Fixes and unit tests for UI language detection

composer.json

@@ -42,7 +42,7 @@
     "twig/extensions": "1.5.*",
     "twbs/bootstrap": "5.1.*",
     "twitter/typeahead.js": "v0.11.*",
-    "willdurand/negotiation": "3.0.*",
+    "willdurand/negotiation": "3.1.*",
     "vakata/jstree": "3.3.*",
     "punic/punic": "3.5.1",
     "ml/json-ld": "1.*",

controller/WebController.php

@@ -76,14 +76,16 @@ public function __construct($model)
     /**
      * Guess the language of the user. Return a language string that is one
      * of the supported languages defined in the $LANGUAGES setting, e.g. "fi".
+     * @param Request $request HTTP request
      * @param string $vocid identifier for the vocabulary eg. 'yso'.
      * @return string returns the language choice as a numeric string value
      */
-    public function guessLanguage($vocid = null)
+    public function guessLanguage($request, $vocid = null)
     {
         // 1. select language based on SKOSMOS_LANGUAGE cookie
-        if (filter_input(INPUT_COOKIE, 'SKOSMOS_LANGUAGE', FILTER_SANITIZE_FULL_SPECIAL_CHARS)) {
-            return filter_input(INPUT_COOKIE, 'SKOSMOS_LANGUAGE', FILTER_SANITIZE_FULL_SPECIAL_CHARS);
+        $languageCookie = $request->getCookie('SKOSMOS_LANGUAGE');
+        if ($languageCookie) {
+            return $languageCookie;
         }
 
         // 2. if vocabulary given, select based on the default language of the vocabulary
@@ -101,9 +103,10 @@ public function guessLanguage($vocid = null)
         $this->negotiator = new \Negotiation\LanguageNegotiator();
         $langcodes = array_keys($this->languages);
         // using a random language from the configured UI languages when there is no accept language header set
-        $acceptLanguage = filter_input(INPUT_SERVER, 'HTTP_ACCEPT_LANGUAGE', FILTER_SANITIZE_FULL_SPECIAL_CHARS) ? filter_input(INPUT_SERVER, 'HTTP_ACCEPT_LANGUAGE', FILTER_SANITIZE_FULL_SPECIAL_CHARS) : $langcodes[0];
+        $acceptLanguage = $request->getServerConstant('HTTP_ACCEPT_LANGUAGE') ? $request->getServerConstant('HTTP_ACCEPT_LANGUAGE') : $langcodes[0];
+
         $bestLang = $this->negotiator->getBest($acceptLanguage, $langcodes);
-        if (isset($bestLang) && in_array($bestLang, $langcodes)) {
+        if (isset($bestLang) && in_array($bestLang->getValue(), $langcodes)) {
             return $bestLang->getValue();
         }
 

dockerfiles/config/config-docker-compose.ttl

@@ -37,7 +37,7 @@
         [ rdfs:label "da" ; rdf:value "da_DK.utf8" ]
         [ rdfs:label "de" ; rdf:value "de_DE.utf8" ]
         [ rdfs:label "en" ; rdf:value "en_GB.utf8" ]
-        [ rdfs:label "en_US" ; rdf:value "en_US.utf8" ]
+        [ rdfs:label "en-US" ; rdf:value "en_US.utf8" ]
         [ rdfs:label "es" ; rdf:value "es_ES.utf8" ]
         [ rdfs:label "fa" ; rdf:value "fa_IR.utf8" ]
         [ rdfs:label "fi" ; rdf:value "fi_FI.utf8" ]
@@ -48,7 +48,7 @@
         [ rdfs:label "nn" ; rdf:value "nn_NO.utf8" ]
         [ rdfs:label "pl" ; rdf:value "pl_PL.utf8" ]
         [ rdfs:label "pt" ; rdf:value "pt_PT.utf8" ]
-        [ rdfs:label "pt_BR" ; rdf:value "pt_BR.utf8" ]
+        [ rdfs:label "pt-BR" ; rdf:value "pt_BR.utf8" ]
         [ rdfs:label "ru" ; rdf:value "ru_RU.utf8" ]
         [ rdfs:label "sv" ; rdf:value "sv_SE.utf8" ]
         [ rdfs:label "zh" ; rdf:value "zh_CN.utf8" ]

