X-Git-Url: https://git.immae.eu/?a=blobdiff_plain;ds=sidebyside;f=src%2FWallabag%2FCoreBundle%2FHelper%2FContentProxy.php;h=4b3e6fbb161fd176ad875b5f3b1d8857037d47df;hb=d06a1ebdfb6f70adea935c727799798f2b8ea3f0;hp=7fb41393127c8bcdaab1a11511a988e93bb6cb16;hpb=917040d4a0b7694ca4f695c02243ffb4283d6901;p=github%2Fwallabag%2Fwallabag.git diff --git a/src/Wallabag/CoreBundle/Helper/ContentProxy.php b/src/Wallabag/CoreBundle/Helper/ContentProxy.php index 7fb41393..4b3e6fbb 100644 --- a/src/Wallabag/CoreBundle/Helper/ContentProxy.php +++ b/src/Wallabag/CoreBundle/Helper/ContentProxy.php @@ -3,8 +3,10 @@ namespace Wallabag\CoreBundle\Helper; use Graby\Graby; +use Psr\Log\LoggerInterface; use Wallabag\CoreBundle\Entity\Entry; use Wallabag\CoreBundle\Tools\Utils; +use Symfony\Component\HttpFoundation\File\MimeType\MimeTypeExtensionGuesser; /** * This kind of proxy class take care of getting the content from an url @@ -13,24 +15,46 @@ use Wallabag\CoreBundle\Tools\Utils; class ContentProxy { protected $graby; + protected $tagger; + protected $logger; + protected $mimeGuesser; + protected $fetchingErrorMessage; + protected $eventDispatcher; - public function __construct(Graby $graby) + public function __construct(Graby $graby, RuleBasedTagger $tagger, LoggerInterface $logger, $fetchingErrorMessage) { $this->graby = $graby; + $this->tagger = $tagger; + $this->logger = $logger; + $this->mimeGuesser = new MimeTypeExtensionGuesser(); + $this->fetchingErrorMessage = $fetchingErrorMessage; } /** * Fetch content using graby and hydrate given entry with results information. * In case we couldn't find content, we'll try to use Open Graph data. * - * @param Entry $entry Entry to update - * @param string $url Url to grab content for + * We can also force the content, in case of an import from the v1 for example, so the function won't + * fetch the content from the website but rather use information given with the $content parameter. + * + * @param Entry $entry Entry to update + * @param string $url Url to grab content for + * @param array $content An array with AT LEAST keys title, html, url, language & content_type to skip the fetchContent from the url * * @return Entry */ - public function updateEntry(Entry $entry, $url) + public function updateEntry(Entry $entry, $url, array $content = []) { - $content = $this->graby->fetchContent($url); + // do we have to fetch the content or the provided one is ok? + if (empty($content) || false === $this->validateContent($content)) { + $fetchedContent = $this->graby->fetchContent($url); + + // when content is imported, we have information in $content + // in case fetching content goes bad, we'll keep the imported information instead of overriding them + if (empty($content) || $fetchedContent['html'] !== $this->fetchingErrorMessage) { + $content = $fetchedContent; + } + } $title = $content['title']; if (!$title && isset($content['open_graph']['og_title'])) { @@ -39,7 +63,7 @@ class ContentProxy $html = $content['html']; if (false === $html) { - $html = '

Unable to retrieve readable content.

'; + $html = $this->fetchingErrorMessage; if (isset($content['open_graph']['og_description'])) { $html .= '

But we found a short description:

'; @@ -50,15 +74,60 @@ class ContentProxy $entry->setUrl($content['url'] ?: $url); $entry->setTitle($title); $entry->setContent($html); - $entry->setLanguage($content['language']); - $entry->setMimetype($content['content_type']); + $entry->setHttpStatus(isset($content['status']) ? $content['status'] : ''); + + if (isset($content['date']) && null !== $content['date'] && '' !== $content['date']) { + $entry->setPublishedAt(new \DateTime($content['date'])); + } + + if (!empty($content['authors'])) { + $entry->setPublishedBy($content['authors']); + } + + if (!empty($content['all_headers'])) { + $entry->setHeaders($content['all_headers']); + } + + $entry->setLanguage(isset($content['language']) ? $content['language'] : ''); + $entry->setMimetype(isset($content['content_type']) ? $content['content_type'] : ''); $entry->setReadingTime(Utils::getReadingTime($html)); - $entry->setDomainName(parse_url($entry->getUrl(), PHP_URL_HOST)); - if (isset($content['open_graph']['og_image'])) { + $domainName = parse_url($entry->getUrl(), PHP_URL_HOST); + if (false !== $domainName) { + $entry->setDomainName($domainName); + } + + if (isset($content['open_graph']['og_image']) && $content['open_graph']['og_image']) { $entry->setPreviewPicture($content['open_graph']['og_image']); } + // if content is an image define as a preview too + if (isset($content['content_type']) && in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) { + $entry->setPreviewPicture($content['url']); + } + + try { + $this->tagger->tag($entry); + } catch (\Exception $e) { + $this->logger->error('Error while trying to automatically tag an entry.', [ + 'entry_url' => $url, + 'error_msg' => $e->getMessage(), + ]); + } + return $entry; } + + /** + * Validate that the given content as enough value to be used + * instead of fetch the content from the url. + * + * @param array $content + * + * @return bool true if valid otherwise false + */ + private function validateContent(array $content) + { + return isset($content['title']) && isset($content['html']) && isset($content['url']) && isset($content['language']) && isset($content['content_type']); + } }