test: add test for FeedParser (#3754)

bridges/ArsTechnicaBridge.php

@@ -30,7 +30,7 @@ class ArsTechnicaBridge extends FeedExpander
     public function collectData()
     {
         $url = 'https://feeds.arstechnica.com/arstechnica/' . $this->getInput('section');
-        $this->collectExpandableDatas($url);
+        $this->collectExpandableDatas($url, 10);
     }
 
     protected function parseItem(array $item)

bridges/UrlebirdBridge.php

@@ -6,7 +6,7 @@ class UrlebirdBridge extends BridgeAbstract
     const NAME = 'urlebird.com';
     const URI = 'https://urlebird.com/';
     const DESCRIPTION = 'Bridge for urlebird.com';
-    const CACHE_TIMEOUT = 10;
+    const CACHE_TIMEOUT = 60 * 5;
     const PARAMETERS = [
         [
             'query' => [
@@ -21,50 +21,70 @@ class UrlebirdBridge extends BridgeAbstract
 
     private $title;
 
-    private function fixURI($uri)
-    {
-        $path = parse_url($uri, PHP_URL_PATH);
-        $encoded_path = array_map('urlencode', explode('/', $path));
-        return str_replace($path, implode('/', $encoded_path), $uri);
-    }
-
     public function collectData()
     {
         switch ($this->getInput('query')[0]) {
-            default:
-                returnServerError('Please, enter valid username or hashtag!');
-                break;
             case '@':
                 $url = 'https://urlebird.com/user/' . substr($this->getInput('query'), 1) . '/';
                 break;
             case '#':
                 $url = 'https://urlebird.com/hash/' . substr($this->getInput('query'), 1) . '/';
                 break;
+            default:
+                returnServerError('Please, enter valid username or hashtag!');
+                break;
         }
 
         $html = getSimpleHTMLDOM($url);
+        $limit = 10;
+
         $this->title = $html->find('title', 0)->innertext;
         $articles = $html->find('div.thumb');
+        $articles = array_slice($articles, 0, $limit);
         foreach ($articles as $article) {
             $item = [];
-            $item['uri'] = $this->fixURI($article->find('a', 2)->href);
+            $itemUrl = $article->find('a', 2)->href;
-            $article_content = getSimpleHTMLDOM($item['uri']);
+            $item['uri'] = $this->encodePathSegments($itemUrl);
-            $item['author'] = $article->find('img', 0)->alt . ' (' .
+
-                $article_content->find('a.user-video', 1)->innertext . ')';
+            $dom = getSimpleHTMLDOM($item['uri']);
-            $item['title'] = $article_content->find('title', 0)->innertext;
+            $videoDiv = $dom->find('div.video', 0);
-            $item['enclosures'][] = $article_content->find('video', 0)->poster;
+
-            $video = $article_content->find('video', 0);
+            // timestamp
+            $timestampH6 = $videoDiv->find('h6', 0);
+            $datetimeString = str_replace('Posted ', '', $timestampH6->plaintext);
+            $item['timestamp'] = $datetimeString;
+
+            $innertext = $dom->find('a.user-video', 1)->innertext;
+            $alt = $article->find('img', 0)->alt;
+            $item['author'] = $alt . ' (' . $innertext . ')';
+
+            $item['title'] = $dom->find('title', 0)->innertext;
+            $item['enclosures'][] = $dom->find('video', 0)->poster;
+
+            $video = $dom->find('video', 0);
             $video->autoplay = null;
+
             $item['content'] = $video->outertext . '<br>' .
-                $article_content->find('div.music', 0) . '<br>' .
+                $dom->find('div.music', 0) . '<br>' .
-                $article_content->find('div.info2', 0)->innertext .
+                $dom->find('div.info2', 0)->innertext .
-                '<br><br><a href="' . $article_content->find('video', 0)->src .
+                '<br><br><a href="' . $dom->find('video', 0)->src .
                 '">Direct video link</a><br><br><a href="' . $item['uri'] .
                 '">Post link</a><br><br>';
+
             $this->items[] = $item;
         }
     }
 
+    private function encodePathSegments($url)
+    {
+        $path = parse_url($url, PHP_URL_PATH);
+        $pathSegments = explode('/', $path);
+        $encodedPathSegments = array_map('urlencode', $pathSegments);
+        $encodedPath = implode('/', $encodedPathSegments);
+        $result = str_replace($path, $encodedPath, $url);
+        return $result;
+    }
+
     public function getName()
     {
         return $this->title ?: parent::getName();

lib/FeedParser.php

@@ -142,6 +142,7 @@ final class FeedParser
         }
 
         if (isset($feedItem->guid)) {
+            // Pluck out a url from guid
             foreach ($feedItem->guid->attributes() as $attribute => $value) {
                 if (
                     $attribute === 'isPermaLink'
@@ -207,9 +208,9 @@ final class FeedParser
             'content'       => null,
             'timestamp'     => null,
             'author'        => null,
-            'uid'           => null,
+            //'uid'           => null,
-            'categories'    => [],
+            //'categories'    => [],
-            'enclosures'    => [],
+            //'enclosures'    => [],
         ];
         if (isset($feedItem->link)) {
             // todo: trim uri

