From: Cormac Parle <cparle@wikimedia.org>
Date: Wed, 18 Oct 2017 12:38:48 +0000 (+0100)
Subject: Treat langtags in SVG switch case-insensitively
X-Git-Tag: 1.31.0-rc.0~1498^2
X-Git-Url: https://git.cyclocoop.org/%7B%24www_url%7Dadmin/compta/banques/?a=commitdiff_plain;h=f6620e2a7576b81ac5d3ca2f598746d1bc70c887;p=lhc%2Fweb%2Fwiklou.git

Treat langtags in SVG switch case-insensitively

See https://tools.ietf.org/html/bcp47#section-2.1.1

Also implement matching of systemLanguage attribs as
specified in the SVG spec

Note that librsvg that we use for rendering pngs of svg
files has a bug, and matches languages in the following
way instead of what is implemented in SVG::getMatchedLanguage()

public function getMatchedLanguage( $userPreferredLanguage, array $svgLanguages ) {
  foreach ( $svgLanguages as $svgLang ) {
    if ($svgLang == $userPreferredLanguage) {
      return $svgLang;
    }
    $dashPosition = strpos( $userPreferredLanguage, '-' );
    if ( $dashPosition !== false ) {
      if (
        strtolower( substr( $svgLang, 0, $dashPosition ) ) ==
        strtolower( substr( $userPreferredLanguage, 0, $dashPosition ) )
      ) {
        return $svgLang;
      }
    }
    return null;
}

Bug: T154132
Change-Id: Ibff66a0844f0cecfae0260c6a7d20aeedc2849a2
---

diff --git a/includes/filerepo/file/File.php b/includes/filerepo/file/File.php
index 32f4504ba6..dd12ab5d95 100644
--- a/includes/filerepo/file/File.php
+++ b/includes/filerepo/file/File.php
@@ -581,6 +581,25 @@ abstract class File implements IDBAccessObject {
 		}
 	}
 
+	/**
+	 * Get the language code from the available languages for this file that matches the language
+	 * requested by the user
+	 *
+	 * @param string $userPreferredLanguage
+	 * @return string|null
+	 */
+	public function getMatchedLanguage( $userPreferredLanguage ) {
+		$handler = $this->getHandler();
+		if ( $handler && method_exists( $handler, 'getMatchedLanguage' ) ) {
+			return $handler->getMatchedLanguage(
+				$userPreferredLanguage,
+				$handler->getAvailableLanguages( $this )
+			);
+		} else {
+			return null;
+		}
+	}
+
 	/**
 	 * In files that support multiple language, what is the default language
 	 * to use if none specified.
diff --git a/includes/media/SVG.php b/includes/media/SVG.php
index bd78b49e5d..2b138930d1 100644
--- a/includes/media/SVG.php
+++ b/includes/media/SVG.php
@@ -97,19 +97,50 @@ class SvgHandler extends ImageHandler {
 			if ( isset( $metadata['translations'] ) ) {
 				foreach ( $metadata['translations'] as $lang => $langType ) {
 					if ( $langType === SVGReader::LANG_FULL_MATCH ) {
-						$langList[] = $lang;
+						$langList[] = strtolower( $lang );
 					}
 				}
 			}
 		}
-		return $langList;
+		return array_unique( $langList );
 	}
 
