 +--
 languages/messages/MessagesUg_arab.php  |  2 --
 languages/messages/MessagesUk.php       |  3 +--
 languages/messages/MessagesUz.php       |  3 +--
 languages/messages/MessagesWar.php      |  2 --
 maintenance/language/messageTypes.inc   |  1 -
 maintenance/language/messages.inc       |  3 ---
 44 files changed, 36 insertions(+), 83 deletions(-)

diff --git a/includes/api/ApiQuerySiteinfo.php b/includes/api/ApiQuerySiteinfo.php
index a7767062e8..ac9e85ab74 100644
--- a/includes/api/ApiQuerySiteinfo.php
+++ b/includes/api/ApiQuerySiteinfo.php
@@ -153,8 +153,12 @@ class ApiQuerySiteinfo extends ApiQueryBase {
 		}
 
 		if ( $wgContLang->linkPrefixExtension() ) {
-			$data['linkprefix'] = wfMessage( 'linkprefix' )->inContentLanguage()->text();
+			$linkPrefixCharset = $wgContLang->linkPrefixCharset();
+			$data['linkprefixcharset'] = $linkPrefixCharset;
+			// For backwards compatability
+			$data['linkprefix'] = "/^((?>.*[^$linkPrefixCharset]|))(.+)$/sDu";
 		} else {
+			$data['linkprefixcharset'] = '';
 			$data['linkprefix'] = '';
 		}
 
diff --git a/includes/cache/LocalisationCache.php b/includes/cache/LocalisationCache.php
index 1bfd17bde5..c5a153a88b 100644
--- a/includes/cache/LocalisationCache.php
+++ b/includes/cache/LocalisationCache.php
@@ -106,7 +106,7 @@ class LocalisationCache {
 		'fallback', 'namespaceNames', 'bookstoreList',
 		'magicWords', 'messages', 'rtl', 'capitalizeAllNouns', 'digitTransformTable',
 		'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
-		'linkTrail', 'namespaceAliases',
+		'linkTrail', 'linkPrefixCharset', 'namespaceAliases',
 		'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
 		'defaultDateFormat', 'extraUserToggles', 'specialPageAliases',
 		'imageFiles', 'preloadedMessages', 'namespaceGenderAliases',
diff --git a/includes/parser/Parser.php b/includes/parser/Parser.php
index 1f14223d69..6e9e06e12f 100644
--- a/includes/parser/Parser.php
+++ b/includes/parser/Parser.php
@@ -1863,7 +1863,9 @@ class Parser {
 		if ( $useLinkPrefixExtension ) {
 			# Match the end of a line for a word that's not followed by whitespace,
 			# e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
-			$e2 = wfMessage( 'linkprefix' )->inContentLanguage()->text();
+			global $wgContLang;
+			$charset = $wgContLang->linkPrefixCharset();
+			$e2 = "/^((?>.*[^$charset]|))(.+)$/sDu";
 		}
 
 		if ( is_null( $this->mTitle ) ) {
diff --git a/languages/Language.php b/languages/Language.php
index 527f382728..dc87bc894e 100644
--- a/languages/Language.php
+++ b/languages/Language.php
@@ -3949,6 +3949,16 @@ class Language {
 		return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
 	}
 
