<?php
/**
- * @copyright Copyright (C) 2010-2021, the Friendica project
+ * @copyright Copyright (C) 2010-2023, the Friendica project
*
* @license GNU AGPL version 3 or any later version
*
use DOMDocument;
use DOMXPath;
use Friendica\Content\OEmbed;
+use Friendica\Content\Text\HTML;
+use Friendica\Protocol\HTTP\MediaType;
use Friendica\Core\Hook;
use Friendica\Core\Logger;
use Friendica\Database\Database;
use Friendica\Database\DBA;
use Friendica\DI;
+use Friendica\Network\HTTPClient\Client\HttpClientAccept;
use Friendica\Network\HTTPException;
+use Friendica\Network\HTTPClient\Client\HttpClientOptions;
/**
* Get information about a given URL
/**
* Fetch the content type of the given url
- * @param string $url URL of the page
+ * @param string $url URL of the page
+ * @param string $accept content-type to accept
+ * @param int $timeout
* @return array content type
*/
- public static function getContentType(string $url)
+ public static function getContentType(string $url, string $accept = HttpClientAccept::DEFAULT, int $timeout = 0): array
{
- $curlResult = DI::httpRequest()->head($url);
+ if (!empty($timeout)) {
+ $options = [HttpClientOptions::TIMEOUT => $timeout];
+ } else {
+ $options = [];
+ }
+
+ try {
+ $curlResult = DI::httpClient()->head($url, array_merge([HttpClientOptions::ACCEPT_CONTENT => $accept], $options));
+ } catch (\Exception $e) {
+ DI::logger()->debug('Got exception', ['url' => $url, 'message' => $e->getMessage()]);
+ return [];
+ }
+
+ // Workaround for systems that can't handle a HEAD request. Don't retry on timeouts.
+ if (!$curlResult->isSuccess() && ($curlResult->getReturnCode() >= 400) && !in_array($curlResult->getReturnCode(), [408, 504])) {
+ $curlResult = DI::httpClient()->get($url, $accept, array_merge([HttpClientOptions::CONTENT_LENGTH => 1000000], $options));
+ }
+
if (!$curlResult->isSuccess()) {
+ Logger::debug('Got HTTP Error', ['http error' => $curlResult->getReturnCode(), 'url' => $url]);
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)));
}
/**
- * Search for chached embeddable data of an url otherwise fetch it
+ * Search for cached embeddable data of an url otherwise fetch it
*
* @param string $url The url of the page which should be scraped
* @param bool $do_oembed The false option is used by the function fetch_oembed()
* @see ParseUrl::getSiteinfo() for more information about scraping
* embeddable content
*/
- public static function getSiteinfoCached($url, $do_oembed = true): array
+ public static function getSiteinfoCached(string $url, bool $do_oembed = true): array
{
if (empty($url)) {
return [
* </body>
* @endverbatim
*/
- public static function getSiteinfo($url, $do_oembed = true, $count = 1)
+ public static function getSiteinfo(string $url, bool $do_oembed = true, int $count = 1): array
{
if (empty($url)) {
return [
];
if ($count > 10) {
- Logger::notice('Endless loop detected', ['url' => $url]);
+ Logger::warning('Endless loop detected', ['url' => $url]);
return $siteinfo;
}
return $siteinfo;
}
- $curlResult = DI::httpRequest()->get($url, false, ['content_length' => 1000000]);
+ $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML, [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");
}
$charset = '';
- // Look for a charset, first in headers
- // Expected form: Content-Type: text/html; charset=ISO-8859-4
- foreach ($curlResult->getContentType() as $type) {
- if (preg_match('/charset=([a-z0-9-_.\/]+)/i', $type, $matches)) {
- $charset = trim(trim(trim(array_pop($matches)), ';,'));
+ try {
+ // Look for a charset, first in headers
+ $mediaType = MediaType::fromContentType($curlResult->getContentType());
+ if (isset($mediaType->parameters['charset'])) {
+ $charset = $mediaType->parameters['charset'];
}
- }
-
- // Then in body that gets precedence
- // Expected forms:
- // - <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
- // - <meta charset="utf-8">
- // - <meta charset=utf-8>
- // - <meta charSet="utf-8">
- // We escape <style> and <script> tags since they can contain irrelevant charset information
- // (see https://github.com/friendica/friendica/issues/9251#issuecomment-698636806)
- Strings::performWithEscapedBlocks($body, '#<(?:style|script).*?</(?:style|script)>#ism', function ($body) use (&$charset) {
- if (preg_match('/charset=["\']?([a-z0-9-_.\/]+)/i', $body, $matches)) {
- $charset = trim(trim(trim(array_pop($matches)), ';,'));
- }
- });
+ } catch(\InvalidArgumentException $e) {}
$siteinfo['charset'] = $charset;
$body = mb_convert_encoding($body, 'HTML-ENTITIES', 'UTF-8');
+ if (empty($body)) {
+ return $siteinfo;
+ }
+
$doc = new DOMDocument();
@$doc->loadHTML($body);
+ $siteinfo['charset'] = HTML::extractCharset($doc) ?? $siteinfo['charset'];
+
XML::deleteNode($doc, 'style');
XML::deleteNode($doc, 'option');
XML::deleteNode($doc, 'h1');
$siteinfo['title'] = trim($list->item(0)->nodeValue);
}
- $twitter_card = false;
- $twitter_image = false;
-
$list = $xpath->query('//meta[@name]');
foreach ($list as $node) {
$meta_tag = [];
break;
case 'twitter:image':
$siteinfo['image'] = $meta_tag['content'];
- $twitter_image = true;
break;
case 'twitter:image:src':
$siteinfo['image'] = $meta_tag['content'];
break;
- case 'twitter:card':
- // Detect photo pages
- if ($meta_tag['content'] == 'summary_large_image') {
- $twitter_card = true;
- }
- 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;
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;
break;
case 'twitter:image':
$siteinfo['image'] = $meta_tag['content'];
- $twitter_image = true;
break;
}
}
}
}
-// Currently deactivated, see https://github.com/friendica/friendica/pull/10148#issuecomment-821512658
- // Prevent to have a photo type without an image
-// if ($twitter_card && $twitter_image && !empty($siteinfo['image'])) {
-// $siteinfo['type'] = 'photo';
-// }
+ if (!empty($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' => $siteinfo['player']['stream']];
+
+ if (!empty($siteinfo['player']['embed'])) {
+ $media['embed'] = $siteinfo['player']['embed'];
+ }
+
+ $siteinfo[$contenttype[0]][] = $media;
+ }
+ }
+ }
if (!empty($siteinfo['image'])) {
$siteinfo['images'] = $siteinfo['images'] ?? [];
Hook::callAll('getsiteinfo', $siteinfo);
+ ksort($siteinfo);
+
return $siteinfo;
}
*
* @param string $page_url
* @param array $siteinfo
- * @return void
+ * @return array
*/
- private static function checkMedia(string $page_url, array $siteinfo)
+ private static function checkMedia(string $page_url, array $siteinfo) : array
{
if (!empty($siteinfo['images'])) {
array_walk($siteinfo['images'], function (&$image) use ($page_url) {
- // According to the specifications someone could place a picture url into the content field as well.
- // But this doesn't seem to happen in the wild, so we don't cover it here.
+ /*
+ * According to the specifications someone could place a picture
+ * URL into the content field as well. But this doesn't seem to
+ * happen in the wild, so we don't cover it here.
