单元测试XML编码问题

Question

请考虑下面的类：

public class SampleXmlGenerator 
{ 
    public byte[] GenerateDocumentBytes() 
    { 
     byte[] fileBytes; 
     using (var xmlStream = new MemoryStream()) 
     { 
      using (var myWriter = new XmlTextWriter(xmlStream, Encoding.GetEncoding("UTF-8"))) 
      { 
       myWriter.Formatting = Formatting.Indented; 
       myWriter.Indentation = 4; 
       myWriter.IndentChar = ' '; 
       myWriter.WriteStartDocument(); 
       myWriter.WriteStartElement("foo"); 
       myWriter.WriteString("bar"); 
       myWriter.WriteEndElement(); // end foo 

       myWriter.Flush(); 

       fileBytes = xmlStream.ToArray(); 
      } 
     } 

     return fileBytes; 
    } 
} 

有了以下的单元测试：

[TestClass] 
public class TestSampleXmlGenerator 
{ 
    [TestMethod] 
    public void TextEmptyDocument() 
    { 
     var actualBytes = new SampleXmlGenerator().GenerateDocumentBytes(); 
     var actualUtf8String = Encoding.UTF8.GetString(actualBytes); 
     Console.Out.WriteLine("// actualUtf8String"); 
     Console.Out.WriteLine(actualUtf8String); 

     var actualDefaultString = Encoding.Default.GetString(actualBytes); 
     Console.Out.WriteLine("// actualDefaultString"); 
     Console.Out.WriteLine(actualDefaultString); 


     var expectedString = @"<?xml version=""1.0"" encoding=""utf-8""?> 
<foo>bar</foo>"; 
     var expectedBytes = Encoding.UTF8.GetBytes(expectedString); 

//  var expectedBytes = Encoding.Convert(Encoding.Default, Encoding.UTF8, Encoding.Default.GetBytes(@"<?xml version=""1.0"" encoding=""utf-8""?> 
//<foo>bar</foo>")); 
//  var expectedString = Encoding.UTF8.GetString(expectedBytes); 

     Console.Out.WriteLine("// expectedString"); 
     Console.Out.WriteLine(expectedString); 

     Assert.AreEqual(expectedBytes.Length, actualBytes.Length); 
     //Assert.AreEqual(expectedString, actualUtf8String); 
    } 
} 

最后输出：

Assert.AreEqual failed. Expected:<54>. Actual:<57>. 

// actualUtf8String 
<?xml version="1.0" encoding="utf-8"?> 
<foo>bar</foo> 

// actualDefaultString 
<?xml version="1.0" encoding="utf-8"?> 
<foo>bar</foo> 

// expectedString 
<?xml version="1.0" encoding="utf-8"?> 
<foo>bar</foo> 

expectedString和actualUtf8String看起来是一样的，但不是。

actualDefaultString显示开头的3个额外字符。

那么是什么给？我如何去测试/比较生成的XML？我应该做什么不同？

Answer 1

你看到的是BOM序言

https://msdn.microsoft.com/en-us/library/system.text.encoding.getpreamble(v=vs.110).aspx

您可以测试它，或者不序列化。

Answer 2

基于从马丁和基思的建议，并与一些额外的研究，我结束了从生成的XML字节的单元测试去除BOM，以下列方式（基于以下SO article）：

 var xmlBytes = new SampleXmlGenerator().GenerateDocumentBytes(); 
     var newXmlDoc = new XmlDocument {PreserveWhitespace = true}; 
     newXmlDoc.Load(new MemoryStream(xmlBytes)); 
     var actualBytes = Encoding.UTF8.GetBytes(newXmlDoc.OuterXml); 

现在单元测试经过！