+	/**
+	 * A regular expression character set to match legal word-prefixing
+	 * characters which should be merged onto a link of the form foo[[bar]].
+	 *
+	 * @return string
+	 */
+	public function linkPrefixCharset() {
+		return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
+	}
+
 	/**
 	 * @return Language
 	 */
diff --git a/languages/messages/MessagesAry.php b/languages/messages/MessagesAry.php
index 9f1b3364d6..3c31b1f38e 100644
--- a/languages/messages/MessagesAry.php
+++ b/languages/messages/MessagesAry.php
@@ -142,8 +142,6 @@ $messages = array(
 'noindex-category' => 'shat mamfhtsach',
 'broken-file-category' => 'ṣfaḫi fiha wṣlat milffaṫ mhrrsa',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'ala',
 'article' => 'sfht mohtawa',
 'newwindow' => '(kayṫḫell fe ċerjem weḫdaĥor)',
diff --git a/languages/messages/MessagesAz.php b/languages/messages/MessagesAz.php
index 7b345b3496..a39894c588 100644
--- a/languages/messages/MessagesAz.php
+++ b/languages/messages/MessagesAz.php
@@ -206,8 +206,6 @@ $messages = array(
 'noindex-category' => 'İndeksləşdirilməyən səhifələr',
 'broken-file-category' => 'İşləməyən fayl keçidləri olan səhifələr',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Haqqında',
 'article' => 'Mündəricat',
 'newwindow' => '(yeni pəncərədə açılır)',
diff --git a/languages/messages/MessagesCe.php b/languages/messages/MessagesCe.php
index 2c1d5851e1..476b8369d3 100644
--- a/languages/messages/MessagesCe.php
+++ b/languages/messages/MessagesCe.php
@@ -459,8 +459,6 @@ $messages = array(
 'broken-file-category' => '{{#switch:{{NAMESPACE}}
  |{{ns:0}}=Болх цабеш файлийн хьажорагаш йолу агӀонаш}}',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Цуьнах лаьцна',
 'article' => 'Яззам',
 'newwindow' => '(керлачу корехь)',
diff --git a/languages/messages/MessagesCrh_cyrl.php b/languages/messages/MessagesCrh_cyrl.php
index a582b187da..c07de244d1 100644
--- a/languages/messages/MessagesCrh_cyrl.php
+++ b/languages/messages/MessagesCrh_cyrl.php
@@ -103,6 +103,7 @@ $dateFormats = array(
 
 $separatorTransformTable = array( ','  => '.', '.' => ',' );
 $linkTrail = '/^([a-zâçğıñöşüа-яё“»]+)(.*)$/sDu';
+$linkPrefixCharset = 'a-zâçğıñöşüA-ZÂÇĞİÑÖŞÜa-яёА-ЯЁ«„';
 
 $messages = array(
 # User preference toggles
@@ -230,8 +231,6 @@ $messages = array(
 'noindex-category' => 'Индекссиз саифелер',
 'broken-file-category' => 'Ичинде бозукъ файл багълантылары олгъан саифелер',
 
-'linkprefix' => '/^(.*?)([a-zâçğıñöşüA-ZÂÇĞİÑÖŞÜa-яёА-ЯЁ«„]+)$/sDu',
-
 'about' => 'Акъкъында',
 'article' => 'Саифе',
 'newwindow' => '(янъы бир пенджереде ачылыр)',
diff --git a/languages/messages/MessagesCrh_latn.php b/languages/messages/MessagesCrh_latn.php
index bd6306f978..6bb6a4cbf6 100644
--- a/languages/messages/MessagesCrh_latn.php
+++ b/languages/messages/MessagesCrh_latn.php
@@ -97,6 +97,7 @@ $dateFormats = array(
 
 $separatorTransformTable = array( ',' => '.', '.' => ',' );
 $linkTrail = '/^([a-zâçğıñöşüа-яё“»]+)(.*)$/sDu';
+$linkPrefixCharset = 'a-zâçğıñöşüA-ZÂÇĞİÑÖŞÜa-яёА-ЯЁ«„';
 
 $messages = array(
 # User preference toggles
@@ -224,8 +225,6 @@ $messages = array(
 'noindex-category' => 'Ä°ndekssiz saifeler',
 'broken-file-category' => 'İçinde bozuq fayl bağlantıları olğan saifeler',
 
-'linkprefix' => '/^(.*?)([a-zâçğıñöşüA-ZÂÇĞİÑÖŞÜa-яёА-ЯЁ«„]+)$/sDu',
-
 'about' => 'Aqqında',
 'article' => 'Saife',
 'newwindow' => '(yañı bir pencerede açılır)',
diff --git a/languages/messages/MessagesCu.php b/languages/messages/MessagesCu.php
index e9c7e059ed..706051ca97 100644
--- a/languages/messages/MessagesCu.php
+++ b/languages/messages/MessagesCu.php
@@ -97,6 +97,7 @@ $dateFormats = array(
 );
 
 $linkTrail = '/^([a-zабвгдеєжѕзїіıићклмнопсстѹфхѡѿцчшщъыьѣюѥѧѩѫѭѯѱѳѷѵґѓђёјйљњќуўџэ҄я“»]+)(.*)$/sDu';
+$linkPrefixCharset = '„«';
 
