Change URL detection regex to exclude trailing content (#6572)

Author: JamesNK

playground/Stress/Stress.ApiService/ConsoleStresser.cs

@@ -78,6 +78,11 @@ public static void Stress()
         Console.WriteLine("https://www.example.com/path/with/percent%25encoded");
         Console.WriteLine("https://www.example.com/path/with/dollar$sign");
         Console.WriteLine("https://www.example.com/path/with/exclamation!mark");
+        Console.WriteLine("https://www.example.com/;path/");
+        Console.WriteLine("https://www.example.com/path/?query;string");
+        Console.WriteLine("https://;www.example.com/");
+        Console.WriteLine("https://www;.example.com/");
+        Console.WriteLine("https://www.exa;mple.com/");
 
         Console.Write("\x1b[0m"); // reset color
 

src/Aspire.Dashboard/ConsoleLogs/UrlParser.cs

@@ -3,6 +3,7 @@
 
 using System.Diagnostics.CodeAnalysis;
 using System.Globalization;
+using System.Net;
 using System.Text;
 using System.Text.RegularExpressions;
 
@@ -34,7 +35,7 @@ public static bool TryParse(string? text, Func<string, string>? nonMatchFragment
                 nextCharIndex = urlMatch.Index + urlMatch.Length;
                 var url = text[urlStart..nextCharIndex];
 
-                builder.Append(CultureInfo.InvariantCulture, $"<a target=\"_blank\" href=\"{url}\">{url}</a>");
+                builder.Append(CultureInfo.InvariantCulture, $"<a target=\"_blank\" href=\"{url}\">{WebUtility.HtmlEncode(url)}</a>");
                 urlMatch = urlMatch.NextMatch();
             }
 
@@ -65,17 +66,9 @@ static void AppendNonMatchFragment(StringBuilder stringBuilder, Func<string, str
     }
 
     // Regular expression that detects http/https URLs in a log entry
-    // Based on the RegEx used in Windows Terminal for the same purpose. Some modifications:
-    // - Can start at a non word boundary. This behavior is similar to how GitHub matches URLs in pretty printed code.
-    // - Limited to only http/https URLs.
-    // - Ignore case. That means it matches URLs starting with http and HTTP.
-    //
-    // Explanation:
-    // https?://                      - http:// or https://
-    // [-A-Za-z0-9+&@#/%?=~_|$!:,.;]* - Any character in the list, matched zero or more times.
-    // [A-Za-z0-9+&@#/%=~_|$]         - Any character in the list, matched exactly once
+    // Based on the RegEx used by GitHub to detect links in content.
     [GeneratedRegex(
-        "https?://[-A-Za-z0-9+&@#/%?=~_|$!:,.;]*[A-Za-z0-9+&@#/%=~_|$]",
-        RegexOptions.IgnoreCase | RegexOptions.CultureInvariant)]
+        @"((?<!\+)https?:\/\/(?:www\.)?(?:[-\p{L}.]+?[.@][a-zA-Z\d]{2,}|localhost)(?:[-\w\p{L}.:%+~#*$!?&/=@]*(?:,(?!\s))*?)*)",
+        RegexOptions.IgnoreCase | RegexOptions.CultureInvariant | RegexOptions.ExplicitCapture)]
     public static partial Regex GenerateUrlRegEx();
 }

tests/Aspire.Dashboard.Tests/ConsoleLogsTests/UrlParserTests.cs

@@ -40,7 +40,7 @@ public void TryParse_ReturnsCorrectResult(string input, bool expectedResult, str
     [InlineData("http://bing.com/", "<a target=\"_blank\" href=\"http://bing.com/\">http://bing.com/</a>")]
     [InlineData("http://bing.com/dir", "<a target=\"_blank\" href=\"http://bing.com/dir\">http://bing.com/dir</a>")]
     [InlineData("http://bing.com/index.aspx", "<a target=\"_blank\" href=\"http://bing.com/index.aspx\">http://bing.com/index.aspx</a>")]
-    [InlineData("http://bing", "<a target=\"_blank\" href=\"http://bing\">http://bing</a>")]
+    [InlineData("http://localhost", "<a target=\"_blank\" href=\"http://localhost\">http://localhost</a>")]
     public void TryParse_SupportedUrlFormats(string input, string? expectedOutput)
     {
         var result = UrlParser.TryParse(input, WebUtility.HtmlEncode, out var modifiedText);
@@ -71,6 +71,15 @@ public void TryParse_ExcludeInvalidTrailingChars(string input, string? expectedO
         Assert.Equal(expectedOutput, modifiedText);
     }
 
+    [Fact]
+    public void TryParse_QueryString()
+    {
+        var result = UrlParser.TryParse("https://www.example.com?query=string&param=value", WebUtility.HtmlEncode, out var modifiedText);
+        Assert.True(result);
+
+        Assert.Equal("<a target=\"_blank\" href=\"https://www.example.com?query=string&param=value\">https://www.example.com?query=string&amp;param=value</a>", modifiedText);
+    }
+
     [Theory]
     [InlineData("http://www.localhost:8080")]
     [InlineData("HTTP://WWW.LOCALHOST:8080")]
@@ -83,4 +92,17 @@ public void GenerateUrlRegEx_MatchUrlAfterContent(string content)
         var match = regex.Match(content);
         Assert.Equal("http://www.localhost:8080", match.Value.ToLowerInvariant());
     }
+
+    [Theory]
+    [InlineData("http://www.localhost:8080!", "http://www.localhost:8080!")]
+    [InlineData("http://www.localhost:8080/path!", "http://www.localhost:8080/path!")]
+    [InlineData("http://www.localhost:8080/path;", "http://www.localhost:8080/path")]
+    [InlineData("http://www.localhost:8080;", "http://www.localhost:8080")]
+    [InlineData("http://www.local;host:8080;", "http://www.local")]
+    public void GenerateUrlRegEx_MatchUrlBeforeContent(string content, string expected)
+    {
+        var regex = UrlParser.GenerateUrlRegEx();
+        var match = regex.Match(content);
+        Assert.Equal(expected, match.Value.ToLowerInvariant());
+    }
 }