Fixed a error #73 “HtmlMinifier.Minify hangs permanently”

WebMarkupMin.NoSamples.sln

@@ -15,6 +15,7 @@ Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "html", "html", "{E16C8CAD-9
 	ProjectSection(SolutionItems) = preProject
 		test\SharedFiles\html\html-document-with-bom-at-start.html = test\SharedFiles\html\html-document-with-bom-at-start.html
 		test\SharedFiles\html\html-document-with-bom-in-body-tag.html = test\SharedFiles\html\html-document-with-bom-in-body-tag.html
+		test\SharedFiles\html\html-document-with-invalid-characters.html = test\SharedFiles\html\html-document-with-invalid-characters.html
 		test\SharedFiles\html\html-document-without-bom.html = test\SharedFiles\html\html-document-without-bom.html
 	EndProjectSection
 EndProject
@@ -49,7 +50,7 @@ Project("{9A19103F-16F7-4668-BE54-9A1E7A4F7556}") = "WebMarkupMin.AspNetCore1",
 EndProject
 Project("{9A19103F-16F7-4668-BE54-9A1E7A4F7556}") = "WebMarkupMin.AspNetCore2", "src\WebMarkupMin.AspNetCore2\WebMarkupMin.AspNetCore2.csproj", "{FB743652-E1C4-40B3-B499-53780DCBAA8F}"
 EndProject
-Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "WebMarkupMin.Benchmarks", "test\WebMarkupMin.Benchmarks\WebMarkupMin.Benchmarks.csproj", "{213B17DB-6990-43CA-9585-A1557C6D27E8}"
+Project("{9A19103F-16F7-4668-BE54-9A1E7A4F7556}") = "WebMarkupMin.Benchmarks", "test\WebMarkupMin.Benchmarks\WebMarkupMin.Benchmarks.csproj", "{213B17DB-6990-43CA-9585-A1557C6D27E8}"
 EndProject
 Project("{9A19103F-16F7-4668-BE54-9A1E7A4F7556}") = "WebMarkupMin.Core.Test", "test\WebMarkupMin.Core.Test\WebMarkupMin.Core.Test.csproj", "{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68}"
 EndProject
@@ -107,14 +108,14 @@ Global
 		{FB743652-E1C4-40B3-B499-53780DCBAA8F}.Debug|Any CPU.Build.0 = Debug|Any CPU
 		{FB743652-E1C4-40B3-B499-53780DCBAA8F}.Release|Any CPU.ActiveCfg = Release|Any CPU
 		{FB743652-E1C4-40B3-B499-53780DCBAA8F}.Release|Any CPU.Build.0 = Release|Any CPU
-		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
-		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68}.Debug|Any CPU.Build.0 = Debug|Any CPU
-		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68}.Release|Any CPU.ActiveCfg = Release|Any CPU
-		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68}.Release|Any CPU.Build.0 = Release|Any CPU
 		{213B17DB-6990-43CA-9585-A1557C6D27E8}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
 		{213B17DB-6990-43CA-9585-A1557C6D27E8}.Debug|Any CPU.Build.0 = Debug|Any CPU
 		{213B17DB-6990-43CA-9585-A1557C6D27E8}.Release|Any CPU.ActiveCfg = Release|Any CPU
 		{213B17DB-6990-43CA-9585-A1557C6D27E8}.Release|Any CPU.Build.0 = Release|Any CPU
+		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
+		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68}.Debug|Any CPU.Build.0 = Debug|Any CPU
+		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68}.Release|Any CPU.ActiveCfg = Release|Any CPU
+		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68}.Release|Any CPU.Build.0 = Release|Any CPU
 	EndGlobalSection
 	GlobalSection(SolutionProperties) = preSolution
 		HideSolutionNode = FALSE
@@ -135,8 +136,8 @@ Global
 		{27D7454A-F570-4ACC-BB7E-1DD5B264AC22} = {227F2DF0-FCC8-4548-A4F9-6DA5BE01918E}
 		{5F014AC3-7F5D-4EBB-850C-01513CBF1D49} = {227F2DF0-FCC8-4548-A4F9-6DA5BE01918E}
 		{FB743652-E1C4-40B3-B499-53780DCBAA8F} = {227F2DF0-FCC8-4548-A4F9-6DA5BE01918E}
-		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68} = {1AEA1D1F-45BC-4A6C-BCD8-71C50DC839BA}
 		{213B17DB-6990-43CA-9585-A1557C6D27E8} = {1AEA1D1F-45BC-4A6C-BCD8-71C50DC839BA}
+		{9125BDF2-885B-45DB-98AA-3DD8C2EA4A68} = {1AEA1D1F-45BC-4A6C-BCD8-71C50DC839BA}
 	EndGlobalSection
 	GlobalSection(ExtensibilityGlobals) = postSolution
 		SolutionGuid = {7E056B7D-640D-49E5-BFD7-D64C25744378}