 $messages = array(
 # User preference toggles
@@ -180,8 +181,6 @@ $messages = array(
 'category-subcat-count' => '{{PLURAL:$2|Сѥи катигорїи тъкъмо сꙗ подъкатигорїꙗ ѥстъ|Сѥи катигорїи {{PLURAL:$1|ѥдина подъкатигорїꙗ ѥстъ|2 подъкатигорїи ѥстє|$1 подъкатигорїѩ сѫтъ}} · а вьсѩ жє подъкатигорїѩ число $2 ѥстъ}}',
 'listingcontinuesabbrev' => '· вѧщє',
 
-'linkprefix' => '/^(.*?)(„|«)$/sD',
-
 'about' => 'опьсаниѥ',
 'article' => 'члѣнъ',
 'newwindow' => '(иномь окънѣ)',
diff --git a/languages/messages/MessagesCv.php b/languages/messages/MessagesCv.php
index ef5f4f1ca4..7a93e72bff 100644
--- a/languages/messages/MessagesCv.php
+++ b/languages/messages/MessagesCv.php
@@ -42,6 +42,7 @@ $namespaceGenderAliases = array();
 
 $linkPrefixExtension = true;
 $linkTrail = '/^([a-zа-яĕçăӳ"»]+)(.*)$/sDu';
+$linkPrefixCharset = 'a-zA-Z"\\x{80}-\\x{10ffff}';
 
 $messages = array(
 # User preference toggles
@@ -153,8 +154,6 @@ $messages = array(
 'category-file-count-limited' => 'Ку категоринче $1 файл.',
 'listingcontinuesabbrev' => '(малалли)',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff«"]+)$/sD',
-
 'about' => 'Ăнлантаркăч',
 'article' => 'Статья',
 'newwindow' => '(çĕнĕ чӳречере)',
diff --git a/languages/messages/MessagesDiq.php b/languages/messages/MessagesDiq.php
index aa5ce1f21c..9892e183ef 100644
--- a/languages/messages/MessagesDiq.php
+++ b/languages/messages/MessagesDiq.php
@@ -458,8 +458,6 @@ $messages = array(
 'broken-file-category' => 'Peleye ke gıreyê dosyeyanê ğeletan muhtewa kenê',
 'categoryviewer-pagedlinks' => '($1) ($2)',
 
-'linkprefix' => "'''MediaWiki niya ro.'''",
-
 'about' => 'Heqa cı de',
 'article' => 'Wesiqe',
 'newwindow' => '(pençereyê newey de beno a)',
diff --git a/languages/messages/MessagesEn.php b/languages/messages/MessagesEn.php
index 6d68d36aa5..7c5b1b9e03 100644
--- a/languages/messages/MessagesEn.php
+++ b/languages/messages/MessagesEn.php
@@ -494,6 +494,12 @@ $specialPageAliases = array(
  */
 $linkTrail = '/^([a-z]+)(.*)$/sD';
 
+/**
+ * Regular expression charset matching the "link prefix", e.g. "foo" in
+ * foo[[bar]]. UTF-8 characters may be used.
+ */
+$linkPrefixCharset = 'a-zA-Z\\x{80}-\\x{10ffff}';
+
 /**
  * List of filenames for some ui images that can be overridden per language
  * basis if needed.
@@ -790,8 +796,6 @@ future releases. Also note that since each list value is wrapped in a unique
 'broken-file-category'           => 'Pages with broken file links',
 'categoryviewer-pagedlinks'      => '($1) ($2)', # only translate this message to other languages if you have to change it
 
-'linkprefix' => '/^((?>.*(?<![a-zA-Z\\x80-\\xff])))(.+)$/sD', # only translate this message to other languages if you have to change it
-
 'about'         => 'About',
 'article'       => 'Content page',
 'newwindow'     => '(opens in new window)',
diff --git a/languages/messages/MessagesGa.php b/languages/messages/MessagesGa.php
index d7b37f5564..c8172dc0e3 100644
--- a/languages/messages/MessagesGa.php
+++ b/languages/messages/MessagesGa.php
@@ -218,8 +218,6 @@ $messages = array(
 'listingcontinuesabbrev' => 'ar lean.',
 'index-category' => 'Leathanaigh innéacsaithe',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Maidir leis',
 'article' => 'Leathanach ábhair',
 'newwindow' => '(a osclófar i bhfuinneog nua)',
diff --git a/languages/messages/MessagesGu.php b/languages/messages/MessagesGu.php
index 69692e85ee..864ffa5fc2 100644
--- a/languages/messages/MessagesGu.php
+++ b/languages/messages/MessagesGu.php
@@ -291,8 +291,6 @@ $messages = array(
 'noindex-category' => 'અનુક્રમણિકા નહી બનાવેલા પાનાં',
 'broken-file-category' => 'ફાઇલોની ત્રૂટક કડીઓવાળાં પાનાં',
 