+ */
if (!empty($image['url'])) {
$image['url'] = self::completeUrl($image['url'], $page_url);
$photodata = Images::getInfoFromURLCached($image['url']);
- if (!empty($photodata) && ($photodata[0] > 50) && ($photodata[1] > 50)) {
+ if (($photodata) && ($photodata[0] > 50) && ($photodata[1] > 50)) {
$image['src'] = $image['url'];
$image['width'] = $photodata[0];
$image['height'] = $photodata[1];
$image['contenttype'] = $photodata['mime'];
+ $image['blurhash'] = $photodata['blurhash'] ?? null;
unset($image['url']);
ksort($image);
} else {
}
if (!empty($embed)) {
$media['embed'] = $embed;
- if (!empty($media['main'])) {
- $siteinfo['embed'] = $embed;
+ if (empty($siteinfo['player']['embed'])) {
+ $siteinfo['player']['embed'] = $embed;
}
}
if (!empty($content)) {
* Convert tags from CSV to an array
*
* @param string $string Tags
+ *
* @return array with formatted Hashtags
*/
- public static function convertTagsToArray($string)
+ public static function convertTagsToArray(string $string): array
{
$arr_tags = str_getcsv($string);
if (count($arr_tags)) {
// add the # sign to every tag
- array_walk($arr_tags, ["self", "arrAddHashes"]);
+ array_walk($arr_tags, [self::class, 'arrAddHashes']);
return $arr_tags;
}
+ return [];
}
/**
*
* @param string $tag The pure tag name
* @param int $k Counter for internal use
+ *
* @return void
*/
- private static function arrAddHashes(&$tag, $k)
+ private static function arrAddHashes(string &$tag, int $k)
{
- $tag = "#" . $tag;
+ $tag = '#' . $tag;
}
/**
*
* @return string The url with a scheme
*/
- private static function completeUrl($url, $scheme)
+ private static function completeUrl(string $url, string $scheme): string
{
$urlarr = parse_url($url);
- // If the url does allready have an scheme
+ // If the url does already have an scheme
// we can stop the process here
- if (isset($urlarr["scheme"])) {
- return($url);
+ if (isset($urlarr['scheme'])) {
+ return $url;
}
$schemearr = parse_url($scheme);
- $complete = $schemearr["scheme"]."://".$schemearr["host"];
+ $complete = $schemearr['scheme'] . '://' . $schemearr['host'];
- if (!empty($schemearr["port"])) {
- $complete .= ":".$schemearr["port"];
+ if (!empty($schemearr['port'])) {
+ $complete .= ':' . $schemearr['port'];
}
- if (!empty($urlarr["path"])) {
- if (strpos($urlarr["path"], "/") !== 0) {
- $complete .= "/";
+ if (!empty($urlarr['path'])) {
+ if (strpos($urlarr['path'], '/') !== 0) {
+ $complete .= '/';
}
- $complete .= $urlarr["path"];
+ $complete .= $urlarr['path'];
}
- if (!empty($urlarr["query"])) {
- $complete .= "?".$urlarr["query"];
+ if (!empty($urlarr['query'])) {
+ $complete .= '?' . $urlarr['query'];
}
- if (!empty($urlarr["fragment"])) {
- $complete .= "#".$urlarr["fragment"];
+ if (!empty($urlarr['fragment'])) {
+ $complete .= '#' . $urlarr['fragment'];
}
- return($complete);
+ return $complete;
}
/**
*
* @param array $siteinfo
* @param array $jsonld
+ *
* @return array siteinfo
*/
- private static function parseParts(array $siteinfo, array $jsonld)
+ private static function parseParts(array $siteinfo, array $jsonld): array
{
if (!empty($jsonld['@graph']) && is_array($jsonld['@graph'])) {
foreach ($jsonld['@graph'] as $part) {
*
* @param array $siteinfo
* @param array $jsonld
+ *
* @return array siteinfo
*/
- private static function parseJsonLd(array $siteinfo, array $jsonld)
+ private static function parseJsonLd(array $siteinfo, array $jsonld): array
{
$type = JsonLD::fetchElement($jsonld, '@type');
if (empty($type)) {
*
* @param array $siteinfo
* @param array $jsonld
+ *
* @return array siteinfo
*/
- private static function parseJsonLdAuthor(array $siteinfo, array $jsonld)
+ private static function parseJsonLdAuthor(array $siteinfo, array $jsonld): array
{
$jsonldinfo = [];
*
* @param array $siteinfo
* @param array $jsonld
+ *
* @return array siteinfo
*/
- private static function parseJsonLdArticle(array $siteinfo, array $jsonld)
+ private static function parseJsonLdArticle(array $siteinfo, array $jsonld): array
{
$jsonldinfo = [];
}
}
+ $content = JsonLD::fetchElement($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]);
*
* @param array $siteinfo
* @param array $jsonld
+ *
* @return array siteinfo
*/
- private static function parseJsonLdWebPage(array $siteinfo, array $jsonld)
+ private static function parseJsonLdWebPage(array $siteinfo, array $jsonld): array
{
$jsonldinfo = [];
}
$content = JsonLD::fetchElement($jsonld, 'description');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['text'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'image');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['image'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['image'] = trim($content);
