Question

我尝试使用正则表达式进行解析，但并未解析所有XML文件!!

标签名称未预定义，这意味着，解决方案应该自己识别XML标签的开头和结尾

示例数据如下：

<XML data> 
Random text 
<XML data> 
Random text

Answer 1

正如我在评论中所说，这是一个难题，尤其是在不了解您的数据的情况下。可能会有很多边缘案例。

以下代码可能适用于您想要的内容。它试图找到元素（'＆lt; element ...'或'＆lt; element /＆gt;'，然后通过查找结束标记来确定XML内容。它通过尝试解析XML来检查它，如果错误则拒绝它XML。

此代码是为了清晰而编写的，而不是性能或良好的结构（它是所有主线的下降）。这应该为您提供一个示例，您可以将其用作您所描述的解析类型的良好起点。

using System;
using System.Linq;
using System.Text;
using System.Xml.Linq;

namespace messyXml {
class Program {

    const String almostXml = @"
@#$%random junk
<Fruits>
  <Apples>
    Pies
  </Apples>
  <Pears>
    Tarts
  </Pears>
</Fruits>
This is some junk about Fruits like Apples and Pears 
which can be made into Pies and Tarts.  I think that
Pears<Apples.  Edge case might be <Parts or /Apples>
<Parts No='123'>
  Pie Plate
</Parts>";

    static void Main(string[] args) {
        Console.WriteLine("Extracting XML from:");
        Console.WriteLine(almostXml);
        Console.WriteLine();
        int i = 0;
        var validXml = new StringBuilder();
        while (i < almostXml.Length) {
            if (almostXml[i] == '<') { // might be an xml start
                int ix = almostXml.IndexOfAny(" >\t".ToArray(), i + 1); 
                // this only check for space, > and tab, you may want to
                // include other whitespace chars

                if (ix < 0) {
                    ix = almostXml.IndexOf("/>", i + 1); // you might have <element/>
                    if (ix >= 0) {
                        // you could check if element name is valid
                        var xml = almostXml.Substring(i, (ix + 2) - i);
                        try {
                            // see if this is really xml
                            var doc = XDocument.Parse(xml);
                            validXml.AppendLine(xml);
                            i = ix + 2;
                            continue; // next iteration of while i 
                            }
                        catch (System.Xml.XmlException) {
                            // do nothing
                            }
                        }
                    ix = almostXml.IndexOf(">", i + 1); // you might have <element/>
                    }
                else { // we found something like <element ...
                    var ix2 = almostXml.IndexOf('>', ix); // where is the '>'
                    if (ix2 >= 0) {
                        // build an end tag
                        var endtag = "</" + almostXml.Substring(i + 1, (ix - i) - 1) + ">";
                        var endix = almostXml.IndexOf(endtag, ix2);
                        if (endix >= 0) {
                            var xml = almostXml.Substring(i, (endix + endtag.Length) - i);
                            try {
                                // see if this is really xml
                                var doc = XDocument.Parse(xml);
                                validXml.AppendLine(xml);
                                i = (endix + endtag.Length);
                                continue; // next iteration of while i
                                }
                            catch (System.Xml.XmlException) {
                                // do nothing
                                }
                            }
                        }
                    }
                }
            i++;
            }
        Console.WriteLine("-----------");
        Console.WriteLine("Valid XML found:");
        Console.WriteLine(validXml.ToString());
        Console.ReadKey();
        }
    }
}

如何从包含多余数据和垃圾的文件中删除有效的XML

1 个答案: