fix incorrect parsing of script element

mgambill · mgambill · commit 7e66df20b44b · 2021-05-16T14:29:07.000-04:00
diff --git a/HtmlParser.Tests/ParserTests.cs b/HtmlParser.Tests/ParserTests.cs
@@ -2,6 +2,7 @@
 using System;
 using System.Collections.Generic;
 using System.Linq;
+using System.Security.Cryptography.X509Certificates;
 using System.Text;
 
 namespace NMyVision.HtmlParserTests
@@ -78,5 +79,29 @@ public void MultipleRunsTest()
             Assert.IsTrue(el.HasChildren());
 
         }
+
+        [TestMethod()]
+        public void LargeHtml()
+        {
+            var html = @"<html lang=""en"">
+    <head>
+        <title>Pug</title>
+        <script type=""text/javascript""> if (foo) bar(1 + 5) </script>
+    </head>
+    <body>
+        <h1>Pug - node template engine</h1>
+        <div id=""container"" class=""col"">
+            <p>You are amazing</p>
+            <p>Pug is a terse and simple templating language.</p>
+        </div>
+    </body>
+</html>";
+
+            var el = new HtmlParser().Parse(html).First();
+
+            Assert.AreEqual( html, el.OuterHTML );
+
+
+        }
     }
 }
diff --git a/HtmlParser/NMyVision.HtmlParser.csproj b/HtmlParser/NMyVision.HtmlParser.csproj
@@ -11,7 +11,7 @@
     <PackageLicenseFile>LICENSE.txt</PackageLicenseFile>
     <RepositoryType></RepositoryType>
     <RepositoryUrl>https://github.com/NMyVision/HtmlParser</RepositoryUrl>
-    <Version>1.0.4</Version>
+    <Version>1.0.5</Version>
 
 
     <!-- Optional: Publish the repository URL in the built .nupkg (in the NuSpec <Repository> element) -->
diff --git a/HtmlParser/Parser.cs b/HtmlParser/Parser.cs
@@ -110,9 +110,11 @@ private IEnumerable<HtmlNode> InternalParse(HtmlNode parent = null)
 					}
 					else if (node.Tag.Equals("script", StringComparison.OrdinalIgnoreCase))
 					{
-						Dequeue();
+						Dequeue(); // >
 						node.Content = GetUpTo("</script");
 						node.EndTag = "script";
+						Dequeue(9); // </script>
+						node.SetEndIndex(index, Source);
 						list.Add(node);
 					}
 				}

Original file line number	Diff line number	Diff line change
`@@ -110,9 +110,11 @@ private IEnumerable<HtmlNode> InternalParse(HtmlNode parent = null)`
`110`	`110`	`}`
`111`	`111`	`else if (node.Tag.Equals("script", StringComparison.OrdinalIgnoreCase))`
`112`	`112`	`{`
`113`		`- Dequeue();`
	`113`	`+ Dequeue(); // >`
`114`	`114`	`node.Content = GetUpTo("</script");`
`115`	`115`	`node.EndTag = "script";`
	`116`	`+ Dequeue(9); // </script>`
	`117`	`+ node.SetEndIndex(index, Source);`
`116`	`118`	`list.Add(node);`
`117`	`119`	`}`
`118`	`120`	`}`