rss_converter_instagram.com.xsl: support scraping Instagram locations pages

[tweeper.git] / src / Tweeper.php

diff --git a/src/Tweeper.php b/src/Tweeper.php
index 944b37c..8ac2fe3 100644 (file)
--- 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 <ao2@ao2.it>
+ * Copyright (C) 2013-2016  Antonio Ospite <ao2@ao2.it>
  *
  * 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
@@ -315,21 +315,30 @@ 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) {