dockerfiles/config/config-docker.ttl

@@ -37,7 +37,7 @@
         [ rdfs:label "da" ; rdf:value "da_DK.utf8" ]
         [ rdfs:label "de" ; rdf:value "de_DE.utf8" ]
         [ rdfs:label "en" ; rdf:value "en_GB.utf8" ]
-        [ rdfs:label "en_US" ; rdf:value "en_US.utf8" ]
+        [ rdfs:label "en-US" ; rdf:value "en_US.utf8" ]
         [ rdfs:label "es" ; rdf:value "es_ES.utf8" ]
         [ rdfs:label "fa" ; rdf:value "fa_IR.utf8" ]
         [ rdfs:label "fi" ; rdf:value "fi_FI.utf8" ]
@@ -48,7 +48,7 @@
         [ rdfs:label "nn" ; rdf:value "nn_NO.utf8" ]
         [ rdfs:label "pl" ; rdf:value "pl_PL.utf8" ]
         [ rdfs:label "pt" ; rdf:value "pt_PT.utf8" ]
-        [ rdfs:label "pt_BR" ; rdf:value "pt_BR.utf8" ]
+        [ rdfs:label "pt-BR" ; rdf:value "pt_BR.utf8" ]
         [ rdfs:label "ru" ; rdf:value "ru_RU.utf8" ]
         [ rdfs:label "sv" ; rdf:value "sv_SE.utf8" ]
         [ rdfs:label "zh" ; rdf:value "zh_CN.utf8" ]

index.php

@@ -26,7 +26,7 @@
 if (sizeof($parts) <= 2) {
     // if language code missing, redirect to guessed language
     // in any case, redirect to <lang>/
-    $lang = sizeof($parts) == 2 && $parts[1] !== '' ? $parts[1] : $controller->guessLanguage();
+    $lang = sizeof($parts) == 2 && $parts[1] !== '' ? $parts[1] : $controller->guessLanguage($request);
     header("Location: " . $lang . "/");
 } else {
   if (array_key_exists($parts[1], $config->getLanguages())) { // global pages
@@ -50,12 +50,12 @@
         try {
             $request->setVocab($parts[1]);
         } catch (Exception | ValueError $e) {
-            $request->setLang($controller->guessLanguage());
+            $request->setLang($controller->guessLanguage($request));
             $controller->invokeGenericErrorPage($request);
             return;
         }
         if (sizeof($parts) == 3) { // language code missing
-            $lang = $controller->guessLanguage();
+            $lang = $controller->guessLanguage($request);
             $newurl = $controller->getBaseHref() . $vocab . "/" . $lang . "/";
             header("Location: " . $newurl);
         } else {
@@ -97,7 +97,7 @@
                     $controller->invokeGenericErrorPage($request);
                 }
             } else { // language code missing, redirect to some language version
-                $lang = $controller->guessLanguage($vocab);
+                $lang = $controller->guessLanguage($request, $vocab);
                 $newurl = $controller->getBaseHref() . $vocab . "/" . $lang . "/" . implode('/', array_slice($parts, 2));
                 $qs = $request->getServerConstant('QUERY_STRING');
                 if ($qs) {

model/Request.php

@@ -17,6 +17,7 @@ class Request
     private $queryParams;
     private $queryParamsPOST;
     private $serverConstants;
+    private $cookies;
 
     /**
      * Initializes the Request Object
@@ -45,6 +46,13 @@ public function __construct($model)
         foreach (filter_input_array(INPUT_SERVER) ?: [] as $key => $val) {
             $this->serverConstants[$key] = $val;
         }
+
+        // Store cookies in a local array, so we can mock them in tests.
+        // We do not apply any filters at this point.
+        $this->cookies = [];
+        foreach (filter_input_array(INPUT_COOKIE) ?: [] as $key => $val) {
+            $this->cookies[$key] = $val;
+        }
     }
 
     /**
@@ -67,6 +75,16 @@ public function setServerConstant($paramName, $value)
         $this->serverConstants[$paramName] = $value;
     }
 
+    /**
+     * Set a cookie to mock it in tests.
+     * @param string $paramName parameter name
+     * @param string $value parameter value
+     */
+    public function setCookie($paramName, $value)
+    {
+        $this->cookies[$paramName] = $value;
+    }
+
     /**
      * Return the requested GET query parameter as a string. Backslashes are stripped for security reasons.
      * @param string $paramName parameter name
@@ -110,6 +128,12 @@ public function getServerConstant($paramName)
         return filter_var($this->serverConstants[$paramName], FILTER_SANITIZE_FULL_SPECIAL_CHARS);
     }
 