-'linkprefix' => '/^(.*?)((?:[a-zA-Z\\x80-\\xff]|ક્|ખ્|ગ્|ઘ્|ચ્|છ્|જ્|ઝ્|ટ્|ઠ્|ડ્|ઢ્|ણ્|ત્|થ્|દ્|ધ્|ન્|પ્|ફ્|બ્|ભ્|મ્|ય્|ર્|લ્|વ્|સ્|શ્|ષ્|હ્|ળ્|ક્ષ્|જ્ઞ્|અ|આ|ઇ|ઈ|ઉ|ઊ|એ|ઐ|ઓ|ઔ|અં|અઃ|અઁ|ઍ|ઑ|ઋ|ઁ|઼|।|્|ા|િ|ી|ુ|ૂ|ે|ૈ|ો|ૌ|ં|ઃ|ૅ|ૉ|ૃ)+)$/sD',
-
 'about' => 'વિષે',
 'article' => 'લેખનું પાનું',
 'newwindow' => '(નવા પાનામાં ખુલશે)',
diff --git a/languages/messages/MessagesId.php b/languages/messages/MessagesId.php
index cee748da37..dd1a209f0f 100644
--- a/languages/messages/MessagesId.php
+++ b/languages/messages/MessagesId.php
@@ -481,8 +481,6 @@ $messages = array(
 'broken-file-category' => 'Halaman dengan gambar rusak',
 'categoryviewer-pagedlinks' => '($1) ($2)',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Tentang',
 'article' => 'Halaman isi',
 'newwindow' => '(buka di jendela baru)',
diff --git a/languages/messages/MessagesIs.php b/languages/messages/MessagesIs.php
index 2e25bdd5aa..8b76026bb0 100644
--- a/languages/messages/MessagesIs.php
+++ b/languages/messages/MessagesIs.php
@@ -231,6 +231,7 @@ $specialPageAliases = array(
 $separatorTransformTable = array( ',' => '.', '.' => ',' );
 $linkPrefixExtension = true;
 $linkTrail = '/^([áðéíóúýþæöa-z-–]+)(.*)$/sDu';
+$linkPrefixCharset = 'áÁðÐéÉíÍóÓúÚýÝþÞæÆöÖA-Za-z–-';
 
 $messages = array(
 # User preference toggles
@@ -371,8 +372,6 @@ $messages = array(
 'noindex-category' => 'Óraðaðar skrár',
 'broken-file-category' => 'Síður með brotna myndatengla',
 
-'linkprefix' => '/^(.*?)([áÁðÐéÉíÍóÓúÚýÝþÞæÆöÖA-Za-z-–]+)$/sDu',
-
 'about' => 'Um',
 'article' => 'Efnissíða',
 'newwindow' => '(opnast í nýjum glugga)',
diff --git a/languages/messages/MessagesKa.php b/languages/messages/MessagesKa.php
index a4cd57a570..dfcdb94732 100644
--- a/languages/messages/MessagesKa.php
+++ b/languages/messages/MessagesKa.php
@@ -310,8 +310,6 @@ $messages = array(
 'broken-file-category' => 'გვერდები ფაილების არასწორი ბმულებით',
 'categoryviewer-pagedlinks' => '($1) ($2)',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'შესახებ',
 'article' => 'სტატია',
 'newwindow' => '(ახალ ფანჯარაში)',
diff --git a/languages/messages/MessagesKaa.php b/languages/messages/MessagesKaa.php
index 147e19af70..b1314d1d71 100644
--- a/languages/messages/MessagesKaa.php
+++ b/languages/messages/MessagesKaa.php
@@ -151,6 +151,7 @@ $dateFormats = array(
 );
 
 $linkTrail = "/^((?:[a-zıʼ’“»]|'(?!'))+)(.*)$/sDu";
+$linkPrefixCharset = 'a-zıA-Zİ\\x80-\\xff';
 
 $messages = array(
 # User preference toggles
@@ -264,8 +265,6 @@ $messages = array(
 'category-file-count-limited' => "Usı kategoriyada to'mendegi {{PLURAL:$1|fayl|$1 fayl}} bar.",
 'listingcontinuesabbrev' => 'dawamı',
 