src/WebMarkupMin.Core/Parsers/HtmlParser.cs

@@ -29,18 +29,12 @@ internal sealed class HtmlParser : MarkupParserBase
 		#region Regular expressions for parsing tags and attributes
 
 		private static readonly Regex _xmlDeclarationRegex = new Regex(@"^<\?xml\s+[^>]+\s*\?>", RegexOptions.IgnoreCase);
-		private static readonly Regex _startTagRegex = new Regex(@"^<(?<tagName>" + CommonRegExps.HtmlTagNamePattern + ")" +
-			"(?:" +
-				"(?:" +
-					@"(?:\s+|(?<=[""']))" + CommonRegExps.HtmlAttributeNamePattern +
-					@"(?:\s*=\s*(?:(?:""[^""]*"")|(?:'[^']*')|[^\s""'`=<>]+)?)?" +
-				")" +
-				@"|(?:\s*(?<invalidCharacters>(?:[^/>\s][^>\s]*?)|(?:/[^>\s]*?(?!>))))" +
-			")*" +
-			@"\s*(?<emptyTagSlash>/)?>");
+		private static readonly Regex _startTagBeginPartRegex = new Regex(@"^<(?<tagName>" + CommonRegExps.HtmlTagNamePattern + ")");
+		private static readonly Regex _startTagEndPartRegex = new Regex(@"^\s*(?<emptyTagSlash>/)?>");
 		private static readonly Regex _endTagRegex = new Regex(@"^<\/(?<tagName>" + CommonRegExps.HtmlTagNamePattern + @")\s*>");
 		private static readonly Regex _attributeRegex =
-			new Regex(@"(?<attributeName>" + CommonRegExps.HtmlAttributeNamePattern + @")" +
+			new Regex(@"(?:\s+|(?<=[""']))" +
+				"(?<attributeName>" + CommonRegExps.HtmlAttributeNamePattern + @")" +
 				"(?:" +
 					@"\s*(?<attributeEqualSign>=)\s*" +
 					"(?:" +
@@ -533,49 +527,51 @@ private bool ProcessStartTag()
 		{
 			bool isProcessed = false;
 			string content = _innerContext.SourceCode;
-			int contentPosition = _innerContext.Position;
-			int contentRemainderLength = _innerContext.RemainderLength;
 
-			var match = _startTagRegex.Match(content, contentPosition, contentRemainderLength);
-			if (match.Success)
+			Match startTagBeginPartMatch = _startTagBeginPartRegex.Match(content, _innerContext.Position,
+				_innerContext.RemainderLength);
+			if (startTagBeginPartMatch.Success)
 			{
-				GroupCollection groups = match.Groups;
-
-				Group startTagNameGroup = groups["tagName"];
-				string startTagName = startTagNameGroup.Value;
+				string startTagName = startTagBeginPartMatch.Groups["tagName"].Value;
 				string startTagNameInLowercase = startTagName;
 				if (Utils.ContainsUppercaseCharacters(startTagName))
 				{
 					startTagNameInLowercase = startTagName.ToLowerInvariant();
 				}
 
