From a4e0c57e95ca4617907a996256fabe5f4c4d6148 Mon Sep 17 00:00:00 2001
From: James Newton-King <james@newtonking.com>
Date: Thu, 31 Oct 2024 21:35:58 +0800
Subject: [PATCH] Update regex

---
 src/Aspire.Dashboard/ConsoleLogs/UrlParser.cs     | 14 +++-----------
 .../ConsoleLogsTests/UrlParserTests.cs            | 15 ++++++++++++++-
 2 files changed, 17 insertions(+), 12 deletions(-)

diff --git a/src/Aspire.Dashboard/ConsoleLogs/UrlParser.cs b/src/Aspire.Dashboard/ConsoleLogs/UrlParser.cs
index 0b19a200ba..02e1d9fd3e 100644
--- a/src/Aspire.Dashboard/ConsoleLogs/UrlParser.cs
+++ b/src/Aspire.Dashboard/ConsoleLogs/UrlParser.cs
@@ -65,17 +65,9 @@ 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();
 }
diff --git a/tests/Aspire.Dashboard.Tests/ConsoleLogsTests/UrlParserTests.cs b/tests/Aspire.Dashboard.Tests/ConsoleLogsTests/UrlParserTests.cs
index 510a245785..557d589138 100644
--- a/tests/Aspire.Dashboard.Tests/ConsoleLogsTests/UrlParserTests.cs
+++ b/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);
@@ -83,4 +83,17 @@ 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());
+    }
 }