-'linkprefix' => '/^(.*?)([a-zıA-Zİ\\x80-\\xff]+)$/sDu',
-
 'about' => 'Haqqında',
 'article' => "Mag'lıwmat beti",
 'newwindow' => "(jan'a aynada)",
diff --git a/languages/messages/MessagesKiu.php b/languages/messages/MessagesKiu.php
index 8f18ce2eec..36c0437f48 100644
--- a/languages/messages/MessagesKiu.php
+++ b/languages/messages/MessagesKiu.php
@@ -177,8 +177,6 @@ $messages = array(
 'noindex-category' => 'Pelê bêendeksıni',
 'broken-file-category' => 'Peli be gıreunê dosyeunê sıkıtau',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Heqa',
 'article' => 'Pela tedeesteyu',
 'newwindow' => '(zerrê pençerê dê newey de beno ra)',
diff --git a/languages/messages/MessagesKm.php b/languages/messages/MessagesKm.php
index c4e951be3c..063439cd58 100644
--- a/languages/messages/MessagesKm.php
+++ b/languages/messages/MessagesKm.php
@@ -402,8 +402,6 @@ $messages = array(
 'broken-file-category' => 'ទំព័រទាំងឡាយដែលដាច់តំណភ្ជាប់',
 'categoryviewer-pagedlinks' => '($1) ($2)',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'អំពី',
 'article' => 'មាតិកាអត្ថបទ',
 'newwindow' => '(បើក​លើ​បង្អួច​ថ្មី)',
diff --git a/languages/messages/MessagesLtg.php b/languages/messages/MessagesLtg.php
index 1b4b7f5c45..558d6c0112 100644
--- a/languages/messages/MessagesLtg.php
+++ b/languages/messages/MessagesLtg.php
@@ -105,8 +105,6 @@ $messages = array(
 'category-article-count' => '{{PLURAL:$2|Itymā kategorejā ir vīn dūtuo puslopa.|{{PLURAL:$1|Paruodeita $1 puslopa|Paruodeitys $1 puslopys}} nu $2.}}',
 'listingcontinuesabbrev' => '(tuoļuojums)',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Aproksts',
 'article' => 'Rakstīņs',
 'newwindow' => '(atdareišona jaunuo puslopā)',
diff --git a/languages/messages/MessagesMk.php b/languages/messages/MessagesMk.php
index 982427e346..8b726e7977 100644
--- a/languages/messages/MessagesMk.php
+++ b/languages/messages/MessagesMk.php
@@ -489,8 +489,6 @@ $messages = array(
 'noindex-category' => 'Неиндексирани страници',
 'broken-file-category' => 'Страници со прекинати врски до податотеки',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'За {{SITENAME}}',
 'article' => 'Статија',
 'newwindow' => '(се отвора во нов прозорец)',
diff --git a/languages/messages/MessagesMs.php b/languages/messages/MessagesMs.php
index 2f07107d7b..dfef63f1e9 100644
--- a/languages/messages/MessagesMs.php
+++ b/languages/messages/MessagesMs.php
@@ -338,8 +338,6 @@ $messages = array(
 'noindex-category' => 'Laman tak diindeks',
 'broken-file-category' => 'Laman yang ada pautan fail yang terputus',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Perihal',
 'article' => 'Laman kandungan',
 'newwindow' => '(dibuka di tetingkap baru)',
diff --git a/languages/messages/MessagesMt.php b/languages/messages/MessagesMt.php
index 6bec1c6da2..a1bea2d548 100644
--- a/languages/messages/MessagesMt.php
+++ b/languages/messages/MessagesMt.php
@@ -271,6 +271,8 @@ $magicWords = array(
 	'formatdate'                => array( '0', 'formatdata', 'dataformat', 'formatdate', 'dateformat' ),
 );
 
+$linkPrefixCharset = 'A-\\x{10ffff}';
+
 $messages = array(
 # User preference toggles
 'tog-underline' => 'Ħoloq sottolinjati:',
@@ -411,8 +413,6 @@ $messages = array(
 'noindex-category' => 'Paġni mhux indiċizzati',
 'broken-file-category' => "Paġni b'ħoloq lejn fajls miksura",
 
