You are here

Home » API reference » Link checker 6.2 » linkchecker_extract_links.test » LinkCheckerLinkExtractionTest

public function LinkCheckerLinkExtractionTest::testLinkCheckerCreateNodeWithLinks in Link checker 6.2

Same name and namespace in other branches
  1. 7 linkchecker.test \LinkCheckerLinkExtractionTest::testLinkCheckerCreateNodeWithLinks()

File

tests/linkchecker_extract_links.test, line 32
Test file for Link checker module.

Class

LinkCheckerLinkExtractionTest
@file Test file for Link checker module.

Code

public function testLinkCheckerCreateNodeWithLinks() {

  // Enable all node type page for link extraction.
  variable_set('linkchecker_scan_nodetypes', array(
    'page',
  ));
  variable_set('linkchecker_scan_blocks', 1);

  // Core enables the URL filter for "Full HTML" by default.
  // -> Blacklist / Disable URL filter for testing.
  variable_set('linkchecker_filter_blacklist', array(
    'filter/2',
  ));

  // Extract from all link checker supported HTML tags.
  variable_set('linkchecker_extract_from_a', 1);
  variable_set('linkchecker_extract_from_audio', 1);
  variable_set('linkchecker_extract_from_embed', 1);
  variable_set('linkchecker_extract_from_iframe', 1);
  variable_set('linkchecker_extract_from_img', 1);
  variable_set('linkchecker_extract_from_object', 1);
  variable_set('linkchecker_extract_from_source', 1);
  variable_set('linkchecker_extract_from_video', 1);
  $body = <<<EOT
<!-- UNSUPPORTED for link checking: -->

<a href="mailto:test@example.com">Send email</a>
<a href="javascript:foo()">Execute JavaScript</a>

<!-- SUPPORTED for link checking: -->

<!-- URL in HTML comment: http://example.com/test-if-url-filter-is-disabled -->

<!-- Relative URLs -->
<img src="test.png" alt="Test image 1" />
<img src="../foo1/test.png" alt="Test image 2" />

<a href="../foo1/bar1">../foo1/bar1</a>
<a href="./foo2/bar2">./foo2/bar2</a>
<a href="../foo3/../foo4/foo5">../foo3/../foo4/foo5</a>
<a href="./foo4/../foo5/foo6">./foo4/../foo5/foo6</a>
<a href="./foo4/./foo5/foo6">./foo4/./foo5/foo6</a>

<!-- URL with uncommon chars that could potentially fail to extract. See http://drupal.org/node/465462. -->
<a href="http://www.lagrandeepicerie.fr/#e-boutique/Les_produits_du_moment,2/coffret_vins_doux_naturels,149">URL with uncommon chars</a>

<!-- object tag: Embed SWF files -->
<object width="150" height="116"
  type="application/x-shockwave-flash"
  data="http://wetterservice.msn.de/phclip.swf?zip=60329&ort=Frankfurt">
    <param name="movie" value="http://wetterservice.msn.de/phclip.swf?zip=60329&ort=Frankfurt" />
    <img src="flash.png" width="150" height="116" alt="" /> <br />
      No weather report visible? At <a href="http://www.msn.de/">MSN</a>
      you are able to find the weather report missing here and the
      Flash plugin can be found at <a href="http://www.adobe.com/">Adobe</a>.
</object>

<!-- object tag: Embed Quicktime Movies on HTML pages -->
<object width="420" height="282"
  classid="clsid:02BF25D5-8C17-4B23-BC80-D3488ABDDC6B"
  codebase="http://www.apple.com/qtactivex/qtplugin.cab">
  <param name="src" value="http://example.net/video/foo1.mov" />
  <param name="href" value="http://example.net/video/foo2.mov" />
  <param name="controller" value="true" />
  <param name="autoplay" value="false" />
  <param name="scale" value="aspect" />
  <!--[if gte IE 7]> <!-->
  <object type="video/quicktime" data="http://example.net/video/foo3.mov" width="420" height="282">
    <param name="controller" value="true" />
    <param name="autoplay" value="false" />
  </object>
  <!--<![endif]-->
</object>

<!-- object tag: Play MP4 videos on HTML pages -->
<object data="http://example.org/video/foo1.mp4" type="video/mp4" width="420" height="288">
  <param name="src" value="http://example.org/video/foo2.mp4" />
  <param name="autoplay" value="false" />
  <param name="autoStart" value="0" />
  <a href="http://example.org/video/foo3.mp4">/video/foo3.mp4</a>
</object>