 	/**
-	 * What language to render file in if none selected.
+	 * SVG's systemLanguage matching rules state:
+	 * 'The `systemLanguage` attribute ... [e]valuates to "true" if one of the languages indicated
+	 * by user preferences exactly equals one of the languages given in the value of this parameter,
+	 * or if one of the languages indicated by user preferences exactly equals a prefix of one of
+	 * the languages given in the value of this parameter such that the first tag character
+	 * following the prefix is "-".'
 	 *
-	 * @param File $file
-	 * @return string Language code.
+	 * Return the first element of $svgLanguages that matches $userPreferredLanguage
+	 *
+	 * @see https://www.w3.org/TR/SVG/struct.html#SystemLanguageAttribute
+	 * @param string $userPreferredLanguage
+	 * @param array $svgLanguages
+	 * @return string|null
+	 */
+	public function getMatchedLanguage( $userPreferredLanguage, array $svgLanguages ) {
+		foreach ( $svgLanguages as $svgLang ) {
+			if ( strcasecmp( $svgLang, $userPreferredLanguage ) === 0 ) {
+				return $svgLang;
+			}
+			$trimmedSvgLang = $svgLang;
+			while ( strpos( $trimmedSvgLang, '-' ) !== false ) {
+				$trimmedSvgLang = substr( $trimmedSvgLang, 0, strrpos( $trimmedSvgLang, '-' ) );
+				if ( strcasecmp( $trimmedSvgLang, $userPreferredLanguage ) === 0 ) {
+					return $svgLang;
+				}
+			}
+		}
+		return null;
+	}
+
+	/**
+	 * What language to render file in if none selected
+	 *
+	 * @param File $file Language code
+	 * @return string
 	 */
 	public function getDefaultRenderLanguage( File $file ) {
 		return 'en';
@@ -479,7 +510,7 @@ class SvgHandler extends ImageHandler {
 			return ( $value > 0 );
 		} elseif ( $name == 'lang' ) {
 			// Validate $code
-			if ( $value === '' || !Language::isValidBuiltInCode( $value ) ) {
+			if ( $value === '' || !Language::isValidCode( $value ) ) {
 				wfDebug( "Invalid user language code\n" );
 
 				return false;
@@ -499,8 +530,7 @@ class SvgHandler extends ImageHandler {
 	public function makeParamString( $params ) {
 		$lang = '';
 		if ( isset( $params['lang'] ) && $params['lang'] !== 'en' ) {
-			$params['lang'] = strtolower( $params['lang'] );
-			$lang = "lang{$params['lang']}-";
+			$lang = 'lang' . strtolower( $params['lang'] ) . '-';
 		}
 		if ( !isset( $params['width'] ) ) {
 			return false;
@@ -511,7 +541,7 @@ class SvgHandler extends ImageHandler {
 
 	public function parseParamString( $str ) {
 		$m = false;
-		if ( preg_match( '/^lang([a-z]+(?:-[a-z]+)*)-(\d+)px$/', $str, $m ) ) {
+		if ( preg_match( '/^lang([a-z]+(?:-[a-z]+)*)-(\d+)px$/i', $str, $m ) ) {
 			return [ 'width' => array_pop( $m ), 'lang' => $m[1] ];
 		} elseif ( preg_match( '/^(\d+)px$/', $str, $m ) ) {
 			return [ 'width' => $m[1], 'lang' => 'en' ];
diff --git a/includes/page/ImagePage.php b/includes/page/ImagePage.php
index 62f5d00c33..935d299fd3 100644
--- a/includes/page/ImagePage.php
+++ b/includes/page/ImagePage.php
@@ -285,6 +285,19 @@ class ImagePage extends Article {
 		return parent::getContentObject();
 	}
 
+	private function getLanguageForRendering( WebRequest $request, File $file ) {
+		$handler = $this->displayImg->getHandler();
+
+		$requestLanguage = $request->getVal( 'lang' );
+		if ( !is_null( $requestLanguage ) ) {
+			if ( $handler && $handler->validateParam( 'lang', $requestLanguage ) ) {
+				return $requestLanguage;
+			}
+		}
+
+		return $handler->getDefaultRenderLanguage( $this->displayImg );
+	}
+
 	protected function openShowImage() {
 		global $wgEnableUploads, $wgSend404Code, $wgSVGMaxSize;
 
@@ -309,14 +322,9 @@ class ImagePage extends Article {
 				$params = [ 'page' => $page ];
 			}
 
