X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=src%2FUtil%2FParseUrl.php;h=24be55cab931ced253b360a51070f93852dcb0d0;hb=17d49422b4f7b3d1ed06859f996876a85028d87b;hp=3ed48a077f79037d017d7361a64c6ab7bac6e5d1;hpb=fbbab822a0582601389cea43d5a2f8d78f94b972;p=friendica.git diff --git a/src/Util/ParseUrl.php b/src/Util/ParseUrl.php index 3ed48a077f..24be55cab9 100644 --- a/src/Util/ParseUrl.php +++ b/src/Util/ParseUrl.php @@ -1,6 +1,6 @@ head($url); + $curlResult = DI::httpClient()->head($url); + + // Workaround for systems that can't handle a HEAD request + if (!$curlResult->isSuccess() && ($curlResult->getReturnCode() == 405)) { + $curlResult = DI::httpClient()->get($url, [HttpClientOptions::CONTENT_LENGTH => 1000000]); + } + if (!$curlResult->isSuccess()) { return []; } - $contenttype = $curlResult->getHeader('Content-Type'); + $contenttype = $curlResult->getHeader('Content-Type')[0] ?? ''; if (empty($contenttype)) { - return []; + return ['application', 'octet-stream']; } return explode('/', current(explode(';', $contenttype))); @@ -197,7 +204,7 @@ class ParseUrl ]; if ($count > 10) { - Logger::log('Endless loop detected for ' . $url, Logger::DEBUG); + Logger::notice('Endless loop detected', ['url' => $url]); return $siteinfo; } @@ -213,26 +220,21 @@ class ParseUrl return $siteinfo; } - $curlResult = DI::httpRequest()->get($url); - if (!$curlResult->isSuccess()) { + $curlResult = DI::httpClient()->get($url, [HttpClientOptions::CONTENT_LENGTH => 1000000]); + if (!$curlResult->isSuccess() || empty($curlResult->getBody())) { + Logger::info('Empty body or error when fetching', ['url' => $url, 'success' => $curlResult->isSuccess(), 'code' => $curlResult->getReturnCode()]); return $siteinfo; } $siteinfo['expires'] = DateTimeFormat::utc(self::DEFAULT_EXPIRATION_SUCCESS); - // If the file is too large then exit - if (($curlResult->getInfo()['download_content_length'] ?? 0) > 1000000) { - return $siteinfo; - } - - if ($cacheControlHeader = $curlResult->getHeader('Cache-Control')) { + if ($cacheControlHeader = $curlResult->getHeader('Cache-Control')[0] ?? '') { if (preg_match('/max-age=([0-9]+)/i', $cacheControlHeader, $matches)) { $maxAge = max(86400, (int)array_pop($matches)); $siteinfo['expires'] = DateTimeFormat::utc("now + $maxAge seconds"); } } - $header = $curlResult->getHeader(); $body = $curlResult->getBody(); if ($do_oembed) { @@ -273,7 +275,7 @@ class ParseUrl $charset = ''; // Look for a charset, first in headers // Expected form: Content-Type: text/html; charset=ISO-8859-4 - if (preg_match('/charset=([a-z0-9-_.\/]+)/i', $header, $matches)) { + if (preg_match('/charset=([a-z0-9-_.\/]+)/i', $curlResult->getContentType(), $matches)) { $charset = trim(trim(trim(array_pop($matches)), ';,')); } @@ -297,7 +299,7 @@ class ParseUrl // See https://github.com/friendica/friendica/issues/5470#issuecomment-418351211 $charset = str_ireplace('latin-1', 'latin1', $charset); - Logger::log('detected charset ' . $charset, Logger::DEBUG); + Logger::info('detected charset', ['charset' => $charset]); $body = iconv($charset, 'UTF-8//TRANSLIT', $body); } @@ -380,18 +382,24 @@ class ParseUrl case 'twitter:image:src': $siteinfo['image'] = $meta_tag['content']; break; - case 'twitter:card': - // Detect photo pages - if ($meta_tag['content'] == 'summary_large_image') { - $siteinfo['type'] = 'photo'; - } - break; case 'twitter:description': $siteinfo['text'] = trim($meta_tag['content']); break; case 