<!-- object tag: Play MP4 videos with Quicktime -->
<object width="420" height="282" codebase="http://www.apple.com/qtactivex/qtplugin.cab">
  <param name="src" value="http://example.org/video/foo4.mp4" />
  <param name="href" value="http://example.org/video/foo5.mp4" />
  <param name="controller" value="true" />
  <param name="autoplay" value="false" />
  <param name="scale" value="aspect" />
  <!--[if gte IE 7]> <!-->
  <object type="video/quicktime" data="http://example.org/video/foo6.mp4" width="420" height="282">
    <param name="controller" value="true" />
    <param name="autoplay" value="false" />
  </object>
  <!--<![endif]-->
</object>

<!-- object tag: Play flash videos on HTML pages -->
<object type="application/x-shockwave-flash" data="http://example.org/video/player1.swf" width="420" height="270">
    <param name="movie" value="http://example.org/video/player2.swf" />
    <param src="movie" value="http://example.org/video/player3.swf" />
    <param name="flashvars" value="file=http://example.org/video/foo1.flv&width=420&height=270" />
</object>

<!-- Embed ActiveX control as objekt -->
<object width="267" height="175" classid="CLSID:05589FA1-C356-11CE-BF01-00AA0055595A">
  <param name="filename" value="ritmo.mid">
</object>

<!-- Add inline frames -->
<iframe src="http://example.com/iframe/" name="ExampleIFrame" width="300" height="200">
  <p>Your browser does not support inline frames.</p>
</iframe>

<!-- https://developer.mozilla.org/en/Using_audio_and_video_in_Firefox -->

<!-- http://www.theora.org/cortado/ -->
<video src="my_ogg_video.ogg" controls width="320" height="240">
  <object type="application/x-java-applet" width="320" height="240">
    <param name="archive" value="http://www.theora.org/cortado.jar">
    <param name="code" value="com.fluendo.player.Cortado.class">
    <param name="url" value="my_ogg_video.ogg">
    <p>You need to install Java to play this file.</p>
  </object>
</video>

<video src="video.ogv" controls>
  <object data="flvplayer1.swf" type="application/x-shockwave-flash">
    <param name="movie" value="flvplayer2.swf" />
  </object>
</video>

<video controls>
  <source src="http://v2v.cc/~j/theora_testsuite/pixel_aspect_ratio.ogg" type="video/ogg">
  <source src="http://v2v.cc/~j/theora_testsuite/pixel_aspect_ratio.mov">
  Your browser does not support the <code>video</code> element.
</video>

<video controls>
  <source src="foo.ogg" type="video/ogg; codecs=&quot;dirac, speex&quot;">
  Your browser does not support the <code>video</code> element.
</video>

<video src="http://v2v.cc/~j/theora_testsuite/320x240.ogg" controls>
  Your browser does not support the <code>video</code> element.
</video>
EOT;

  // Save folder names in variables for reuse.
  $folder1 = $this
    ->randomName(10);
  $folder2 = $this
    ->randomName(5);

  // Allow the user to use the 'Full HTML' format.
  db_query("UPDATE {filter_formats} SET roles = '%s' WHERE format = %d", ',2,', 2);

  // Fill node array.
  $edit = array();
  $edit['title'] = $this
    ->randomName(32);
  $edit['body'] = $body;
  $edit['path'] = $folder1 . '/' . $folder2;
  $edit['format'] = 2;

  // Extract only full qualified URLs.
  variable_set('linkchecker_fqdn_only', 1);

  // Verify path input field appears on add "Basic page" form.
  $this
    ->drupalGet('node/add/page');

  // Verify path input is present.
  $this
    ->assertFieldByName('path', '', t('Path input field present on add Basic page form.'));

  // Save node.
  $this
    ->drupalPost('node/add/page', $edit, t('Save'));
  $this
    ->assertRaw(t('@type %title has been created.', array(
    '@type' => 'Page',
    '%title' => $edit['title'],
  )), t('Node was created.'));