-				Group invalidCharactersGroup = groups["invalidCharacters"];
-				if (invalidCharactersGroup.Success)
-				{
-					int invalidCharactersPosition = invalidCharactersGroup.Index;
-					int invalidCharactersOffset = invalidCharactersPosition - contentPosition;
+				_innerContext.IncreasePosition(startTagBeginPartMatch.Length);
 
-					_innerContext.IncreasePosition(invalidCharactersOffset);
+				IList<HtmlAttribute> attributes = ProcessAttributes();
 
-					throw new MarkupParsingException(
-						string.Format(Strings.ErrorMessage_InvalidCharactersInStartTag, startTagName),
-						_innerContext.NodeCoordinates, _innerContext.GetSourceFragment());
-				}
+				int currentPosition = _innerContext.Position;
+				int currentRemainderLength = _innerContext.RemainderLength;
 
-				int attributesPosition = startTagNameGroup.Index + startTagNameGroup.Length;
-				int attributesOffset = attributesPosition - contentPosition;
-				int startTagRemainderLength = contentPosition + match.Length - attributesPosition;
+				Match startTagEndPartMatch = _startTagEndPartRegex.Match(content, currentPosition,
+					currentRemainderLength);
+				if (startTagEndPartMatch.Success)
+				{
+					bool isEmptyTag = startTagEndPartMatch.Groups["emptyTagSlash"].Success;
 
-				_innerContext.IncreasePosition(attributesOffset);
+					ParseStartTag(startTagName, startTagNameInLowercase, attributes, isEmptyTag);
 
-				IList<HtmlAttribute> attributes = ParseAttributes(content, attributesPosition,
-					startTagRemainderLength, _innerContext.NodeCoordinates);
-				bool isEmptyTag = groups["emptyTagSlash"].Success;
+					_innerContext.IncreasePosition(startTagEndPartMatch.Length);
+					isProcessed = true;
+				}
+				else
+				{
+					int invalidCharPosition = SourceCodeNavigator.FindNextNonWhitespaceChar(content,
+						currentPosition, currentRemainderLength);
 
-				ParseStartTag(startTagName, startTagNameInLowercase, attributes, isEmptyTag);
+					int invalidCharOffset = invalidCharPosition - currentPosition;
+					if (invalidCharOffset > 0)
+					{
+						_innerContext.IncreasePosition(invalidCharOffset);
+					}
 
-				_innerContext.IncreasePosition(startTagRemainderLength);
-				isProcessed = true;
+					throw new MarkupParsingException(
+						string.Format(Strings.ErrorMessage_InvalidCharactersInStartTag, startTagName),
+						_innerContext.NodeCoordinates, _innerContext.GetSourceFragment());
+				}
 			}
 
 			return isProcessed;
@@ -855,23 +851,26 @@ private void ParseStartTag(string tagName, string tagNameInLowercase, IList<Html
 		}
 
 		/// <summary>
-		/// Parses a attributes
+		/// Process a attributes
 		/// </summary>
-		/// <param name="sourceCode">Source code</param>
-		/// <param name="attributesPosition">Start position of attributes</param>
-		/// <param name="startTagRemainderLength">Length of attributes and remaining characters in start tag</param>
-		/// <param name="attributesCoordinates">Attributes coordinates</param>
 		/// <returns>List of attributes</returns>
-		private IList<HtmlAttribute> ParseAttributes(string sourceCode, int attributesPosition,
-			int startTagRemainderLength, SourceCodeNodeCoordinates attributesCoordinates)
+		private IList<HtmlAttribute> ProcessAttributes()
 		{
-			Match match = _attributeRegex.Match(sourceCode, attributesPosition, startTagRemainderLength);
+			string content = _innerContext.SourceCode;
+			int currentPosition = _innerContext.Position;
+			int currentRemainderLength = _innerContext.RemainderLength;
+			SourceCodeNodeCoordinates currentCoordinates = _innerContext.NodeCoordinates;
+
+			Match match = _attributeRegex.Match(content, currentPosition, currentRemainderLength);
 			var attributes = new List<HtmlAttribute>();
-			int currentPosition = attributesPosition;
-			SourceCodeNodeCoordinates currentCoordinates = attributesCoordinates;
 