tests/FeedParserTest.php

@@ -0,0 +1,128 @@
+<?php
+
+declare(strict_types=1);
+
+namespace RssBridge\Tests;
+
+use PHPUnit\Framework\TestCase;
+
+class FeedParserTest extends TestCase
+{
+    public function testRss1()
+    {
+        $xml = <<<XML
+        <?xml version="1.0" encoding="utf-8"?> 
+        <rdf:RDF 
+          xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" 
+          xmlns:cc="http://creativecommons.org/ns#"
+          xmlns="http://purl.org/rss/1.0/"
+        > 
+        <channel rdf:about="http://meerkat.oreillynet.com/?_fl=rss1.0">
+            <title>hello feed</title>
+            <link>http://meerkat.oreillynet.com</link>
+            <description>Meerkat: An Open Wire Service</description>
+            
+            <items>
+                <rdf:Seq>
+                    <rdf:li resource="http://c.moreover.com/click/here.pl?r123" />
+                </rdf:Seq>
+            </items>
+        </channel>
+
+        <item rdf:about="http://c.moreover.com/click/here.pl?r123">
+            <title>XML: A Disruptive Technology</title> 
+            <link>http://c.moreover.com/click/here.pl?r123</link>
+            <description>desc</description>
+        </item> 
+        </rdf:RDF>
+        XML;
+
+        $sut = new \FeedParser();
+        $feed = $sut->parseFeed($xml);
+
+        $this->assertSame('hello feed', $feed['title']);
+        $this->assertSame('http://meerkat.oreillynet.com', $feed['uri']);
+        $this->assertSame(null, $feed['icon']);
+
+        $item = $feed['items'][0];
+        $this->assertSame('XML: A Disruptive Technology', $item['title']);
+        $this->assertSame('http://c.moreover.com/click/here.pl?r123', $item['uri']);
+        $this->assertSame('desc', $item['content']);
+    }
+
+    public function testRss2()
+    {
+        $xml = <<<XML
+        <?xml version="1.0"?>
+        <rss version="2.0">
+            <channel>
+                <title>hello feed</title>
+                <link>https://example.com/</link>
+                <image>
+                    <url>https://example.com/2.ico</url>
+                </image>
+
+                <item>
+                    <title>hello world</title>
+                    <link>https://example.com/1</link>
+                    <description>desc2</description>
+                    <pubDate>Tue, 26 Apr 2022 00:00:00 +0200</pubDate>
+                    <author>root</author>
+                    <enclosure url="https://example.com/1.png"></enclosure>
+                </item>
+            </channel>
+        </rss>
+        XML;
+
+        $sut = new \FeedParser();
+        $feed = $sut->parseFeed($xml);
+
+        $this->assertSame('hello feed', $feed['title']);
+        $this->assertSame('https://example.com/', $feed['uri']);
+        $this->assertSame('https://example.com/2.ico', $feed['icon']);
+
+        $item = $feed['items'][0];
+        $this->assertSame('hello world', $item['title']);
+        $this->assertSame('https://example.com/1', $item['uri']);
+        $this->assertSame(1650924000, $item['timestamp']);
+        $this->assertSame('root', $item['author']);
+        $this->assertSame('desc2', $item['content']);
+        $this->assertSame(['https://example.com/1.png'], $item['enclosures']);
+    }
+
+    public function testAtom()
+    {
+        $xml = <<<XML
+        <?xml version="1.0" encoding="UTF-8"?>
+        <feed xmlns="http://www.w3.org/2005/Atom" xmlns:media="http://search.yahoo.com/mrss/">
+            <title>hello feed</title>
+            <link href="https://example.com/1"></link>
+            <icon>https://example.com/2.ico</icon>
+
+            <entry>
+                <title>hello world</title>
+                <link href="https://example.com/1"></link>
+                <author>
+                    <name>root</name>
+                </author>
+                <content type="html">html</content>
+                <updated>2015-11-05T14:38:49+01:00</updated>
+            </entry>
+        </feed>
+        XML;
+
+        $sut = new \FeedParser();
+        $feed = $sut->parseFeed($xml);
+
+        $this->assertSame('hello feed', $feed['title']);
+        $this->assertSame('https://example.com/1', $feed['uri']);
+        $this->assertSame('https://example.com/2.ico', $feed['icon']);
+
+        $item = $feed['items'][0];
+        $this->assertSame('hello world', $item['title']);
+        $this->assertSame('https://example.com/1', $item['uri']);
+        $this->assertSame(1446730729, $item['timestamp']);
+        $this->assertSame('root', $item['author']);
+        $this->assertSame('html', $item['content']);
+    }
+}