  // Verify if the content links are extracted correctly.
  $urls_fqdn = array(
    'http://www.lagrandeepicerie.fr/#e-boutique/Les_produits_du_moment,2/coffret_vins_doux_naturels,149',
    'http://wetterservice.msn.de/phclip.swf?zip=60329&ort=Frankfurt',
    'http://www.msn.de/',
    'http://www.adobe.com/',
    'http://www.apple.com/qtactivex/qtplugin.cab',
    'http://example.net/video/foo1.mov',
    'http://example.net/video/foo2.mov',
    'http://example.net/video/foo3.mov',
    'http://example.org/video/foo1.mp4',
    'http://example.org/video/foo2.mp4',
    'http://example.org/video/foo3.mp4',
    'http://example.org/video/foo4.mp4',
    'http://example.org/video/foo5.mp4',
    'http://example.org/video/foo6.mp4',
    'http://example.org/video/player1.swf',
    'http://example.org/video/player2.swf',
    'http://example.org/video/player3.swf',
    'http://example.com/iframe/',
    'http://www.theora.org/cortado.jar',
    'http://v2v.cc/~j/theora_testsuite/pixel_aspect_ratio.ogg',
    'http://v2v.cc/~j/theora_testsuite/pixel_aspect_ratio.mov',
    'http://v2v.cc/~j/theora_testsuite/320x240.ogg',
  );
  foreach ($urls_fqdn as $org_url => $check_url) {
    $link = $this
      ->getLinkCheckerLink($check_url);
    if ($link) {
      $this
        ->assertIdentical($link->url, $check_url, t('Absolute URL %org_url matches expected result %check_url.', array(
        '%org_url' => $org_url,
        '%check_url' => $check_url,
      )));
    }
    else {
      $this
        ->fail(t('URL %check_url not found.', array(
        '%check_url' => $check_url,
      )));
    }
  }

  // Check if the number of links is correct.
  // - Verifies if all HTML tag regexes matched.
  // - Verifies that the linkchecker filter blacklist works well.
  $urls_in_database = $this
    ->getLinkCheckerLinksCount();
  $urls_expected_count = count($urls_fqdn);
  $this
    ->assertEqual($urls_in_database, $urls_expected_count, t('Found @urls_in_database URLs in database matches expected result of @urls_expected_count.', array(
    '@urls_in_database' => $urls_in_database,
    '@urls_expected_count' => $urls_expected_count,
  )));

  // Extract all URLs including relative path.
  variable_set('clean_url', 1);
  variable_set('linkchecker_fqdn_only', 0);
  $node = $this
    ->drupalGetNodeByTitle($edit['title']);
  $this
    ->assertTrue($node, t('Node found in database.'));
  $this
    ->drupalPost('node/' . $node->nid . '/edit', $edit, t('Save'));
  $this
    ->assertRaw(t('@type %title has been updated.', array(
    '@type' => 'Page',
    '%title' => $edit['title'],
  )));

  // FIXME: Path alias seems not saved!???
  // $this->assertIdentical($node->path, $edit['path'], t('URL alias "@node-path" matches path "@edit-path".', array('@node-path' => $node->path, '@edit-path' => $edit['path'])));
  // DEBUG
  $linkchecker_links = db_query("SELECT * FROM {linkchecker_links}");
  while ($row = db_fetch_array($linkchecker_links)) {
    $rows[] = $row['url'];
  }
  $this
    ->verbose(theme_item_list($rows, 'URLs in database:'));

  // $this->fail('DEBUG: ' .implode('<br />', $rows));
  // Verify if the content links are extracted correctly.
  global $base_root, $base_path;
  $urls_relative = array(
    '../foo1/test.png' => $base_root . $base_path . 'foo1/test.png',
    'test.png' => $base_root . $base_path . $folder1 . '/test.png',
    '../foo1/bar1' => $base_root . $base_path . 'foo1/bar1',
    './foo2/bar2' => $base_root . $base_path . $folder1 . '/foo2/bar2',
    '../foo3/../foo4/foo5' => $base_root . $base_path . 'foo4/foo5',
    './foo4/../foo5/foo6' => $base_root . $base_path . $folder1 . '/foo5/foo6',
    './foo4/./foo5/foo6' => $base_root . $base_path . $folder1 . '/foo4/foo5/foo6',
    'flash.png' => $base_root . $base_path . $folder1 . '/flash.png',
    'ritmo.mid' => $base_root . $base_path . $folder1 . '/ritmo.mid',
    'my_ogg_video.ogg' => $base_root . $base_path . $folder1 . '/my_ogg_video.ogg',
    'video.ogv' => $base_root . $base_path . $folder1 . '/video.ogv',
    'flvplayer1.swf' => $base_root . $base_path . $folder1 . '/flvplayer1.swf',
    'flvplayer2.swf' => $base_root . $base_path . $folder1 . '/flvplayer2.swf',
    'foo.ogg' => $base_root . $base_path . $folder1 . '/foo.ogg',
  );
  $this
    ->verbose(theme_item_list($urls_relative, 'Verify if following relative URLs exists:'));
  foreach ($urls_relative as $org_url => $check_url) {
    $link = $this
      ->getLinkCheckerLink($check_url);
    if ($link) {
      $this
        ->assertIdentical($link->url, $check_url, t('Relative URL %org_url matches expected result %check_url.', array(
        '%org_url' => $org_url,
        '%check_url' => $check_url,
      )));
    }
    else {
      $this
        ->fail(t('URL %check_url not found.', array(
        '%check_url' => $check_url,
      )));
    }
  }