 			while (match.Success)
 			{
+				if (match.Index != _innerContext.Position)
+				{
+					break;
+				}
+
 				GroupCollection groups = match.Groups;
 				Group attributeNameGroup = groups["attributeName"];
 				Group attributeEqualSignGroup = groups["attributeEqualSign"];
@@ -913,7 +912,7 @@ private IList<HtmlAttribute> ParseAttributes(string sourceCode, int attributesPo
 					int lineBreakCount;
 					int charRemainderCount;
 
-					SourceCodeNavigator.CalculateLineBreakCount(sourceCode, currentPosition,
+					SourceCodeNavigator.CalculateLineBreakCount(content, currentPosition,
 						attributeNamePosition - currentPosition, out lineBreakCount, out charRemainderCount);
 					attributeNameCoordinates = SourceCodeNavigator.CalculateAbsoluteNodeCoordinates(
 						currentCoordinates, lineBreakCount, charRemainderCount);
@@ -934,7 +933,7 @@ private IList<HtmlAttribute> ParseAttributes(string sourceCode, int attributesPo
 					int lineBreakCount;
 					int charRemainderCount;
 
-					SourceCodeNavigator.CalculateLineBreakCount(sourceCode, currentPosition,
+					SourceCodeNavigator.CalculateLineBreakCount(content, currentPosition,
 						attributeValuePosition - currentPosition, out lineBreakCount, out charRemainderCount);
 					attributeValueCoordinates = SourceCodeNavigator.CalculateAbsoluteNodeCoordinates(
 						currentCoordinates, lineBreakCount, charRemainderCount);
@@ -947,6 +946,7 @@ private IList<HtmlAttribute> ParseAttributes(string sourceCode, int attributesPo
 					HtmlAttributeType.Unknown, attributeNameCoordinates, attributeValueCoordinates);
 				attributes.Add(attribute);
 
+				_innerContext.IncreasePosition(match.Length);
 				match = match.NextMatch();
 			}
 

src/WebMarkupMin.Core/Utilities/SourceCodeNavigator.cs

@@ -109,6 +109,31 @@ private static void FindPreviousLineBreak(string sourceCode, int startPosition,
 			}
 		}
 
+		/// <summary>
+		/// Finds a next non-whitespace character
+		/// </summary>
+		/// <param name="sourceCode">Source code</param>
+		/// <param name="startPosition">Position in the input string that defines the leftmost
+		/// position to be searched</param>
+		/// <param name="length">Number of characters in the substring to include in the search</param>
+		/// <returns>Position of non-whitespace character</returns>
+		internal static int FindNextNonWhitespaceChar(string sourceCode, int startPosition, int length)
+		{
+			int charPosition;
+			int endPosition = startPosition + length - 1;
+
+			for (charPosition = startPosition; charPosition <= endPosition; charPosition++)
+			{
+				char charValue = sourceCode[charPosition];
+				if (!char.IsWhiteSpace(charValue))
+				{
+					break;
+				}
+			}
+
+			return charPosition;
+		}
+
 		/// <summary>
 		/// Calculates a line break count
 		/// </summary>

src/WebMarkupMin.Core/WebMarkupMin.Core.csproj

@@ -18,7 +18,8 @@ Minification of markup produces by removing extra whitespaces, comments and redu
 
 Also supports minification of views of popular JavaScript template engines: KnockoutJS, Kendo UI MVVM and AngularJS 1.X.</Description>
 		<PackageTags>WebMarkupMin;Markup;HTML;XHTML;XML;Minification;Minifier;Minify;Performance;Optimization;Compression</PackageTags>
-		<PackageReleaseNotes>Fixed a error #77 “HtmlMinifier.Minify throws InvalidOperationException”.</PackageReleaseNotes>
+		<PackageReleaseNotes>1. Fixed a error #73 “HtmlMinifier.Minify hangs permanently”;
+2. Fixed a error #77 “HtmlMinifier.Minify throws InvalidOperationException”.</PackageReleaseNotes>
 	</PropertyGroup>
 
