X-Git-Url: https://git.immae.eu/?a=blobdiff_plain;f=src%2FWallabag%2FCoreBundle%2FHelper%2FContentProxy.php;h=bfaa197695941af57dde87e751f95d1a1d0cc30c;hb=9fe87bc2e20fa95573287a61ef9798cc15648187;hp=ed4a220dcdbb33e265b6ea2606ef6dbfb1c5d376;hpb=ba2157b270095757eb5af8062df72ce16c497637;p=github%2Fwallabag%2Fwallabag.git
diff --git a/src/Wallabag/CoreBundle/Helper/ContentProxy.php b/src/Wallabag/CoreBundle/Helper/ContentProxy.php
index ed4a220d..bfaa1976 100644
--- a/src/Wallabag/CoreBundle/Helper/ContentProxy.php
+++ b/src/Wallabag/CoreBundle/Helper/ContentProxy.php
@@ -3,11 +3,10 @@
namespace Wallabag\CoreBundle\Helper;
use Graby\Graby;
-use Psr\Log\LoggerInterface as Logger;
+use Psr\Log\LoggerInterface;
use Wallabag\CoreBundle\Entity\Entry;
-use Wallabag\CoreBundle\Entity\Tag;
use Wallabag\CoreBundle\Tools\Utils;
-use Wallabag\CoreBundle\Repository\TagRepository;
+use Symfony\Component\HttpFoundation\File\MimeType\MimeTypeExtensionGuesser;
/**
* This kind of proxy class take care of getting the content from an url
@@ -18,56 +17,104 @@ class ContentProxy
protected $graby;
protected $tagger;
protected $logger;
- protected $tagRepository;
+ protected $mimeGuesser;
+ protected $fetchingErrorMessage;
+ protected $eventDispatcher;
- public function __construct(Graby $graby, RuleBasedTagger $tagger, TagRepository $tagRepository, Logger $logger)
+ public function __construct(Graby $graby, RuleBasedTagger $tagger, LoggerInterface $logger, $fetchingErrorMessage)
{
$this->graby = $graby;
$this->tagger = $tagger;
$this->logger = $logger;
- $this->tagRepository = $tagRepository;
+ $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.
+ * Update entry using either fetched or provided content.
*
- * 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
+ * @param Entry $entry Entry to update
+ * @param string $url Url of the content
+ * @param array $content Array with content provided for import with AT LEAST keys title, html, url to skip the fetchContent from the url
+ * @param bool $disableContentUpdate Whether to skip trying to fetch content using Graby
*/
- public function updateEntry(Entry $entry, $url, array $content = [])
+ public function updateEntry(Entry $entry, $url, array $content = [], $disableContentUpdate = false)
{
- // do we have to fetch the content or the provided one is ok?
- if (empty($content) || false === $this->validateContent($content)) {
- $content = $this->graby->fetchContent($url);
+ if (!empty($content['html'])) {
+ $content['html'] = $this->graby->cleanupHtml($content['html'], $url);
+ }
+
+ if ((empty($content) || false === $this->validateContent($content)) && false === $disableContentUpdate) {
+ $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;
+ }
}
+ // be sure to keep the url in case of error
+ // so we'll be able to refetch it in the future
+ $content['url'] = !empty($content['url']) ? $content['url'] : $url;
+
+ $this->stockEntry($entry, $content);
+ }
+
+ /**
+ * Stock entry with fetched or imported content.
+ * Will fall back to OpenGraph data if available.
+ *
+ * @param Entry $entry Entry to stock
+ * @param array $content Array with at least title, url & html
+ */
+ private function stockEntry(Entry $entry, array $content)
+ {
$title = $content['title'];
- if (!$title && isset($content['open_graph']['og_title'])) {
+ if (!$title && !empty($content['open_graph']['og_title'])) {
$title = $content['open_graph']['og_title'];
}
$html = $content['html'];
if (false === $html) {
- $html = 'Unable to retrieve readable content.
';
+ $html = $this->fetchingErrorMessage;
- if (isset($content['open_graph']['og_description'])) {
+ if (!empty($content['open_graph']['og_description'])) {
$html .= 'But we found a short description:
';
$html .= $content['open_graph']['og_description'];
}
}
- $entry->setUrl($content['url'] ?: $url);
+ $entry->setUrl($content['url']);
$entry->setTitle($title);
$entry->setContent($html);
- $entry->setLanguage($content['language']);
- $entry->setMimetype($content['content_type']);
+ $entry->setHttpStatus(isset($content['status']) ? $content['status'] : '');
+
+ if (!empty($content['date'])) {
+ $date = $content['date'];
+
+ // is it a timestamp?
+ if (filter_var($date, FILTER_VALIDATE_INT) !== false) {
+ $date = '@'.$content['date'];
+ }
+
+ try {
+ $entry->setPublishedAt(new \DateTime($date));
+ } catch (\Exception $e) {
+ $this->logger->warning('Error while defining date', ['e' => $e, 'url' => $content['url'], 'date' => $content['date']]);
+ }
+ }
+
+ if (!empty($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));
$domainName = parse_url($entry->getUrl(), PHP_URL_HOST);
@@ -75,59 +122,27 @@ class ContentProxy
$entry->setDomainName($domainName);
}
- if (isset($content['open_graph']['og_image'])) {
+ if (!empty($content['open_graph']['og_image'])) {
$entry->setPreviewPicture($content['open_graph']['og_image']);
}
+ // if content is an image define as a preview too
+ if (!empty($content['content_type']) && in_array($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.', array(
- 'entry_url' => $url,
+ $this->logger->error('Error while trying to automatically tag an entry.', [
+ 'entry_url' => $content['url'],
'error_msg' => $e->getMessage(),
- ));
- }
-
- return $entry;
- }
-
- /**
- * Assign some tags to an entry.
- *
- * @param Entry $entry
- * @param array|string $tags An array of tag or a string coma separated of tag
- */
- public function assignTagsToEntry(Entry $entry, $tags)
- {
- if (!is_array($tags)) {
- $tags = explode(',', $tags);
- }
-
- foreach ($tags as $label) {
- $label = trim($label);
-
- // avoid empty tag
- if (0 === strlen($label)) {
- continue;
- }
-
- $tagEntity = $this->tagRepository->findOneByLabel($label);
-
- if (is_null($tagEntity)) {
- $tagEntity = new Tag();
- $tagEntity->setLabel($label);
- }
-
- // only add the tag on the entry if the relation doesn't exist
- if (false === $entry->getTags()->contains($tagEntity)) {
- $entry->addTag($tagEntity);
- }
+ ]);
}
}
/**
- * Validate that the given content as enough value to be used
- * instead of fetch the content from the url.
+ * Validate that the given content has at least a title, an html and a url.
*
* @param array $content
*
@@ -135,6 +150,6 @@ class ContentProxy
*/
private function validateContent(array $content)
{
- return isset($content['title']) && isset($content['html']) && isset($content['url']) && isset($content['language']) && isset($content['content_type']);
+ return !empty($content['title']) && !empty($content['html']) && !empty($content['url']);
}
}