  // Check if the number of links is correct.
  $urls_in_database = $this
    ->getLinkCheckerLinksCount();
  $urls_expected_count = count($urls_fqdn + $urls_relative);
  $this
    ->assertEqual($urls_in_database, $urls_expected_count, t('Found @urls_in_database URLs in database matches expected result of @urls_expected_count.', array(
    '@urls_in_database' => $urls_in_database,
    '@urls_expected_count' => $urls_expected_count,
  )));

  // Verify if link check has been enabled for normal URLs.
  $urls = array(
    'http://www.lagrandeepicerie.fr/#e-boutique/Les_produits_du_moment,2/coffret_vins_doux_naturels,149',
    'http://wetterservice.msn.de/phclip.swf?zip=60329&ort=Frankfurt',
    'http://www.msn.de/',
    'http://www.adobe.com/',
    'http://www.apple.com/qtactivex/qtplugin.cab',
    'http://www.theora.org/cortado.jar',
    'http://v2v.cc/~j/theora_testsuite/pixel_aspect_ratio.ogg',
    'http://v2v.cc/~j/theora_testsuite/pixel_aspect_ratio.mov',
    'http://v2v.cc/~j/theora_testsuite/320x240.ogg',
    $base_root . $base_path . 'foo1/test.png',
    $base_root . $base_path . $folder1 . '/test.png',
    $base_root . $base_path . 'foo1/bar1',
    $base_root . $base_path . $folder1 . '/foo2/bar2',
    $base_root . $base_path . 'foo4/foo5',
    $base_root . $base_path . $folder1 . '/foo5/foo6',
    $base_root . $base_path . $folder1 . '/foo4/foo5/foo6',
    $base_root . $base_path . $folder1 . '/flash.png',
    $base_root . $base_path . $folder1 . '/ritmo.mid',
    $base_root . $base_path . $folder1 . '/my_ogg_video.ogg',
    $base_root . $base_path . $folder1 . '/video.ogv',
    $base_root . $base_path . $folder1 . '/flvplayer1.swf',
    $base_root . $base_path . $folder1 . '/flvplayer2.swf',
    $base_root . $base_path . $folder1 . '/foo.ogg',
  );
  foreach ($urls as $url) {
    $this
      ->assertTrue($this
      ->getLinkcheckerLink($url)->status, t('Link check for %url is enabled.', array(
      '%url' => $url,
    )));
  }

  // Verify if link check has been disabled for example.com/net/org URLs.
  $documentation_urls = array(
    'http://example.net/video/foo1.mov',
    'http://example.net/video/foo2.mov',
    'http://example.net/video/foo3.mov',
    'http://example.org/video/foo1.mp4',
    'http://example.org/video/foo2.mp4',
    'http://example.org/video/foo3.mp4',
    'http://example.org/video/foo4.mp4',
    'http://example.org/video/foo5.mp4',
    'http://example.org/video/foo6.mp4',
    'http://example.org/video/player1.swf',
    'http://example.org/video/player2.swf',
    'http://example.org/video/player3.swf',
    'http://example.com/iframe/',
  );
  foreach ($documentation_urls as $documentation_url) {
    $this
      ->assertFalse($this
      ->getLinkcheckerLink($documentation_url)->status, t('Link check for %url is disabled.', array(
      '%url' => $documentation_url,
    )));
  }
}