 	<Import Project="../../build/common.props" />

src/WebMarkupMin.Core/readme.txt

@@ -46,7 +46,8 @@
    =============
    RELEASE NOTES
    =============
-   Fixed a error #77 “HtmlMinifier.Minify throws InvalidOperationException”.
+   1. Fixed a error #73 “HtmlMinifier.Minify hangs permanently”;
+   2. Fixed a error #77 “HtmlMinifier.Minify throws InvalidOperationException”.
 
    =============
    DOCUMENTATION

test/WebMarkupMin.Core.Test/Html/ParsingTests.cs

@@ -1,11 +1,21 @@
 using System.Collections.Generic;
+using System.IO;
 
 using Xunit;
 
 namespace WebMarkupMin.Core.Test.Html
 {
-	public class ParsingTests
+	public class ParsingTests : FileSystemTestsBase
 	{
+		private readonly string _htmlFilesDirectoryPath;
+
+
+		public ParsingTests()
+		{
+			_htmlFilesDirectoryPath = Path.GetFullPath(Path.Combine(_baseDirectoryPath, @"../SharedFiles/html/"));
+		}
+
+
 		[Fact]
 		public void ParsingNonTrivialMarkupIsCorrect()
 		{
@@ -259,6 +269,8 @@ public void ProcessingInvalidCharactersInStartTagIsCorrect()
 				"</table."
 				;
 			const string input11 = "<link id=\"favicon\" rel=?\"shortcut icon\" type=?\"image/?png\" href=?\"#\">";
+			string input12 = File.ReadAllText(
+				Path.Combine(_htmlFilesDirectoryPath, "html-document-with-invalid-characters.html"));
 
 			// Act
 			IList<MinificationErrorInfo> errors1 = minifier.Minify(input1).Errors;
@@ -272,6 +284,7 @@ public void ProcessingInvalidCharactersInStartTagIsCorrect()
 			IList<MinificationErrorInfo> errors9 = minifier.Minify(input9).Errors;
 			IList<MinificationErrorInfo> errors10 = minifier.Minify(input10).Errors;
 			IList<MinificationErrorInfo> errors11 = minifier.Minify(input11).Errors;
+			IList<MinificationErrorInfo> errors12 = minifier.Minify(input12).Errors;
 
 			// Assert
 			Assert.Equal(1, errors1.Count);
@@ -284,7 +297,7 @@ public void ProcessingInvalidCharactersInStartTagIsCorrect()
 
 			Assert.Equal(1, errors3.Count);
 			Assert.Equal(1, errors3[0].LineNumber);
-			Assert.Equal(205, errors3[0].ColumnNumber);
+			Assert.Equal(120, errors3[0].ColumnNumber);
 
 			Assert.Equal(1, errors4.Count);
 			Assert.Equal(1, errors4[0].LineNumber);
@@ -304,7 +317,7 @@ public void ProcessingInvalidCharactersInStartTagIsCorrect()
 
 			Assert.Equal(1, errors8.Count);
 			Assert.Equal(2, errors8[0].LineNumber);
-			Assert.Equal(80, errors8[0].ColumnNumber);
+			Assert.Equal(46, errors8[0].ColumnNumber);
 
 			Assert.Equal(1, errors9.Count);
 			Assert.Equal(1, errors9[0].LineNumber);
@@ -316,7 +329,11 @@ public void ProcessingInvalidCharactersInStartTagIsCorrect()
 
 			Assert.Equal(1, errors11.Count);
 			Assert.Equal(1, errors11[0].LineNumber);
-			Assert.Equal(68, errors11[0].ColumnNumber);
+			Assert.Equal(25, errors11[0].ColumnNumber);
+
+			Assert.Equal(1, errors12.Count);
+			Assert.Equal(1, errors12[0].LineNumber);
+			Assert.Equal(3, errors12[0].ColumnNumber);
 		}
 
 		[Fact]