如何查询文本文件以获取模式的不同实例？

Question

我正在创建包含任意数量字符（人类字符/声音）的文档（参见 this），如下所示：

<span class="sam" title="This is Sam speaking">
<span class="higbie" title="This is Calvin Higbie speaking">
<span class="ballou" title="This is Mr. Ballou speaking">

对于某些上下文，这里是一个文档的 sn-p：

  <p><span class="others" title="This is 'an elderly pilgrim' speaking">"Jack, do you see that range of mountains over yonder that bounds the Jordan valley?  The mountains of Moab, Jack!  Think of it, my
  boy--the actual mountains of Moab--renowned in Scripture history!
  We are actually standing face to face with those illustrious crags
  and peaks--and for all we know" [dropping his voice impressively],
  "our eyes may be resting at this very moment upon the spot WHERE
  LIES THE MYSTERIOUS GRAVE OF MOSES!  Think of it, Jack!"</span></p>

当一个文档完成后，我想为这种标记模式生成一个不同的列表。 IOW，我想检查遵循该模式的每一段 HTML，但只返回每个不同的人/演讲者的一个实例。我不想要其中的 400 个：

<span class="sam" title="This is Sam speaking">

...（只有一个）。

在伪 SQL 术语中，我想要的是这样的：

SELECT DISTINCT SOMETHING FROM FILE WHERE SLIDING_WINDOW_OF_TEXT STARTSWITH("<span class=\"") AND SLIDING_WINDOW_OF_TEXT ENDSWITH("  speaking\">")

我不知道这是否是最好的使用正则表达式攻击的东西，或者是否有类似“LinqToText”之类的东西，或者其他东西......

Answer 1

这并不难。您可以使用LINQ 来获取Distinct() 值。添加引用和using System.Linq;/using System.Xml.Linq;。这是一个工作示例（在 VS2012 中测试）：

var MyRegex = new Regex(@"(?i)<span class=([""']).+?\1 title=([""']).+?\2>", RegexOptions.CultureInvariant | RegexOptions.Compiled);
var str = @"<p><span class=""others"" title=""This is 'an elderly pilgrim' speaking""><span class=""others"" title=""This is 'an elderly pilgrim' speaking""><span class=""others"" title=""This is 'an elderly pilgrim' speaking""><span class=""others"" title=""This is 'an elderly pilgrim' speaking""><span class=""higbie"" title=""This is Calvin Higbie speaking""><span class=""ballou"" title=""This is Mr. Ballou speaking""><span class=""ballou"" title=""This is Mr. Ballou speaking""><span class=""higbie"" title=""This is Calvin Higbie speaking""></p>";
var distinct_values = MyRegex.Matches(str).Cast<Match>().Select(p => p.Value).Distinct().ToList();

返回 3 个（不是 8 个）匹配项：

NO-LINQ SOLUTION

如果您不能使用 LINQ（例如在 Mono 中），您可以使用以下代码，该代码利用来自 System.Collections.Generic 的 List<string>：

using System.IO;
using System;
using System.Collections.Generic;
using System.Text.RegularExpressions;

class Program
{
    static void Main()
    {
        var MyRegex = new Regex(@"(?i)<span class=([""']).+?\1 title=([""']).+?\2>", RegexOptions.CultureInvariant | RegexOptions.Compiled);
        var str = @"<p><span class=""others"" title=""This is 'an elderly pilgrim' speaking""><span class=""others"" title=""This is 'an elderly pilgrim' speaking""><span class=""others"" title=""This is 'an elderly pilgrim' speaking""><span class=""others"" title=""This is 'an elderly pilgrim' speaking""><span class=""higbie"" title=""This is Calvin Higbie speaking""><span class=""ballou"" title=""This is Mr. Ballou speaking""><span class=""ballou"" title=""This is Mr. Ballou speaking""><span class=""higbie"" title=""This is Calvin Higbie speaking""></p>";
  //      var distinct_values = MyRegex.Matches(str).
//                    Cast<Match>().Select(p => p.Value).Distinct().ToList();
        var new_arr = new List<string>();
        var matches = MyRegex.Matches(str);
        for (int i=0; i<matches.Count; i++)
            if (!new_arr.Contains(matches[i].Value))
               new_arr.Add(matches[i].Value);

        Console.WriteLine(string.Join("\n", new_arr));
    }
}

输出：

<span class="others" title="This is 'an elderly pilgrim' speaking">                                                                                                 
<span class="higbie" title="This is Calvin Higbie speaking">                                                                                                        
<span class="ballou" title="This is Mr. Ballou speaking"> 

Answer 2

我建议您查看Html Agility Pack，它可以让您查询 html。这是一个例子：(Write query to parse HTML DOCUMENT with HtmlAgilityPack.)

您还可以使用 LinqToXml 将 html 元素作为 xml 节点进行查询。