-'linkprefix' => '/^(.*?)([a-żA-Ż\\x80-\\xff]+)$/sD',
-
 'about' => 'Dwar',
 'article' => 'artiklu',
 'newwindow' => "(tinfetaħ f'tieqa ġdida)",
diff --git a/languages/messages/MessagesNe.php b/languages/messages/MessagesNe.php
index 6abbfdace5..99a16bf136 100644
--- a/languages/messages/MessagesNe.php
+++ b/languages/messages/MessagesNe.php
@@ -193,8 +193,6 @@ $messages = array(
 'noindex-category' => 'क्रमांकन नगरिएका पृष्ठहरु',
 'broken-file-category' => 'टुटेको फाइल लिंकसितको पृष्ठ',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'बारेमा',
 'article' => 'सामाग्री पृष्ठ',
 'newwindow' => '(नयाँ विन्डोमा खुल्छ)',
diff --git a/languages/messages/MessagesNn.php b/languages/messages/MessagesNn.php
index fe944b1903..47bc87c391 100644
--- a/languages/messages/MessagesNn.php
+++ b/languages/messages/MessagesNn.php
@@ -457,8 +457,6 @@ $messages = array(
 'noindex-category' => 'Ikkje-indekserte sider',
 'broken-file-category' => 'Sider med brotne fillenkjer',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Om',
 'article' => 'Innhaldsside',
 'newwindow' => '(vert opna i eit nytt vindauge)',
diff --git a/languages/messages/MessagesQqq.php b/languages/messages/MessagesQqq.php
index a8b66b29df..9d58538a43 100644
--- a/languages/messages/MessagesQqq.php
+++ b/languages/messages/MessagesQqq.php
@@ -426,8 +426,6 @@ The pagination links in category viewer. Parameters:
 * $1 - the previous link, uses {{msg-mw|Prevn}}
 * $2 - the next link, uses {{msg-mw|Nextn}}',
 
-'linkprefix' => '{{optional}}',
-
 'about' => '{{Identical|About}}',
 'article' => "A 'content page' is a page that forms part of the purpose of the wiki. It includes the main page and pages in the main namespace and any other namespaces that are included when the wiki is customised. For example on Wikimedia Commons 'content pages' include pages in the file and category namespaces. On Wikinews 'content pages' include pages in the Portal namespace. For technical definition of 'content namespaces' see [[mw:Manual:Using_custom_namespaces#Content_namespaces|MediaWiki]].
 
