WIP: HTML Validity Tests

_test/tests/general/general_html.test.php

@@ -0,0 +1,118 @@
+<?php
+
+/**
+ * Check some page output for validity
+ *
+ * @group internet
+ */
+class general_html_test extends DokuWikiTest
+{
+    /** @var string[] we consider these hits shortcomings in the validator and not errors */
+    protected $allowedErrors = [
+        'The string “ugc” is not a registered keyword.',
+    ];
+
+    /**
+     * List of requests to check for validity
+     *
+     * @return array
+     */
+    public function requestProvider()
+    {
+        return [
+            ['/doku.php', 'GET', []],
+            ['/doku.php', 'GET', ['do' => 'recent']],
+            ['/doku.php', 'GET', ['do' => 'index']],
+            ['/doku.php', 'GET', ['do' => 'login']],
+            ['/doku.php', 'GET', ['do' => 'search', 'q' => 'wiki']],
+            ['/doku.php', 'GET', ['id' => 'wiki:syntax']],
+            ['/doku.php', 'GET', ['id' => 'wiki:syntax', 'ns' => 'wiki', 'image' => 'wiki:dokuwiki-128.png', 'do' => 'media']],
+            ['/lib/exe/detail.php', 'GET', ['id' => 'wiki:syntax', 'media' => 'wiki:dokuwiki-128.png']],
+        ];
+    }
+
+    /**
+     * Sends the given HTML to the validator and returns the result
+     *
+     * @param string $html
+     * @return array
+     * @throws Exception when communication failed
+     */
+    protected function validate($html)
+    {
+        $http = new \dokuwiki\HTTP\DokuHTTPClient();
+        $http->headers['Content-Type'] = 'text/html; charset=utf-8';
+        $result = $http->post('https://validator.w3.org/nu/?out=json&level=error', $html);
+
+        if ($result === false) {
+            throw new \Exception($http->error);
+        }
+
+        $result = json_decode($result, true);
+        if ($result === null) {
+            throw new \Exception('could not decode JSON');
+        }
+
+        return $result;
+    }
+
+    /**
+     * Reformat the errors for nicer display in output
+     *
+     * @param array $result
+     * @return string[]
+     */
+    protected function listErrors($result)
+    {
+        $errors = [];
+        foreach ($result['messages'] as $msg) {
+            if ($this->isAllowedError($msg['message'])) continue;
+            $errors[] = "☛ " . $msg['message'] . "\n" . $msg['extract'] . "\n";
+        }
+        return $errors;
+    }
+
+    /**
+     * Is the given string an allowed error that should be skipped?
+     *
+     * @param string $string
+     * @return bool
+     */
+    protected function isAllowedError($string)
+    {
+        $re = join('|', array_map('preg_quote_cb', $this->allowedErrors));
+        return (bool)preg_match("/$re/", $string);
+    }
+
+    /**
+     * @dataProvider requestProvider
+     * @param string $url
+     * @param string $method
+     * @param array $data
+     * @group internet
+     */
+    public function test_Validity($url, $method, $data)
+    {
+        $request = new TestRequest();
+        if ($method == 'GET') {
+            $response = $request->get($data, $url);
+        } elseif ($method == 'POST') {
+            $response = $request->post($data, $url);
+        } else {
+            throw new \RuntimeException("unknown method given: $method");
+        }
+
+        $html = $response->getContent();
+        try {
+            $result = $this->validate($html);
+        } catch (\Exception $e) {
+            $this->markTestSkipped($e->getMessage());
+            return;
+        }
+
+        $errors = $this->listErrors($result);
+        $info = "Invalid HTML found:\n" . join("\n", $errors);
+
+        $this->assertEquals(0, count($errors), $info);
+    }
+}

inc/changelog.php

@@ -228,9 +228,9 @@ function getRecents($first,$num,$ns='',$flags=0){
 
     // read all recent changes. (kept short)
     if ($flags & RECENTS_MEDIA_CHANGES) {
-        $lines = @file($conf['media_changelog']);
+        $lines = @file($conf['media_changelog']) ?: [];
     } else {
-        $lines = @file($conf['changelog']);
+        $lines = @file($conf['changelog']) ?: [];
     }
     if (!is_array($lines)) {
         $lines = array();
@@ -240,7 +240,7 @@ function getRecents($first,$num,$ns='',$flags=0){
     $media_lines = array();
 
     if ($flags & RECENTS_MEDIA_PAGES_MIXED) {
-        $media_lines = @file($conf['media_changelog']);
+        $media_lines = @file($conf['media_changelog']) ?: [];
         if (!is_array($media_lines)) {
             $media_lines = array();
         }

inc/html.php

@@ -50,7 +50,7 @@ function html_login($svg = false){
 
     print p_locale_xhtml('login');
     print '<div class="centeralign">'.NL;
-    $form = new Doku_Form(array('id' => 'dw__login'));
+    $form = new Doku_Form(array('id' => 'dw__login', 'action'=>wl($ID)));
     $form->startFieldset($lang['btn_login']);
     $form->addHidden('id', $ID);
     $form->addHidden('do', 'login');
@@ -707,7 +707,7 @@ function html_recent($first = 0, $show_changes = 'both') {
             '</p></div>';
     }
 
-    $form = new Doku_Form(array('id' => 'dw__recent', 'method' => 'GET', 'class' => 'changes'));
+    $form = new Doku_Form(array('id' => 'dw__recent', 'method' => 'GET', 'class' => 'changes', 'action'=>wl($ID)));
     $form->addHidden('sectok', null);
     $form->addHidden('do', 'recent');
     $form->addHidden('id', $ID);