Add retry logic for external link timeout exceptions (#80)

External links that timeout due to slow response times now get retried
using the existing retry configuration (--max-external-retries and
--retry-delay-seconds). This prevents false positives for legitimate
sites that are simply responding slowly.

- Catch TaskCanceledException and HttpRequestException with timeout indicators
- Schedule retries with jitter for timeout failures
- Maintain existing behavior for SSL errors and DNS failures (no retry)

Author: Aaronontheweb

src/LinkValidator/Actors/CrawlerActor.cs

@@ -186,9 +186,10 @@ private async Task<ICrawlResult> CrawlExternalPage(AbsoluteUri url, int retryCou
         var scheduler = Context.System.Scheduler;
         var self = Self;
 
+        using var cts = new CancellationTokenSource(_crawlConfiguration.RequestTimeout);
+        
         try
         {
-            using var cts = new CancellationTokenSource(_crawlConfiguration.RequestTimeout);
             var response = await _httpClient.GetAsync(url.Value, cts.Token);
 
             if (response.StatusCode == HttpStatusCode.TooManyRequests)
@@ -212,6 +213,42 @@ private async Task<ICrawlResult> CrawlExternalPage(AbsoluteUri url, int retryCou
 
             return new ExternalLinkCrawled(url, response.StatusCode);
         }
+        catch (TaskCanceledException tcEx) when (tcEx.InnerException is TimeoutException || cts.Token.IsCancellationRequested)
+        {
+            // This is a timeout - retry if we haven't exceeded max retries
+            if (retryCount < _crawlConfiguration.MaxExternalRetries)
+            {
+                var jitteredDelay = AddJitter(_crawlConfiguration.DefaultExternalRetryDelay);
+                _log.Warning("Request timeout for {0} (retry {1}), scheduling retry in {2}", 
+                    url, retryCount, jitteredDelay);
+                
+                scheduler.ScheduleTellOnce(jitteredDelay, self, new RetryExternalRequest(url, retryCount + 1), self);
+                
+                return new ExternalLinkRetryScheduled(url, HttpStatusCode.RequestTimeout);
+            }
+            
+            _log.Warning("Max retries ({0}) exceeded for external URL {1} after timeout", 
+                _crawlConfiguration.MaxExternalRetries, url);
+            return new ExternalLinkCrawled(url, HttpStatusCode.RequestTimeout);
+        }
+        catch (HttpRequestException httpEx) when (httpEx.InnerException is TaskCanceledException || httpEx.Message.Contains("timeout", StringComparison.OrdinalIgnoreCase))
+        {
+            // This is also a timeout - retry if we haven't exceeded max retries
+            if (retryCount < _crawlConfiguration.MaxExternalRetries)
+            {
+                var jitteredDelay = AddJitter(_crawlConfiguration.DefaultExternalRetryDelay);
+                _log.Warning("Request timeout for {0} (retry {1}), scheduling retry in {2}", 
+                    url, retryCount, jitteredDelay);
+                
+                scheduler.ScheduleTellOnce(jitteredDelay, self, new RetryExternalRequest(url, retryCount + 1), self);
+                
+                return new ExternalLinkRetryScheduled(url, HttpStatusCode.RequestTimeout);
+            }
+            
+            _log.Warning("Max retries ({0}) exceeded for external URL {1} after timeout", 
+                _crawlConfiguration.MaxExternalRetries, url);
+            return new ExternalLinkCrawled(url, HttpStatusCode.RequestTimeout);
+        }
         catch (Exception ex)
         {
             _log.Warning(ex, "Failed to crawl {0}", url);