diff --git a/languages/messages/MessagesRo.php b/languages/messages/MessagesRo.php
index ad808244b1..0dddd8b9e2 100644
--- a/languages/messages/MessagesRo.php
+++ b/languages/messages/MessagesRo.php
@@ -438,8 +438,6 @@ $messages = array(
 'noindex-category' => 'Pagini neindexate',
 'broken-file-category' => 'Pagini cu legături invalide către fișiere',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Despre',
 'article' => 'Articol',
 'newwindow' => '(se deschide într-o fereastră nouă)',
diff --git a/languages/messages/MessagesRoa_tara.php b/languages/messages/MessagesRoa_tara.php
index bc1f7730e0..0203dd1dea 100644
--- a/languages/messages/MessagesRoa_tara.php
+++ b/languages/messages/MessagesRoa_tara.php
@@ -165,8 +165,6 @@ $messages = array(
 'broken-file-category' => 'Pàggene cu collegaminde a le file scuasciate',
 'categoryviewer-pagedlinks' => '($1) ($2)',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Sus a',
 'article' => 'Pàgene de le condenute',
 'newwindow' => "(iapre jndr'à 'na fenestra nova)",
diff --git a/languages/messages/MessagesSc.php b/languages/messages/MessagesSc.php
index bfc2b2be00..d97bb14691 100644
--- a/languages/messages/MessagesSc.php
+++ b/languages/messages/MessagesSc.php
@@ -177,8 +177,6 @@ $messages = array(
 'index-category' => 'Pàginas indicizadas',
 'noindex-category' => 'Pàginas no indicitzadas',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'A propòsitu de',
 'article' => 'Artìculu',
 'newwindow' => '(aberit in una bentana noa)',
diff --git a/languages/messages/MessagesSi.php b/languages/messages/MessagesSi.php
index 4a30265408..a020dd0a49 100644
--- a/languages/messages/MessagesSi.php
+++ b/languages/messages/MessagesSi.php
@@ -331,8 +331,6 @@ $messages = array(
 'noindex-category' => 'සූචිගත නොකළ පිටු',
 'broken-file-category' => 'භින්න වූ ගොනු සබැඳි සහිත පිටු',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'පිළිබඳ',
 'article' => 'පටුන',
 'newwindow' => '(නව කවුළුවක විවෘතවේ)',
diff --git a/languages/messages/MessagesSr_ec.php b/languages/messages/MessagesSr_ec.php
index b28c453e9c..118a9dd7e4 100644
--- a/languages/messages/MessagesSr_ec.php
+++ b/languages/messages/MessagesSr_ec.php
@@ -536,8 +536,6 @@ $messages = array(
 'noindex-category' => 'Непописане странице',
 'broken-file-category' => 'Странице с неисправним везама до датотека',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'О нама',
 'article' => 'Страница са садржајем',
 'newwindow' => '(отвара се у новом прозору)',
diff --git a/languages/messages/MessagesSr_el.php b/languages/messages/MessagesSr_el.php
index bb7f381146..97d433a4eb 100644
--- a/languages/messages/MessagesSr_el.php
+++ b/languages/messages/MessagesSr_el.php
@@ -444,8 +444,6 @@ $messages = array(
 'noindex-category' => 'Nepopisane stranice',
 'broken-file-category' => 'Stranice s neispravnim vezama do datoteka',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'O nama',
 'article' => 'Stranica sa sadržajem',
 'newwindow' => '(otvara u novom prozoru)',
diff --git a/languages/messages/MessagesTl.php b/languages/messages/MessagesTl.php
index becffb7c3c..028870da92 100644
--- a/languages/messages/MessagesTl.php
+++ b/languages/messages/MessagesTl.php
@@ -285,8 +285,6 @@ $messages = array(
 'broken-file-category' => 'Mga pahina na may sirang mga kawing ng talaksan',
 'categoryviewer-pagedlinks' => '($1) ($2)',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Patungkol',
 'article' => 'Pahina ng nilalaman',
 'newwindow' => '(magbubukas sa bagong bintana)',
diff --git a/languages/messages/MessagesTt_cyrl.php b/languages/messages/MessagesTt_cyrl.php
index f25aebf6e5..f6b9404b98 100644
--- a/languages/messages/MessagesTt_cyrl.php
+++ b/languages/messages/MessagesTt_cyrl.php
@@ -348,8 +348,6 @@ $messages = array(
 'noindex-category' => 'Индексланмаган битләр',
 'broken-file-category' => 'Файлларга эшләми торган сылтамалар булган битләр',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Тасвирлама',
 'article' => 'Мәкалә',
 'newwindow' => '(яңа тәрәзәдә ачыла)',
diff --git a/languages/messages/MessagesTt_latn.php b/languages/messages/MessagesTt_latn.php
index 6d17c364f3..8f96b5be6b 100644
--- a/languages/messages/MessagesTt_latn.php
+++ b/languages/messages/MessagesTt_latn.php
@@ -216,8 +216,6 @@ $messages = array(
 'index-category' => 'İndekslanğan bitlär',
 'noindex-category' => 'İndekslanmağan bitlär',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Taswirlama',
 'article' => 'Mäqälä',
 'newwindow' => '(yaña täräzädä açıla)',
diff --git a/languages/messages/MessagesUdm.php b/languages/messages/MessagesUdm.php
index ac03870dce..9900e248bf 100644
--- a/languages/messages/MessagesUdm.php
+++ b/languages/messages/MessagesUdm.php
@@ -44,6 +44,7 @@ $namespaceAliases = array(
 $namespaceGenderAliases = array();
 
 $linkTrail = '/^([a-zа-яёӝӟӥӧӵ“»]+)(.*)$/sDu';
+$linkPrefixCharset = '„«';
 $fallback8bitEncoding = 'windows-1251';
 $separatorTransformTable = array( ',' => "\xc2\xa0", '.' => ',' );
 
@@ -151,8 +152,6 @@ $messages = array(
 'index-category' => 'Индексировать кароно бамъёс',
 'noindex-category' => 'Индексировать каронтэм бамъёс',
 