'twitter:title': $siteinfo['title'] = trim($meta_tag['content']); break; + case 'twitter:player': + $siteinfo['player']['embed'] = trim($meta_tag['content']); + break; + case 'twitter:player:stream': + $siteinfo['player']['stream'] = trim($meta_tag['content']); + break; + case 'twitter:player:width': + $siteinfo['player']['width'] = intval($meta_tag['content']); + break; + case 'twitter:player:height': + $siteinfo['player']['height'] = intval($meta_tag['content']); + break; case 'dc.title': $siteinfo['title'] = trim($meta_tag['content']); break; @@ -450,6 +458,12 @@ class ParseUrl case 'og:site_name': $siteinfo['publisher_name'] = trim($meta_tag['content']); break; + case 'og:locale': + $siteinfo['language'] = trim($meta_tag['content']); + break; + case 'og:type': + $siteinfo['pagetype'] = trim($meta_tag['content']); + break; case 'twitter:description': $siteinfo['text'] = trim($meta_tag['content']); break; @@ -466,32 +480,37 @@ class ParseUrl $list = $xpath->query("//script[@type='application/ld+json']"); foreach ($list as $node) { if (!empty($node->nodeValue)) { - $nodevalue = html_entity_decode($node->nodeValue, ENT_COMPAT, 'UTF-8'); - if ($jsonld = json_decode($nodevalue, true)) { + if ($jsonld = json_decode($node->nodeValue, true)) { $siteinfo = self::parseParts($siteinfo, $jsonld); } } } - // Prevent to have a photo type without an image - if ((empty($siteinfo['image']) || !empty($siteinfo['text'])) && ($siteinfo['type'] == 'photo')) { - $siteinfo['type'] = 'link'; - } - - if (!empty($siteinfo['image']) && empty($siteinfo['images'])) { - $src = self::completeUrl($siteinfo['image'], $url); + if (!empty($siteinfo['player']['stream'])) { + // Only add player data to media arrays if there is no duplicate + $content_urls = array_merge(array_column($siteinfo['audio'] ?? [], 'content'), array_column($siteinfo['video'] ?? [], 'content')); + if (!in_array($siteinfo['player']['stream'], $content_urls)) { + $contenttype = self::getContentType($siteinfo['player']['stream']); + if (!empty($contenttype[0]) && in_array($contenttype[0], ['audio', 'video'])) { + $media = ['content' => $siteinfo['player']['stream']]; - unset($siteinfo['image']); - - $photodata = Images::getInfoFromURLCached($src); + if (!empty($siteinfo['player']['embed'])) { + $media['embed'] = $siteinfo['player']['embed']; + } - if (($photodata) && ($photodata[0] > 10) && ($photodata[1] > 10)) { - $siteinfo['images'][] = ['src' => $src, - 'width' => $photodata[0], - 'height' => $photodata[1]]; + $siteinfo[$contenttype[0]][] = $media; + } } } + if (!empty($siteinfo['image'])) { + $siteinfo['images'] = $siteinfo['images'] ?? []; + array_unshift($siteinfo['images'], ['url' => $siteinfo['image']]); + unset($siteinfo['image']); + } + + $siteinfo = self::checkMedia($url, $siteinfo); + if (!empty($siteinfo['text']) && mb_strlen($siteinfo['text']) > self::MAX_DESC_COUNT) { $siteinfo['text'] = mb_substr($siteinfo['text'], 0, self::MAX_DESC_COUNT) . '…'; $pos = mb_strrpos($siteinfo['text'], '.'); @@ -504,6 +523,104 @@ class ParseUrl Hook::callAll('getsiteinfo', $siteinfo); + ksort($siteinfo); + + return $siteinfo; + } + + /** + * Check the attached media elements. + * Fix existing data and add missing data. + * + * @param string $page_url + * @param array $siteinfo + * @return array + */ + private static function checkMedia(string $page_url, array $siteinfo) : array + { + if (!empty($siteinfo['images'])) { + array_walk($siteinfo['images'], function (&$image) use ($page_url) { + // According to the specifications someone could place a picture url into the content field as well. + // But this doesn't seem to happen in the wild, so we don't cover it here. + if (!empty($image['url'])) { + $image['url'] = self::completeUrl($image['url'], $page_url); + $photodata = Images::getInfoFromURLCached($image['url']); + if (!empty($photodata) && ($photodata[0] > 50) && ($photodata[1] > 50)) { + $image['src'] = $image['url']; + $image['width'] = $photodata[0]; + $image['height'] = $photodata[1]; + $image['contenttype'] = $photodata['mime']; + unset($image['url']); + ksort($image); + } else { + $image = []; + } + } else { + $image = []; + } + }); + + $siteinfo['images'] = array_values(array_filter($siteinfo['images'])); + } + + foreach (['audio', 'video'] as $element) { + if (!empty($siteinfo[$element])) { + array_walk($siteinfo[$element], function (&$media) use ($page_url, &$siteinfo) { + $url = ''; + $embed = ''; + $content = ''; + $contenttype = ''; + foreach (['embed', 'content', 'url'] as $field) { + if (!empty($media[$field])) { + $media[$field] = self::completeUrl($media[$field], $page_url); + $type = self::getContentType($media[$field]); + if (($type[0] ?? '') == 'text') { + if ($field == 'embed') { + $embed = $media[$field]; + } else { + $url = $media[$field]; + } + } elseif (!empty($type[0])) { + $content = $media[$field]; + $contenttype = implode('/', $type); + } + } + unset($media[$field]); + } + + foreach (['image', 'preview'] as $field) { + if (!empty($media[$field])) { + $media[$field] = self::completeUrl($media[$field], $page_url); + } + } + + if (!empty($url)) { + $media['url'] = $url; + } + if (!empty($embed)) { + $media['embed'] = $embed; + if (empty($siteinfo['player']['embed'])) { + $siteinfo['player']['embed'] = $embed; + } + } + if (!empty($content)) { + $media['src'] = $content; + } + if (!empty($contenttype)) { + $media['contenttype'] = $contenttype; + } + if (empty($url) && empty($content) && empty($embed)) { + $media = []; + } + ksort($media); + }); + + $siteinfo[$element] = array_values(array_filter($siteinfo[$element])); + } + if (empty($siteinfo[$element])) { + unset($siteinfo[$element]); + } + } return $siteinfo; } @@ -600,7 +717,9 @@ class ParseUrl { if (!empty($jsonld['@graph']) && is_array($jsonld['@graph'])) { foreach ($jsonld['@graph'] as $part) { - $siteinfo = self::parseParts($siteinfo, $part); + if (!empty($part) && is_array($part)) { + $siteinfo = self::parseParts($siteinfo, $part); + } } } elseif (!empty($jsonld['@type'])) { $siteinfo = self::parseJsonLd($siteinfo, $jsonld); @@ -614,11 +733,19 @@ class ParseUrl } if ($numeric_keys) { foreach ($jsonld as $part) { - $siteinfo = self::parseParts($siteinfo, $part); - } + if (!empty($part) && is_array($part)) { + $siteinfo = self::parseParts($siteinfo, $part); + } + } } } + array_walk_recursive($siteinfo, function (&$element) { + if (is_string($element)) { + $element = trim(strip_tags(html_entity_decode($element, ENT_COMPAT, 'UTF-8'))); + } + }); + return $siteinfo; } @@ -673,7 +800,7 @@ class ParseUrl case 'QAPage': case 'RealEstateListing': case 'SearchResultsPage': - case 'MediaGallery': + case 'MediaGallery': case 'ImageGallery': case 'VideoGallery': case 'RadioEpisode': @@ -702,7 +829,7 @@ class ParseUrl case 'PerformingGroup': case 'DanceGroup'; case 'MusicGroup': - case 'TheaterGroup': + case 'TheaterGroup': return self::parseJsonLdWebPerson($siteinfo, $jsonld); case 'AudioObject': case 