-			$renderLang = $request->getVal( 'lang' );
+			$renderLang = $this->getLanguageForRendering( $request, $this->displayImg );
 			if ( !is_null( $renderLang ) ) {
-				$handler = $this->displayImg->getHandler();
-				if ( $handler && $handler->validateParam( 'lang', $renderLang ) ) {
-					$params['lang'] = $renderLang;
-				} else {
-					$renderLang = null;
-				}
+				$params['lang'] = $renderLang;
 			}
 
 			$width_orig = $this->displayImg->getWidth( $page );
@@ -544,12 +552,7 @@ EOT
 
 			$renderLangOptions = $this->displayImg->getAvailableLanguages();
 			if ( count( $renderLangOptions ) >= 1 ) {
-				$currentLanguage = $renderLang;
-				$defaultLang = $this->displayImg->getDefaultRenderLanguage();
-				if ( is_null( $currentLanguage ) ) {
-					$currentLanguage = $defaultLang;
-				}
-				$out->addHTML( $this->doRenderLangOpt( $renderLangOptions, $currentLanguage, $defaultLang ) );
+				$out->addHTML( $this->doRenderLangOpt( $renderLangOptions, $renderLang ) );
 			}
 
 			// Add cannot animate thumbnail warning
@@ -1047,60 +1050,31 @@ EOT
 	 * Output a drop-down box for language options for the file
 	 *
 	 * @param array $langChoices Array of string language codes
-	 * @param string $curLang Language code file is being viewed in.
-	 * @param string $defaultLang Language code that image is rendered in by default
+	 * @param string $renderLang Language code for the language we want the file to rendered in.
 	 * @return string HTML to insert underneath image.
 	 */
-	protected function doRenderLangOpt( array $langChoices, $curLang, $defaultLang ) {
+	protected function doRenderLangOpt( array $langChoices, $renderLang ) {
 		global $wgScript;
-		sort( $langChoices );
-		$curLang = LanguageCode::bcp47( $curLang );
-		$defaultLang = LanguageCode::bcp47( $defaultLang );
 		$opts = '';
-		$haveCurrentLang = false;
-		$haveDefaultLang = false;
-
-		// We make a list of all the language choices in the file.
-		// Additionally if the default language to render this file
-		// is not included as being in this file (for example, in svgs
-		// usually the fallback content is the english content) also
-		// include a choice for that. Last of all, if we're viewing
-		// the file in a language not on the list, add it as a choice.
+
+		$matchedRenderLang = $this->displayImg->getMatchedLanguage( $renderLang );
+
 		foreach ( $langChoices as $lang ) {
-			$code = LanguageCode::bcp47( $lang );
-			$name = Language::fetchLanguageName( $code, $this->getContext()->getLanguage()->getCode() );
-			if ( $name !== '' ) {
-				$display = $this->getContext()->msg( 'img-lang-opt', $code, $name )->text();
-			} else {
-				$display = $code;
-			}
-			$opts .= "\n" . Xml::option( $display, $code, $curLang === $code );
-			if ( $curLang === $code ) {
-				$haveCurrentLang = true;
-			}
-			if ( $defaultLang === $code ) {
-				$haveDefaultLang = true;
-			}
-		}
-		if ( !$haveDefaultLang ) {
-			// Its hard to know if the content is really in the default language, or
-			// if its just unmarked content that could be in any language.
-			$opts = Xml::option(
-					$this->getContext()->msg( 'img-lang-default' )->text(),
-				$defaultLang,
-				$defaultLang === $curLang
-			) . $opts;
-		}
-		if ( !$haveCurrentLang && $defaultLang !== $curLang ) {
-			$name = Language::fetchLanguageName( $curLang, $this->getContext()->getLanguage()->getCode() );
-			if ( $name !== '' ) {
-				$display = $this->getContext()->msg( 'img-lang-opt', $curLang, $name )->text();
-			} else {
-				$display = $curLang;
-			}
-			$opts = Xml::option( $display, $curLang, true ) . $opts;
+			$opts .= $this->createXmlOptionStringForLanguage(
+				$lang,
+				$matchedRenderLang === $lang
+			);
 		}
 