-'linkprefix' => '/^((?>.*(?<!(?:„|«)$)))(.+)$/sDu',
-
 'about' => 'Та сярысь',
 'article' => 'Статья',
 'mypage' => 'Ас бам',
diff --git a/languages/messages/MessagesUg_arab.php b/languages/messages/MessagesUg_arab.php
index ab5ee42361..aad3464883 100644
--- a/languages/messages/MessagesUg_arab.php
+++ b/languages/messages/MessagesUg_arab.php
@@ -173,8 +173,6 @@ $messages = array(
 'broken-file-category' => 'ھۆججەت ئۇلىنىشى بۇزۇلغان بەتلەر',
 'categoryviewer-pagedlinks' => '($1) ($2)',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'ھەققىدە',
 'article' => 'مۇندەرىجە',
 'newwindow' => '(يېڭى كۆزنەكتە ئاچ)',
diff --git a/languages/messages/MessagesUk.php b/languages/messages/MessagesUk.php
index 105bef1bdd..b9795cc4bb 100644
--- a/languages/messages/MessagesUk.php
+++ b/languages/messages/MessagesUk.php
@@ -361,6 +361,7 @@ $magicWords = array(
 );
 
 $linkTrail = '/^([a-zабвгґдеєжзиіїйклмнопрстуфхцчшщьєюяёъы“»]+)(.*)$/sDu';
+$linkPrefixCharset = '„«';
 
 $messages = array(
 # User preference toggles
@@ -504,8 +505,6 @@ $messages = array(
 'broken-file-category' => 'Сторінки, що посилаються на неіснуючі файли',
 'categoryviewer-pagedlinks' => '($1) ($2)',
 
-'linkprefix' => '/^(.*?)(„|«)$/sD',
-
 'about' => 'Про',
 'article' => 'Стаття',
 'newwindow' => '(відкривається в новому вікні)',
diff --git a/languages/messages/MessagesUz.php b/languages/messages/MessagesUz.php
index e8746b379a..b154b148f9 100644
--- a/languages/messages/MessagesUz.php
+++ b/languages/messages/MessagesUz.php
@@ -107,6 +107,7 @@ $magicWords = array(
 );
 
 $linkTrail = '/^([a-zʻʼ“»]+)(.*)$/sDu';
+$linkPrefixCharset = 'a-zA-Z\\x80-\\xffʻʼ«„';
 
 $messages = array(
 # User preference toggles
@@ -236,8 +237,6 @@ $messages = array(
 'noindex-category' => 'Indekslanmaydigan sahifalar',
 'broken-file-category' => 'Ishlamaydigan fayl havolalari bor sahifalar',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xffʻʼ«„]+)$/sDu',
-
 'about' => 'Haqida',
 'article' => 'Sahifa',
 'newwindow' => '(yangi oynada ochiladi)',
diff --git a/languages/messages/MessagesWar.php b/languages/messages/MessagesWar.php
index bd99d85963..14db0861a5 100644
--- a/languages/messages/MessagesWar.php
+++ b/languages/messages/MessagesWar.php
@@ -203,8 +203,6 @@ $messages = array(
 'noindex-category' => 'Mga diri nakatudlokan nga pagkli',
 'broken-file-category' => 'Mga pakli nga mayda utod nga mga sumpay hin paypay',
 
-'linkprefix' => '/^(.*?)([a-zA-Z\\x80-\\xff]+)$/sD',
-
 'about' => 'Mahitungod han',
 'article' => 'Pakli hin sulod',
 'newwindow' => '(nabuklad hin bag-o nga tamboan o bintana)',
diff --git a/maintenance/language/messageTypes.inc b/maintenance/language/messageTypes.inc
index 8676d74146..267666d38b 100644
--- a/maintenance/language/messageTypes.inc
+++ b/maintenance/language/messageTypes.inc
@@ -266,7 +266,6 @@ $wgIgnoredMessages = array(
 
 /** Optional messages, which may be translated only if changed in the target language. */
 $wgOptionalMessages = array(
-	'linkprefix',
 	'feed-atom',
 	'feed-rss',
 	'unit-pixel',
diff --git a/maintenance/language/messages.inc b/maintenance/language/messages.inc
index a4fc922bcd..1ac9500164 100644
--- a/maintenance/language/messages.inc
+++ b/maintenance/language/messages.inc
@@ -169,9 +169,6 @@ $wgMessageStructure = array(
 		'broken-file-category',
 		'categoryviewer-pagedlinks',
 	),
-	'mainpage' => array(
-		'linkprefix',
-	),
 	'miscellaneous1' => array(
 		'about',
 		'article',
-- 
2.20.1