Updated SgmlDomBuilder so that it respects significant whitespaces.

marek-stoj · May 29, 2013 · 8872e22 · 8872e22
1 parent aa51285
commit 8872e22
Show file tree

Hide file tree

Showing 3 changed files with 15 additions and 2 deletions.
diff --git a/.gitignore b/.gitignore
@@ -36,3 +36,4 @@
 /Src/NReadability/NReadability.6.1.ReSharper.user
 /Src/NReadability/NReadability.sln.DotSettings.user
 *.suo
+Src/NReadability/NReadability.sln.DotSettings
diff --git a/Src/NReadability/NReadability.Tests/SgmlDomBuilderTests.cs b/Src/NReadability/NReadability.Tests/SgmlDomBuilderTests.cs
@@ -1,4 +1,5 @@
-using System.Xml.Linq;
+using System;
+using System.Xml.Linq;
 using NUnit.Framework;
 
 namespace NReadability.Tests
@@ -54,6 +55,17 @@ public void Builder_handles_invalid_entity_references()
       Assert.DoesNotThrow(() => _sgmlDomSerializer.SerializeDocument(document));
     }
 
+    [Test]
+    public void Builder_respects_significant_whitespaces()
+    {
+      XDocument document =
+        _sgmlDomBuilder.BuildDocument("<a>Link1</a> <a>Link2</a>");
+
+      string html = _sgmlDomSerializer.SerializeDocument(document);
+
+      Assert.IsTrue(html.Contains("<a>Link1</a> <a>Link2</a>"));
+    }
+
     #endregion
   }
 }
diff --git a/Src/NReadability/NReadability/SgmlDomBuilder.cs b/Src/NReadability/NReadability/SgmlDomBuilder.cs
@@ -96,7 +96,7 @@ private static XDocument LoadDocument(string htmlContent)
       {
         sgmlReader.CaseFolding = CaseFolding.ToLower;
         sgmlReader.DocType = "HTML";
-        sgmlReader.WhitespaceHandling = WhitespaceHandling.None;
+        sgmlReader.WhitespaceHandling = WhitespaceHandling.Significant;
 
         using (var sr = new StreamReader(new MemoryStream(Encoding.UTF8.GetBytes(htmlContent))))
         {