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 Wallabag\CoreBundle\Entity\Entry;
   8 use Wallabag\CoreBundle\Tools\Utils;
   9 use Symfony\Component\HttpFoundation\File\MimeType\MimeTypeExtensionGuesser;
  10
  11 /**
  12  * This kind of proxy class take care of getting the content from an url
  13  * and update the entry with what it found.
  14  */
  15 class ContentProxy
  16 {
  17     protected $graby;
  18     protected $tagger;
  19     protected $logger;
  20     protected $mimeGuesser;
  21     protected $fetchingErrorMessage;
  22     protected $eventDispatcher;
  23
  24     public function __construct(Graby $graby, RuleBasedTagger $tagger, LoggerInterface $logger, $fetchingErrorMessage)
  25     {
  26         $this->graby = $graby;
  27         $this->tagger = $tagger;
  28         $this->logger = $logger;
  29         $this->mimeGuesser = new MimeTypeExtensionGuesser();
  30         $this->fetchingErrorMessage = $fetchingErrorMessage;
  31     }
  32
  33     /**
  34      * Fetch content using graby and hydrate given $entry with results information.
  35      * In case we couldn't find content, we'll try to use Open Graph data.
  36      *
  37      * We can also force the content, in case of an import from the v1 for example, so the function won't
  38      * fetch the content from the website but rather use information given with the $content parameter.
  39      *
  40      * @param Entry  $entry   Entry to update
  41      * @param string $url     Url to grab content for
  42      * @param array  $content An array with AT LEAST keys title, html, url to skip the fetchContent from the url
  43      */
  44     public function updateEntry(Entry $entry, $url, array $content = [])
  45     {
  46         // ensure content is a bit cleaned up
  47         if (!empty($content['html'])) {
  48             $content['html'] = $this->graby->cleanupHtml($content['html'], $url);
  49         }
  50
  51         // do we have to fetch the content or the provided one is ok?
  52         if (empty($content) || false === $this->validateContent($content)) {
  53             $fetchedContent = $this->graby->fetchContent($url);
  54
  55             // when content is imported, we have information in $content
  56             // in case fetching content goes bad, we'll keep the imported information instead of overriding them
  57             if (empty($content) || $fetchedContent['html'] !== $this->fetchingErrorMessage) {
  58                 $content = $fetchedContent;
  59             }
  60         }
  61
  62         $title = $content['title'];
  63         if (!$title && !empty($content['open_graph']['og_title'])) {
  64             $title = $content['open_graph']['og_title'];
  65         }
  66
  67         $html = $content['html'];
  68         if (false === $html) {
  69             $html = $this->fetchingErrorMessage;
  70
  71             if (!empty($content['open_graph']['og_description'])) {
  72                 $html .= '<p><i>But we found a short description: </i></p>';
  73                 $html .= $content['open_graph']['og_description'];
  74             }
  75         }
  76
  77         $entry->setUrl($content['url'] ?: $url);
  78         $entry->setTitle($title);
  79         $entry->setContent($html);
  80         $entry->setHttpStatus(isset($content['status']) ? $content['status'] : '');
  81
  82         if (!empty($content['date'])) {
  83             $date = $content['date'];
  84
  85             // is it a timestamp?
  86             if (filter_var($date, FILTER_VALIDATE_INT) !== false) {
  87                 $date = '@'.$content['date'];
  88             }
  89
  90             try {
  91                 $entry->setPublishedAt(new \DateTime($date));
  92             } catch (\Exception $e) {
  93                 $this->logger->warning('Error while defining date', ['e' => $e, 'url' => $url, 'date' => $content['date']]);
  94             }
  95         }
  96
  97         if (!empty($content['authors'])) {
  98             $entry->setPublishedBy($content['authors']);
  99         }
 100
 101         if (!empty($content['all_headers'])) {
 102             $entry->setHeaders($content['all_headers']);
 103         }
 104
 105         $entry->setLanguage(isset($content['language']) ? $content['language'] : '');
 106         $entry->setMimetype(isset($content['content_type']) ? $content['content_type'] : '');
 107         $entry->setReadingTime(Utils::getReadingTime($html));
 108
 109         $domainName = parse_url($entry->getUrl(), PHP_URL_HOST);
 110         if (false !== $domainName) {
 111             $entry->setDomainName($domainName);
 112         }
 113
 114         if (!empty($content['open_graph']['og_image'])) {
 115             $entry->setPreviewPicture($content['open_graph']['og_image']);
 116         }
 117
 118         // if content is an image define as a preview too
 119         if (!empty($content['content_type']) && in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) {
 120             $entry->setPreviewPicture($content['url']);
 121         }
 122
 123         try {
 124             $this->tagger->tag($entry);
 125         } catch (\Exception $e) {
 126             $this->logger->error('Error while trying to automatically tag an entry.', [
 127                 'entry_url' => $url,
 128                 'error_msg' => $e->getMessage(),
 129             ]);
 130         }
 131     }
 132
 133     /**
 134      * Validate that the given content as enough value to be used
 135      * instead of fetch the content from the url.
 136      *
 137      * @param array $content
 138      *
 139      * @return bool true if valid otherwise false
 140      */
 141     private function validateContent(array $content)
 142     {
 143         return !empty($content['title']) && !empty($content['html']) && !empty($content['url']);
 144     }
 145 }