+    public function getCookie($paramName)
+    {
+        if (!isset($this->cookies[$paramName])) return null;
+        return filter_var($this->cookies[$paramName], FILTER_SANITIZE_FULL_SPECIAL_CHARS);
+    }
+
     public function getLang()
     {
         return $this->lang;

tests/GlobalConfigTest.php

@@ -58,7 +58,7 @@ public function testGetBaseHref()
 
     public function testGetLanguages()
     {
-        $this->assertEquals(array('en' => 'en_GB.utf8'), $this->config->getLanguages());
+        $this->assertEquals(array('en' => 'en_GB.utf8', 'fi' => 'fi_FI.utf8', 'fr' => 'fr_FR.utf8'), $this->config->getLanguages());
     }
 
     public function testGetSearchResultsSize()

tests/WebControllerTest.php

@@ -226,4 +226,40 @@ public function testFormatChangeList() {
         $expected = array ('hurr durr' => array ('uri' => 'http://www.skosmos.skos/changes/d3', 'prefLabel' => 'Hurr Durr', 'date' => DateTime::__set_state(array('date' => '2010-02-12 10:26:39.000000', 'timezone_type' => 3, 'timezone' => 'UTC')), 'datestring' => 'Feb 12, 2010'), 'second date' => array ('uri' => 'http://www.skosmos.skos/changes/d2', 'prefLabel' => 'Second date', 'date' => DateTime::__set_state(array('date' => '2010-02-12 15:26:39.000000', 'timezone_type' => 3, 'timezone' => 'UTC')), 'datestring' => 'Feb 12, 2010'));
         $this->assertEquals($expected, $months['February 2010']);
     }
+
+    public function testGuessLanguageFirstInConfig() {
+        $request = new Request($this->model);
+        $guessedLanguage = $this->webController->guessLanguage($request);
+        $this->assertEquals($guessedLanguage, 'en');
+    }
+
+    public function testGuessLanguageCookie() {
+        $request = new Request($this->model);
+        $request->setCookie('SKOSMOS_LANGUAGE', 'fr');
+        $guessedLanguage = $this->webController->guessLanguage($request);
+        $this->assertEquals($guessedLanguage, 'fr');
+    }
+
+    public function testGuessLanguageVocabDefault() {
+        $request = new Request($this->model);
+        $guessedLanguage = $this->webController->guessLanguage($request, 'groups');
+        $this->assertEquals($guessedLanguage, 'fi');
+    }
+
+    public function testGuessLanguageAcceptLanguageSimple() {
+        $request = new Request($this->model);
+        $request->setServerConstant('HTTP_ACCEPT_LANGUAGE', 'fr');
+        $guessedLanguage = $this->webController->guessLanguage($request);
+        $this->assertEquals($guessedLanguage, 'fr');
+    }
+
+    public function testGuessLanguageAcceptLanguageBestMatch() {
+        $request = new Request($this->model);
+        $request->setServerConstant('HTTP_ACCEPT_LANGUAGE', 'sv, de;q=0.9, fi;q=0.8, fr;q=0.5');
+        $guessedLanguage = $this->webController->guessLanguage($request);
+        // configured/available languages are en, fi, fr
+        // the best matching language for the given Accept-Language is fi
+        $this->assertEquals($guessedLanguage, 'fi');
+    }
+
 }

tests/testconfig.ttl

@@ -36,7 +36,9 @@
     # customize the base element. Set this if the automatic base url detection doesn't work. For example setups behind a proxy.
     skosmos:baseHref "http://tests.localhost/Skosmos/" ;
     # interface languages available, and the corresponding system locales
-    skosmos:languages ( [ rdfs:label "en" ; rdf:value "en_GB.utf8" ] ) ;
+    skosmos:languages ( [ rdfs:label "en" ; rdf:value "en_GB.utf8" ]
+                        [ rdfs:label "fi" ; rdf:value "fi_FI.utf8" ]
+                        [ rdfs:label "fr" ; rdf:value "fr_FR.utf8" ] ) ;
     # how many results (maximum) to load at a time on the search results page
     skosmos:searchResultsSize 5 ;
     # how many items (maximum) to retrieve in transitive property queries