lla*_*rov 21 .net xml linq unit-testing
在单元测试中我将一个XElement对象与我期望的对象进行比较.我使用的方法是调用.ToString()上XElement的对象,并将其与硬编码字符串值进行比较.这种方法非常不舒服,因为我总是要注意字符串中的格式.
我检查了XElement.DeepEquals()方法,但由于任何原因它没有帮助.
有谁知道我应该使用哪种方法最好?
Whe*_*lie 27
我发现这篇优秀文章很有用.它包含一个代码示例,它实现了XNode.DeepEquals在比较之前规范化XML树的替代方法,这使得非语义内容无关紧要.
为了说明,XNode.DeepEquals对于这些语义上等效的文档,返回false 的实现:
XElement root1 = XElement.Parse("<Root a='1' b='2'><Child>1</Child></Root>");
XElement root2 = XElement.Parse("<Root b='2' a='1'><Child>1</Child></Root>");
Run Code Online (Sandbox Code Playgroud)
但是,使用本文的实现DeepEqualsWithNormalization,您将获得该值,true因为属性的排序不被认为是重要的.该实现包括在下面.
using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using System.Text;
using System.Xml;
using System.Xml.Linq;
using System.Xml.Schema;
public static class MyExtensions
{
public static string ToStringAlignAttributes(this XDocument document)
{
XmlWriterSettings settings = new XmlWriterSettings();
settings.Indent = true;
settings.OmitXmlDeclaration = true;
settings.NewLineOnAttributes = true;
StringBuilder stringBuilder = new StringBuilder();
using (XmlWriter xmlWriter = XmlWriter.Create(stringBuilder, settings))
document.WriteTo(xmlWriter);
return stringBuilder.ToString();
}
}
class Program
{
private static class Xsi
{
public static XNamespace xsi = "http://www.w3.org/2001/XMLSchema-instance";
public static XName schemaLocation = xsi + "schemaLocation";
public static XName noNamespaceSchemaLocation = xsi + "noNamespaceSchemaLocation";
}
public static XDocument Normalize(XDocument source, XmlSchemaSet schema)
{
bool havePSVI = false;
// validate, throw errors, add PSVI information
if (schema != null)
{
source.Validate(schema, null, true);
havePSVI = true;
}
return new XDocument(
source.Declaration,
source.Nodes().Select(n =>
{
// Remove comments, processing instructions, and text nodes that are
// children of XDocument. Only white space text nodes are allowed as
// children of a document, so we can remove all text nodes.
if (n is XComment || n is XProcessingInstruction || n is XText)
return null;
XElement e = n as XElement;
if (e != null)
return NormalizeElement(e, havePSVI);
return n;
}
)
);
}
public static bool DeepEqualsWithNormalization(XDocument doc1, XDocument doc2,
XmlSchemaSet schemaSet)
{
XDocument d1 = Normalize(doc1, schemaSet);
XDocument d2 = Normalize(doc2, schemaSet);
return XNode.DeepEquals(d1, d2);
}
private static IEnumerable<XAttribute> NormalizeAttributes(XElement element,
bool havePSVI)
{
return element.Attributes()
.Where(a => !a.IsNamespaceDeclaration &&
a.Name != Xsi.schemaLocation &&
a.Name != Xsi.noNamespaceSchemaLocation)
.OrderBy(a => a.Name.NamespaceName)
.ThenBy(a => a.Name.LocalName)
.Select(
a =>
{
if (havePSVI)
{
var dt = a.GetSchemaInfo().SchemaType.TypeCode;
switch (dt)
{
case XmlTypeCode.Boolean:
return new XAttribute(a.Name, (bool)a);
case XmlTypeCode.DateTime:
return new XAttribute(a.Name, (DateTime)a);
case XmlTypeCode.Decimal:
return new XAttribute(a.Name, (decimal)a);
case XmlTypeCode.Double:
return new XAttribute(a.Name, (double)a);
case XmlTypeCode.Float:
return new XAttribute(a.Name, (float)a);
case XmlTypeCode.HexBinary:
case XmlTypeCode.Language:
return new XAttribute(a.Name,
((string)a).ToLower());
}
}
return a;
}
);
}
private static XNode NormalizeNode(XNode node, bool havePSVI)
{
// trim comments and processing instructions from normalized tree
if (node is XComment || node is XProcessingInstruction)
return null;
XElement e = node as XElement;
if (e != null)
return NormalizeElement(e, havePSVI);
// Only thing left is XCData and XText, so clone them
return node;
}
private static XElement NormalizeElement(XElement element, bool havePSVI)
{
if (havePSVI)
{
var dt = element.GetSchemaInfo();
switch (dt.SchemaType.TypeCode)
{
case XmlTypeCode.Boolean:
return new XElement(element.Name,
NormalizeAttributes(element, havePSVI),
(bool)element);
case XmlTypeCode.DateTime:
return new XElement(element.Name,
NormalizeAttributes(element, havePSVI),
(DateTime)element);
case XmlTypeCode.Decimal:
return new XElement(element.Name,
NormalizeAttributes(element, havePSVI),
(decimal)element);
case XmlTypeCode.Double:
return new XElement(element.Name,
NormalizeAttributes(element, havePSVI),
(double)element);
case XmlTypeCode.Float:
return new XElement(element.Name,
NormalizeAttributes(element, havePSVI),
(float)element);
case XmlTypeCode.HexBinary:
case XmlTypeCode.Language:
return new XElement(element.Name,
NormalizeAttributes(element, havePSVI),
((string)element).ToLower());
default:
return new XElement(element.Name,
NormalizeAttributes(element, havePSVI),
element.Nodes().Select(n => NormalizeNode(n, havePSVI))
);
}
}
else
{
return new XElement(element.Name,
NormalizeAttributes(element, havePSVI),
element.Nodes().Select(n => NormalizeNode(n, havePSVI))
);
}
}
}
Run Code Online (Sandbox Code Playgroud)
我开始和@llasarov一样走路,但也不喜欢使用字符串.我在这里发现了XElement.DeepEquals(),所以找到问题对我有帮助.
我可以看到,如果你的测试返回一个庞大的XML结构可能会很困难,但在我看来,这不应该做 - 测试应该检查尽可能小的结构.
假设您有一个方法,您希望返回一个看起来像的元素<Test Sample="Value" />.您可以使用XElement和XAttribute构造函数轻松构建期望值,如下所示:
[TestMethod()]
public void MyXmlMethodTest()
{
// Use XElement API to build expected element.
XElement expected = new XElement("Test", new XAttribute("Sample", "Value"));
// Call the method being tested.
XElement actual = MyXmlMethod();
// Assert using XNode.DeepEquals
Assert.IsTrue(XNode.DeepEquals(expected, actual));
}
Run Code Online (Sandbox Code Playgroud)
即使存在少量元素和属性,这也是可管理且一致的.
| 归档时间: |
|
| 查看次数: |
15524 次 |
| 最近记录: |