core: fix invalid characters in xml/rss. resolves #9118 (#9636)

src/Jackett.Test/Models/ResultPageTests.cs

@@ -0,0 +1,83 @@
+using System;
+using System.Collections.Generic;
+using System.Text.RegularExpressions;
+using System.Threading.Tasks;
+using Jackett.Common.Indexers;
+using Jackett.Common.Models;
+using Newtonsoft.Json.Linq;
+using NUnit.Framework;
+using Assert = NUnit.Framework.Assert;
+
+namespace Jackett.Test.Models
+{
+    class TestIndexer : BaseIndexer
+    {
+        public TestIndexer()
+            : base(id: "test_id",
+                   name: "test_name",
+                   description: "test_description",
+                   link: "https://test.link/",
+                   configService: null,
+                   logger: null,
+                   configData: null,
+                   p: null)
+        {
+        }
+
+        public override TorznabCapabilities TorznabCaps { get; protected set; }
+        public override Task<IndexerConfigurationStatus> ApplyConfiguration(JToken configJson) => throw new NotImplementedException();
+        protected override Task<IEnumerable<ReleaseInfo>> PerformQuery(TorznabQuery query) => throw new NotImplementedException();
+    }
+
+    [TestFixture]
+    public class ResultPageTests
+    {
+        [Test]
+        public void TestXmlWithInvalidCharacters()
+        {
+            // 0x1A can't be represented in XML => https://stackoverflow.com/a/8506173
+            // some ascii and unicode characters
+            var text = "Title Ñ 理" + Convert.ToChar("\u001a") + Convert.ToChar("\u2813");
+            var validText = "Title Ñ 理" + Convert.ToChar("\u2813");
+
+            // link with characters that requires URL encode
+            var link = new Uri("https://example.com/" + text);
+            var validLink = "https://example.com/Title%20%C3%91%20%E7%90%86%1A%E2%A0%93";
+
+            var resultPage = new ResultPage(
+                new ChannelInfo // characters in channel info are safe because are provided by us
+                {
+                    Link = link,
+                    ImageUrl = link,
+                    ImageLink = link
+                })
+                {
+                    Releases = new List<ReleaseInfo>
+                    {
+                        new ReleaseInfo // these fields are from websites and they can be problematic
+                        {
+                            Title = text,
+                            Guid = link,
+                            Link = link,
+                            Comments = link,
+                            PublishDate = new DateTime(2020, 09, 22),
+                            Description = text,
+                            Author = text,
+                            BookTitle = text,
+                            BannerUrl = link,
+                            InfoHash = text,
+                            MagnetUri = link,
+                            Origin = new TestIndexer()
+                        }
+                    }
+                };
+            var xml = resultPage.ToXml(link);
+
+            Assert.AreEqual(5, Regex.Matches(xml, validText).Count);
+            Assert.AreEqual(9, Regex.Matches(xml, validLink).Count);
+
+            // this should be in another test but it's here to avoid creating the whole object again
+            Assert.True(xml.Contains("Tue, 22 Sep 2020 00:00:00 "));
+        }
+    }
+}