X-Git-Url: https://git.immae.eu/?a=blobdiff_plain;f=src%2FWallabag%2FCoreBundle%2FHelper%2FContentProxy.php;h=8019df42a6c51bd642c78074f6eceb6ec0e91657;hb=f74061f1e1e70f8f1ec1fb72f7cf6e1ed388023c;hp=ba90b7310af83fd13774a1d87b38fb8fb241736a;hpb=12b5a8cea12c7c8b10ede7fd00d1f07ca06b1033;p=github%2Fwallabag%2Fwallabag.git diff --git a/src/Wallabag/CoreBundle/Helper/ContentProxy.php b/src/Wallabag/CoreBundle/Helper/ContentProxy.php index ba90b731..8019df42 100644 --- a/src/Wallabag/CoreBundle/Helper/ContentProxy.php +++ b/src/Wallabag/CoreBundle/Helper/ContentProxy.php @@ -32,14 +32,21 @@ class ContentProxy * 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)) { + $content = $this->graby->fetchContent($url); + } $title = $content['title']; if (!$title && isset($content['open_graph']['og_title'])) { @@ -62,7 +69,11 @@ class ContentProxy $entry->setLanguage($content['language']); $entry->setMimetype($content['content_type']); $entry->setReadingTime(Utils::getReadingTime($html)); - $entry->setDomainName(parse_url($entry->getUrl(), PHP_URL_HOST)); + + $domainName = parse_url($entry->getUrl(), PHP_URL_HOST); + if (false !== $domainName) { + $entry->setDomainName($domainName); + } if (isset($content['open_graph']['og_image'])) { $entry->setPreviewPicture($content['open_graph']['og_image']); @@ -71,10 +82,10 @@ class ContentProxy try { $this->tagger->tag($entry); } catch (\Exception $e) { - $this->logger->error('Error while trying to automatically tag an entry.', array( + $this->logger->error('Error while trying to automatically tag an entry.', [ 'entry_url' => $url, 'error_msg' => $e->getMessage(), - )); + ]); } return $entry; @@ -84,14 +95,24 @@ class ContentProxy * Assign some tags to an entry. * * @param Entry $entry - * @param array|string $tags An array of tag or a string coma separated of tag + * @param array|string $tags An array of tag or a string coma separated of tag + * @param array $entitiesReady Entities from the EntityManager which are persisted but not yet flushed + * It is mostly to fix duplicate tag on import @see http://stackoverflow.com/a/7879164/569101 */ - public function assignTagsToEntry(Entry $entry, $tags) + public function assignTagsToEntry(Entry $entry, $tags, array $entitiesReady = []) { if (!is_array($tags)) { $tags = explode(',', $tags); } + // keeps only Tag entity from the "not yet flushed entities" + $tagsNotYetFlushed = []; + foreach ($entitiesReady as $entity) { + if ($entity instanceof Tag) { + $tagsNotYetFlushed[$entity->getLabel()] = $entity; + } + } + foreach ($tags as $label) { $label = trim($label); @@ -100,11 +121,15 @@ class ContentProxy continue; } - $tagEntity = $this->tagRepository->findOneByLabel($label); + if (isset($tagsNotYetFlushed[$label])) { + $tagEntity = $tagsNotYetFlushed[$label]; + } else { + $tagEntity = $this->tagRepository->findOneByLabel($label); - if (is_null($tagEntity)) { - $tagEntity = new Tag(); - $tagEntity->setLabel($label); + if (is_null($tagEntity)) { + $tagEntity = new Tag(); + $tagEntity->setLabel($label); + } } // only add the tag on the entry if the relation doesn't exist @@ -113,4 +138,17 @@ class ContentProxy } } } + + /** + * 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']); + } }