+		// Allow for the default case in an svg <switch> that is displayed if no
+		// systemLanguage attribute matches
+		$opts .= "\n" .
+			Xml::option(
+				$this->getContext()->msg( 'img-lang-default' )->text(),
+				'und',
+				is_null( $matchedRenderLang )
+			);
+
 		$select = Html::rawElement(
 			'select',
 			[ 'id' => 'mw-imglangselector', 'name' => 'lang' ],
@@ -1119,6 +1093,27 @@ EOT
 		return $langSelectLine;
 	}
 
+	/**
+	 * @param $lang string
+	 * @param $selected bool
+	 * @return string
+	 */
+	private function createXmlOptionStringForLanguage( $lang, $selected ) {
+		$code = LanguageCode::bcp47( $lang );
+		$name = Language::fetchLanguageName( $code, $this->getContext()->getLanguage()->getCode() );
+		if ( $name !== '' ) {
+			$display = $this->getContext()->msg( 'img-lang-opt', $code, $name )->text();
+		} else {
+			$display = $code;
+		}
+		return "\n" .
+			Xml::option(
+				$display,
+				$lang,
+				$selected
+			);
+	}
+
 	/**
 	 * Get the width and height to display image at.
 	 *
diff --git a/tests/parser/parserTests.txt b/tests/parser/parserTests.txt
index 3c861ea10b..41b09879e0 100644
--- a/tests/parser/parserTests.txt
+++ b/tests/parser/parserTests.txt
@@ -15046,9 +15046,9 @@ SVG thumbnails with invalid language code
 !! options
 parsoid=wt2html,wt2wt,html2html
 !! wikitext
-[[File:Foobar.svg|thumb|caption|lang=invalid.language.code]]
+[[File:Foobar.svg|thumb|caption|lang=invalid:language:code]]
 !! html/php
-<div class="thumb tright"><div class="thumbinner" style="width:182px;"><a href="/wiki/File:Foobar.svg" class="image"><img alt="" src="http://example.com/images/thumb/f/ff/Foobar.svg/180px-Foobar.svg.png" width="180" height="135" class="thumbimage" srcset="http://example.com/images/thumb/f/ff/Foobar.svg/270px-Foobar.svg.png 1.5x, http://example.com/images/thumb/f/ff/Foobar.svg/360px-Foobar.svg.png 2x" /></a>  <div class="thumbcaption"><div class="magnify"><a href="/wiki/File:Foobar.svg" class="internal" title="Enlarge"></a></div>lang=invalid.language.code</div></div></div>
+<div class="thumb tright"><div class="thumbinner" style="width:182px;"><a href="/wiki/File:Foobar.svg" class="image"><img alt="" src="http://example.com/images/thumb/f/ff/Foobar.svg/180px-Foobar.svg.png" width="180" height="135" class="thumbimage" srcset="http://example.com/images/thumb/f/ff/Foobar.svg/270px-Foobar.svg.png 1.5x, http://example.com/images/thumb/f/ff/Foobar.svg/360px-Foobar.svg.png 2x" /></a>  <div class="thumbcaption"><div class="magnify"><a href="/wiki/File:Foobar.svg" class="internal" title="Enlarge"></a></div>lang=invalid:language:code</div></div></div>
 
 !! html/parsoid
 <figure class="mw-default-size" typeof="mw:Image/Thumb"><a href="./File:Foobar.svg"><img resource="./File:Foobar.svg" src="//example.com/images/f/ff/Foobar.svg" data-file-width="240" data-file-height="180" data-file-type="drawing" height="165" width="220"/></a><figcaption>lang=invalid.language.code</figcaption></figure>
diff --git a/tests/phpunit/includes/media/SVGTest.php b/tests/phpunit/includes/media/SVGTest.php
index 4a986b4cc1..9fd640f56e 100644
--- a/tests/phpunit/includes/media/SVGTest.php
+++ b/tests/phpunit/includes/media/SVGTest.php
@@ -5,6 +5,11 @@
  */
 class SvgTest extends MediaWikiMediaTestCase {
 