}
*
* @param array $siteinfo
* @param array $jsonld
+ *
* @return array siteinfo
*/
- private static function parseJsonLdWebSite(array $siteinfo, array $jsonld)
+ private static function parseJsonLdWebSite(array $siteinfo, array $jsonld): array
{
$jsonldinfo = [];
$content = JsonLD::fetchElement($jsonld, 'name');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['publisher_name'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'description');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['publisher_description'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'url');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['publisher_url'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['image'] = trim($content);
}
*
* @param array $siteinfo
* @param array $jsonld
+ *
* @return array siteinfo
*/
- private static function parseJsonLdWebOrganization(array $siteinfo, array $jsonld)
+ private static function parseJsonLdWebOrganization(array $siteinfo, array $jsonld): array
{
$jsonldinfo = [];
$content = JsonLD::fetchElement($jsonld, 'name');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['publisher_name'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'description');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['publisher_description'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'url');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['publisher_url'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'logo', 'url', '@type', 'ImageObject');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['publisher_img'] = trim($content);
+ } elseif (!empty($content) && is_array($content)) {
+ $jsonldinfo['publisher_img'] = trim($content[0]);
}
$content = JsonLD::fetchElement($jsonld, 'brand', 'name', '@type', 'Organization');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['publisher_name'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'brand', 'url', '@type', 'Organization');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['publisher_url'] = trim($content);
}
*
* @param array $siteinfo
* @param array $jsonld
+ *
* @return array siteinfo
*/
- private static function parseJsonLdWebPerson(array $siteinfo, array $jsonld)
+ private static function parseJsonLdWebPerson(array $siteinfo, array $jsonld): array
{
$jsonldinfo = [];
$content = JsonLD::fetchElement($jsonld, 'name');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['author_name'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'description');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['author_description'] = trim($content);
}
}
$content = JsonLD::fetchElement($jsonld, 'url');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$jsonldinfo['author_url'] = trim($content);
}
$content = JsonLD::fetchElement($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);
}
*
* @param array $siteinfo
* @param array $jsonld
+ *
* @return array siteinfo
*/
- private static function parseJsonLdMediaObject(array $siteinfo, array $jsonld, string $name)
+ private static function parseJsonLdMediaObject(array $siteinfo, array $jsonld, string $name): array
{
$media = [];
$content = JsonLD::fetchElement($jsonld, 'caption');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$media['caption'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'url');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$media['url'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'mainEntityOfPage');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$media['main'] = Strings::compareLink($content, $siteinfo['url']);
}
$content = JsonLD::fetchElement($jsonld, 'description');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$media['description'] = trim($content);
}
}
$content = JsonLD::fetchElement($jsonld, 'contentUrl');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$media['content'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'embedUrl');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$media['embed'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'height');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$media['height'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'width');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$media['width'] = trim($content);
}
$content = JsonLD::fetchElement($jsonld, 'image');
- if (!empty($content)) {
+ if (!empty($content) && is_string($content)) {
$media['image'] = trim($content);
}