'Audio': @@ -734,11 +861,6 @@ class ParseUrl $jsonldinfo['publisher_name'] = trim($content); } - $content = JsonLD::fetchElement($jsonld, 'publisher', 'sameAs'); - if (!empty($content) && is_string($content)) { - $jsonldinfo['publisher_url'] = trim($content); - } - $content = JsonLD::fetchElement($jsonld, 'publisher', 'url'); if (!empty($content) && is_string($content)) { $jsonldinfo['publisher_url'] = trim($content); @@ -751,11 +873,6 @@ class ParseUrl $jsonldinfo['publisher_name'] = trim($content); } - $content = JsonLD::fetchElement($brand, 'sameAs'); - if (!empty($content) && is_string($content)) { - $jsonldinfo['publisher_url'] = trim($content); - } - $content = JsonLD::fetchElement($brand, 'url'); if (!empty($content) && is_string($content)) { $jsonldinfo['publisher_url'] = trim($content); @@ -856,13 +973,23 @@ class ParseUrl $siteinfo['keywords'][] = trim($keyword); } } - } else { + } elseif (!empty($jsonld['keywords'])) { $content = JsonLD::fetchElementArray($jsonld, 'keywords'); if (!empty($content) && is_array($content)) { $jsonldinfo['keywords'] = $content; } } + $content = JsonLD::fetchElement($jsonld, 'datePublished'); + if (!empty($content) && is_string($content)) { + $jsonldinfo['published'] = DateTimeFormat::utc($content); + } + + $content = JsonLD::fetchElement($jsonld, 'dateModified'); + if (!empty($content) && is_string($content)) { + $jsonldinfo['modified'] = DateTimeFormat::utc($content); + } + $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld); Logger::info('Fetched article information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]); @@ -969,11 +1096,6 @@ class ParseUrl $jsonldinfo['publisher_description'] = trim($content); } - $content = JsonLD::fetchElement($jsonld, 'sameAs'); - if (!empty($content) && is_string($content)) { - $jsonldinfo['publisher_url'] = trim($content); - } - $content = JsonLD::fetchElement($jsonld, 'url'); if (!empty($content)) { $jsonldinfo['publisher_url'] = trim($content); @@ -1031,7 +1153,11 @@ class ParseUrl } $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject'); - if (!empty($content)) { + if (!empty($content) && !is_string($content)) { + Logger::notice('Unexpected return value for the author image', ['content' => $content]); + } + + if (!empty($content) && is_string($content)) { $jsonldinfo['author_img'] = trim($content); } @@ -1058,7 +1184,12 @@ class ParseUrl $content = JsonLD::fetchElement($jsonld, 'url'); if (!empty($content)) { - $media['src'] = trim($content); + $media['url'] = trim($content); + } + + $content = JsonLD::fetchElement($jsonld, 'mainEntityOfPage'); + if (!empty($content)) { + $media['main'] = Strings::compareLink($content, $siteinfo['url']); } $content = JsonLD::fetchElement($jsonld, 'description'); @@ -1091,16 +1222,20 @@ class ParseUrl $media['width'] = trim($content); } - $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl'); - if (!empty($content)) { - $media['preview'] = trim($content); - } - $content = JsonLD::fetchElement($jsonld, 'image'); if (!empty($content)) { $media['image'] = trim($content); } + $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl'); + if (!empty($content) && (($media['image'] ?? '') != trim($content))) { + if (!empty($media['image'])) { + $media['preview'] = trim($content); + } else { + $media['image'] = trim($content); + } + } + Logger::info('Fetched Media information', ['url' => $siteinfo['url'], 'fetched' => $media]); $siteinfo[$name][] = $media; return $siteinfo;