+	/**
+	 * @var SvgHandler
+	 */
+	private $handler;
+
 	protected function setUp() {
 		parent::setUp();
 
@@ -38,4 +43,71 @@ class SvgTest extends MediaWikiMediaTestCase {
 			[ 'Wikimedia-logo.svg', [] ]
 		];
 	}
+
+	/**
+	 * @param string $userPreferredLanguage
+	 * @param array $svgLanguages
+	 * @param string $expectedMatch
+	 * @dataProvider providerGetMatchedLanguage
+	 * @covers SvgHandler::getMatchedLanguage
+	 */
+	public function testGetMatchedLanguage( $userPreferredLanguage, $svgLanguages, $expectedMatch ) {
+		$match = $this->handler->getMatchedLanguage( $userPreferredLanguage, $svgLanguages );
+		$this->assertEquals( $expectedMatch, $match );
+	}
+
+	public function providerGetMatchedLanguage() {
+		return [
+			'no match' => [
+				'userPreferredLanguage' => 'en',
+				'svgLanguages' => [ 'de-DE', 'zh', 'ga', 'fr', 'sr-Latn-ME' ],
+				'expectedMatch' => null,
+			],
+			'no subtags' => [
+				'userPreferredLanguage' => 'en',
+				'svgLanguages' => [ 'de', 'zh', 'en', 'fr' ],
+				'expectedMatch' => 'en',
+			],
+			'user no subtags, svg 1 subtag' => [
+				'userPreferredLanguage' => 'en',
+				'svgLanguages' => [ 'de-DE', 'en-GB', 'en-US', 'fr' ],
+				'expectedMatch' => 'en-GB',
+			],
+			'user no subtags, svg >1 subtag' => [
+				'userPreferredLanguage' => 'sr',
+				'svgLanguages' => [ 'de-DE', 'sr-Cyrl-BA', 'sr-Latn-ME', 'en-US', 'fr' ],
+				'expectedMatch' => 'sr-Cyrl-BA',
+			],
+			'user 1 subtag, svg no subtags' => [
+				'userPreferredLanguage' => 'en-US',
+				'svgLanguages' => [ 'de', 'en', 'en', 'fr' ],
+				'expectedMatch' => null,
+			],
+			'user 1 subtag, svg 1 subtag' => [
+				'userPreferredLanguage' => 'en-US',
+				'svgLanguages' => [ 'de-DE', 'en-GB', 'en-US', 'fr' ],
+				'expectedMatch' => 'en-US',
+			],
+			'user 1 subtag, svg >1 subtag' => [
+				'userPreferredLanguage' => 'sr-Latn',
+				'svgLanguages' => [ 'de-DE', 'sr-Cyrl-BA', 'sr-Latn-ME', 'fr' ],
+				'expectedMatch' => 'sr-Latn-ME',
+			],
+			'user >1 subtag, svg >1 subtag' => [
+				'userPreferredLanguage' => 'sr-Latn-ME',
+				'svgLanguages' => [ 'de-DE', 'sr-Cyrl-BA', 'sr-Latn-ME', 'en-US', 'fr' ],
+				'expectedMatch' => 'sr-Latn-ME',
+			],
+			'user >1 subtag, svg <=1 subtag' => [
+				'userPreferredLanguage' => 'sr-Latn-ME',
+				'svgLanguages' => [ 'de-DE', 'sr-Cyrl', 'sr-Latn', 'en-US', 'fr' ],
+				'expectedMatch' => null,
+			],
+			'ensure case-insensitive' => [
+				'userPreferredLanguage' => 'sr-latn',
+				'svgLanguages' => [ 'de-DE', 'sr-Cyrl', 'sr-Latn-ME', 'en-US', 'fr' ],
+				'expectedMatch' => 'sr-Latn-ME',
+			],
+		];
+	}
 }