src/Wallabag/CoreBundle/Helper/ContentProxy.php

   1 <?php
   2
   3 namespace Wallabag\CoreBundle\Helper;
   4
   5 use Graby\Graby;
   6 use Psr\Log\LoggerInterface;
   7 use Symfony\Component\EventDispatcher\EventDispatcher;
   8 use Wallabag\CoreBundle\Entity\Entry;
   9 use Wallabag\CoreBundle\Tools\Utils;
  10 use Symfony\Component\HttpFoundation\File\MimeType\MimeTypeExtensionGuesser;
  11
  12 /**
  13  * This kind of proxy class take care of getting the content from an url
  14  * and update the entry with what it found.
  15  */
  16 class ContentProxy
  17 {
  18     protected $graby;
  19     protected $tagger;
  20     protected $logger;
  21     protected $mimeGuesser;
  22     protected $fetchingErrorMessage;
  23     protected $eventDispatcher;
  24
  25     public function __construct(Graby $graby, RuleBasedTagger $tagger, LoggerInterface $logger, $fetchingErrorMessage)
  26     {
  27         $this->graby = $graby;
  28         $this->tagger = $tagger;
  29         $this->logger = $logger;
  30         $this->mimeGuesser = new MimeTypeExtensionGuesser();
  31         $this->fetchingErrorMessage = $fetchingErrorMessage;
  32     }
  33
  34     /**
  35      * Fetch content using graby and hydrate given entry with results information.
  36      * In case we couldn't find content, we'll try to use Open Graph data.
  37      *
  38      * We can also force the content, in case of an import from the v1 for example, so the function won't
  39      * fetch the content from the website but rather use information given with the $content parameter.
  40      *
  41      * @param Entry  $entry   Entry to update
  42      * @param string $url     Url to grab content for
  43      * @param array  $content An array with AT LEAST keys title, html, url, language & content_type to skip the fetchContent from the url
  44      *
  45      * @return Entry
  46      */
  47     public function updateEntry(Entry $entry, $url, array $content = [])
  48     {
  49         // do we have to fetch the content or the provided one is ok?
  50         if (empty($content) || false === $this->validateContent($content)) {
  51             $fetchedContent = $this->graby->fetchContent($url);
  52
  53             // when content is imported, we have information in $content
  54             // in case fetching content goes bad, we'll keep the imported information instead of overriding them
  55             if (empty($content) || $fetchedContent['html'] !== $this->fetchingErrorMessage) {
  56                 $content = $fetchedContent;
  57             }
  58         }
  59
  60         $title = $content['title'];
  61         if (!$title && isset($content['open_graph']['og_title'])) {
  62             $title = $content['open_graph']['og_title'];
  63         }
  64
  65         $html = $content['html'];
  66         if (false === $html) {
  67             $html = $this->fetchingErrorMessage;
  68
  69             if (isset($content['open_graph']['og_description'])) {
  70                 $html .= '<p><i>But we found a short description: </i></p>';
  71                 $html .= $content['open_graph']['og_description'];
  72             }
  73         }
  74
  75         $entry->setUrl($content['url'] ?: $url);
  76         $entry->setTitle($title);
  77         $entry->setContent($html);
  78         $entry->setHttpStatus(isset($content['status']) ? $content['status'] : '');
  79
  80         if (isset($content['date']) && null !== $content['date'] && '' !== $content['date']) {
  81             $entry->setPublishedAt(new \DateTime($content['date']));
  82         }
  83
  84         if (!empty($content['authors'])) {
  85             $entry->setPublishedBy($content['authors']);
  86         }
  87
  88         if (!empty($content['all_headers'])) {
  89             $entry->setHeaders($content['all_headers']);
  90         }
  91
  92         $entry->setLanguage(isset($content['language']) ? $content['language'] : '');
  93         $entry->setMimetype(isset($content['content_type']) ? $content['content_type'] : '');
  94         $entry->setReadingTime(Utils::getReadingTime($html));
  95
  96         $domainName = parse_url($entry->getUrl(), PHP_URL_HOST);
  97         if (false !== $domainName) {
  98             $entry->setDomainName($domainName);
  99         }
 100
 101         if (isset($content['open_graph']['og_image']) && $content['open_graph']['og_image']) {
 102             $entry->setPreviewPicture($content['open_graph']['og_image']);
 103         }
 104
 105         // if content is an image define as a preview too
 106         if (isset($content['content_type']) && in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) {
 107             $entry->setPreviewPicture($content['url']);
 108         }
 109
 110         try {
 111             $this->tagger->tag($entry);
 112         } catch (\Exception $e) {
 113             $this->logger->error('Error while trying to automatically tag an entry.', [
 114                 'entry_url' => $url,
 115                 'error_msg' => $e->getMessage(),
 116             ]);
 117         }
 118
 119         return $entry;
 120     }
 121
 122     /**
 123      * Validate that the given content as enough value to be used
 124      * instead of fetch the content from the url.
 125      *
 126      * @param array $content
 127      *
 128      * @return bool true if valid otherwise false
 129      */
 130     private function validateContent(array $content)
 131     {
 132         return isset($content['title']) && isset($content['html']) && isset($content['url']) && isset($content['language']) && isset($content['content_type']);
 133     }
 134 }