X-Git-Url: https://git.immae.eu/?a=blobdiff_plain;f=src%2FWallabag%2FCoreBundle%2FHelper%2FContentProxy.php;h=2628af190b3119a952394f382368c0ab914bfee9;hb=d76a5a6d60b6ee0d1f7efd0c8a70204f821ed99e;hp=fd97e0ff8a92473b42ac97abe06f00a4656ad439;hpb=b236d3f627a21bc9b02e7726bbb72d632937a45e;p=github%2Fwallabag%2Fwallabag.git diff --git a/src/Wallabag/CoreBundle/Helper/ContentProxy.php b/src/Wallabag/CoreBundle/Helper/ContentProxy.php index fd97e0ff..2628af19 100644 --- a/src/Wallabag/CoreBundle/Helper/ContentProxy.php +++ b/src/Wallabag/CoreBundle/Helper/ContentProxy.php @@ -24,8 +24,9 @@ class ContentProxy protected $mimeGuesser; protected $fetchingErrorMessage; protected $eventDispatcher; + protected $storeArticleHeaders; - public function __construct(Graby $graby, RuleBasedTagger $tagger, ValidatorInterface $validator, LoggerInterface $logger, $fetchingErrorMessage) + public function __construct(Graby $graby, RuleBasedTagger $tagger, ValidatorInterface $validator, LoggerInterface $logger, $fetchingErrorMessage, $storeArticleHeaders = false) { $this->graby = $graby; $this->tagger = $tagger; @@ -33,6 +34,7 @@ class ContentProxy $this->logger = $logger; $this->mimeGuesser = new MimeTypeExtensionGuesser(); $this->fetchingErrorMessage = $fetchingErrorMessage; + $this->storeArticleHeaders = $storeArticleHeaders; } /** @@ -51,6 +53,7 @@ class ContentProxy if ((empty($content) || false === $this->validateContent($content)) && false === $disableContentUpdate) { $fetchedContent = $this->graby->fetchContent($url); + $fetchedContent['title'] = $this->sanitizeUTF8Text($fetchedContent['title']); // 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 @@ -66,6 +69,28 @@ class ContentProxy $this->stockEntry($entry, $content); } + /** + * Remove invalid UTF-8 characters from the given string in following steps: + * - try to interpret the given string as ISO-8859-1, convert it to UTF-8 and return it (if its valid) + * - simply remove every invalid UTF-8 character and return the result (https://stackoverflow.com/a/1433665) + * @param String $rawText + * @return string + */ + private function sanitizeUTF8Text(String $rawText) { + if (mb_check_encoding($rawText, 'utf-8')) { + return $rawText; // return because its valid utf-8 text + } + + // we assume that $text is encoded in ISO-8859-1 (and not the similar Windows-1252 or other encoding) + $convertedText = utf8_encode($rawText); + if (mb_check_encoding($convertedText, 'utf-8')) { + return $convertedText; + } + + // last resort: simply remove invalid UTF-8 character because $rawText can have some every exotic encoding + return iconv("UTF-8", "UTF-8//IGNORE", $rawText); + } + /** * Use a Symfony validator to ensure the language is well formatted. * @@ -83,7 +108,7 @@ class ContentProxy (new LocaleConstraint()) ); - if (0 === count($errors)) { + if (0 === \count($errors)) { $entry->setLanguage($value); return; @@ -105,7 +130,7 @@ class ContentProxy (new UrlConstraint()) ); - if (0 === count($errors)) { + if (0 === \count($errors)) { $entry->setPreviewPicture($value); return; @@ -122,20 +147,58 @@ class ContentProxy */ public function updatePublishedAt(Entry $entry, $value) { - $date = $value instanceof \DateTime ? $value->date : $value; + $date = $value; // is it a timestamp? - if (filter_var($date, FILTER_VALIDATE_INT) !== false) { - $date = '@' . $value; + if (false !== filter_var($date, FILTER_VALIDATE_INT)) { + $date = '@' . $date; } try { - $entry->setPublishedAt(new \DateTime($date)); + // is it already a DateTime? + // (it's inside the try/catch in case of fail to be parse time string) + if (!$date instanceof \DateTime) { + $date = new \DateTime($date); + } + + $entry->setPublishedAt($date); } catch (\Exception $e) { $this->logger->warning('Error while defining date', ['e' => $e, 'url' => $entry->getUrl(), 'date' => $value]); } } + /** + * Helper to extract and save host from entry url. + * + * @param Entry $entry + */ + public function setEntryDomainName(Entry $entry) + { + $domainName = parse_url($entry->getUrl(), PHP_URL_HOST); + if (false !== $domainName) { + $entry->setDomainName($domainName); + } + } + + /** + * Helper to set a default title using: + * - url basename, if applicable + * - hostname. + * + * @param Entry $entry + */ + public function setDefaultEntryTitle(Entry $entry) + { + $url = parse_url($entry->getUrl()); + $path = pathinfo($url['path'], PATHINFO_BASENAME); + + if (empty($path)) { + $path = $url['host']; + } + + $entry->setTitle($path); + } + /** * Stock entry with fetched or imported content. * Will fall back to OpenGraph data if available. @@ -147,10 +210,7 @@ class ContentProxy { $entry->setUrl($content['url']); - $domainName = parse_url($entry->getUrl(), PHP_URL_HOST); - if (false !== $domainName) { - $entry->setDomainName($domainName); - } + $this->setEntryDomainName($entry); if (!empty($content['title'])) { $entry->setTitle($content['title']); @@ -175,11 +235,11 @@ class ContentProxy $entry->setHttpStatus($content['status']); } - if (!empty($content['authors']) && is_array($content['authors'])) { + if (!empty($content['authors']) && \is_array($content['authors'])) { $entry->setPublishedBy($content['authors']); } - if (!empty($content['all_headers'])) { + if (!empty($content['all_headers']) && $this->storeArticleHeaders) { $entry->setHeaders($content['all_headers']); } @@ -196,7 +256,7 @@ class ContentProxy } // if content is an image, define it as a preview too - if (!empty($content['content_type']) && in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) { + if (!empty($content['content_type']) && \in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) { $this->updatePreviewPicture($entry, $content['url']); }