X-Git-Url: https://git.ao2.it/tweeper.git/blobdiff_plain/efd6e189034c7bf376c669349a227ad3bb4f0f67..d566fc7fcf6841909831da3d60b335294befe62d:/src/Tweeper.php diff --git a/src/Tweeper.php b/src/Tweeper.php index 944b37c..50ff148 100644 --- a/src/Tweeper.php +++ b/src/Tweeper.php @@ -6,7 +6,7 @@ namespace Tweeper; * @file * Tweeper - a Twitter to RSS web scraper. * - * Copyright (C) 2013-2015 Antonio Ospite + * Copyright (C) 2013-2018 Antonio Ospite * * This program is free software: you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -36,7 +36,7 @@ date_default_timezone_set('UTC'); */ class Tweeper { - private static $userAgent = "Mozilla/5.0 (Windows NT 6.1; rv:22.0) Gecko/20130405 Firefox/22.0"; + private static $userAgent = "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101 Firefox/52.0"; /** * Constructor sets up {@link $generate_enclosure}. @@ -249,7 +249,13 @@ class Tweeper { return NULL; } - return Tweeper::jsonToXml($matches[1], 'instagram'); + // The "qe" object contains elements which will result in invalid XML + // element names, so remove it. + $data = json_decode($matches[1], $assoc = TRUE); + unset($data["qe"]); + $json = json_encode($data); + + return Tweeper::jsonToXml($json, 'instagram'); } /** @@ -315,21 +321,30 @@ class Tweeper { /** * Convert the site content to RSS. */ - public function tweep($src_url) { + public function tweep($src_url, $host=NULL, $validate_scheme=TRUE) { $url = parse_url($src_url); - if (FALSE === $url || empty($url["host"])) { + if (FALSE === $url) { trigger_error("Invalid URL: $src_url", E_USER_ERROR); return NULL; } - $scheme = $url["scheme"]; - if (!in_array($scheme, array("http", "https"))) { - trigger_error("unsupported scheme: $scheme", E_USER_ERROR); - return NULL; + if (TRUE === $validate_scheme) { + $scheme = $url["scheme"]; + if (!in_array($scheme, array("http", "https"))) { + trigger_error("unsupported scheme: $scheme", E_USER_ERROR); + return NULL; + } } - // Strip the leading www. to be more forgiving on input URLs. - $host = preg_replace('/^www\./', '', $url["host"]); + // if the host is not given derive it from the URL + if (NULL === $host) { + if (empty($url["host"])) { + trigger_error("Invalid host in URL: $src_url", E_USER_ERROR); + return NULL; + } + // Strip the leading www. to be more forgiving on input URLs. + $host = preg_replace('/^www\./', '', $url["host"]); + } $xsltProcessor = $this->loadStylesheet($host); if (NULL === $xsltProcessor) {