在c＃中解析HTML部分

Question

I need to parse sections from a string of HTML. For example:

<p>Lorem ipsum dolor sit amet, consectetur adipiscing elit.</p>
<p>[section=quote]</p>
<p>Mauris at turpis nec dolor bibendum sollicitudin ac quis neque.</p>
<p>[/section]</p>

Parsing the quote section should return:

<p>Mauris at turpis nec dolor bibendum sollicitudin ac quis neque.</p>

Currently I'm using a regular expression to grab the content inside [section=quote]...[/section], but since the sections are entered using a WYSIWYG editor, the section tags themselves get wrapped in a paragraph tag, so the parsed result is:

</p>
<p>Mauris at turpis nec dolor bibendum sollicitudin ac quis neque.</p>
<p>

The Regular Expression I'm using currently is:

\[section=(.+?)\](.+?)\[/section\]

And I'm also doing some additional cleanup prior to parsing the sections:

protected string CleanHtml(string input) {
    // remove whitespace
    input = Regex.Replace(input, @"\s*(<[^>]+>)\s*", "$1", RegexOptions.Singleline);
    // remove empty p elements
    input = Regex.Replace(input, @"<p\s*/>|<p>\s*</p>", string.Empty);
    return input;
}

Can anyone provide a regular expression that would achieve what I am looking for or am I wasting my time trying to do this with Regex? I've seen references to the Html Agility Pack - would this be better for something like this?

[Update]

Thanks to Oscar I have used a combination of the HTML Agility pack and Regex to parse the sections. It still needs a bit of refining but it's nearly there.

public void ParseSections(string content)
{
    this.SourceContent = content;
    this.NonSectionedContent = content;

    content = CleanHtml(content);

    if (!sectionRegex.IsMatch(content))
        return;

    var doc = new HtmlDocument();
    doc.LoadHtml(content);

    bool flag = false;
    string sectionName = string.Empty;
    var sectionContent = new StringBuilder();
    var unsectioned = new StringBuilder();

    foreach (var n in doc.DocumentNode.SelectNodes("//p")) {               
        if (startSectionRegex.IsMatch(n.InnerText)) { 
            flag = true;
            sectionName = startSectionRegex.Match(n.InnerText).Groups[1].Value.ToLowerInvariant();
            continue;
        }
        if (endSectionRegex.IsMatch(n.InnerText)) {
            flag = false;
            this.Sections.Add(sectionName, sectionContent.ToString());
            sectionContent.Clear();
            continue;
        }

        if (flag)
            sectionContent.Append(n.OuterHtml);
        else
            unsectioned.Append(n.OuterHtml);
    }

    this.NonSectionedContent = unsectioned.ToString();
}

Answer 1

The following works, using HtmlAgilityPack library:

using HtmlAgilityPack;

...

HtmlDocument doc = new HtmlDocument();
doc.Load(@"C:\file.html");


bool flag = false;
var sb = new StringBuilder();
foreach (var n in doc.DocumentNode.SelectNodes("//p"))
{
    switch (n.InnerText)
    {
        case "[section=quote]":
            flag = true;
            continue;
        case "[/section]":
            flag = false;
            break;
    }
    if (flag)
    {
        sb.AppendLine(n.OuterHtml);
    }
}

Console.Write(sb);
Console.ReadLine();

If you just want to print Mauris at turpis nec dolor bibendum sollicitudin ac quis neque. without <p>...</p> , you can replace n.OuterHtml by n.InnerHtml .

Of course, you should check if doc.DocumentNode.SelectNodes("//p") is null .
If you want to load the html from an online source instead of a file, you can do:

var htmlWeb = new HtmlWeb();  
var doc = htmlWeb.Load("http://..../page.html");

If [section=quote] an [/section] could be inside any tag (not always <p> ), you can replace doc.DocumentNode.SelectNodes("//p") by doc.DocumentNode.SelectNodes("//*") .

Answer 2

How about replacing

<p>[section=quote]</p>

with

[section=quote]

and

<p>[/section]</p>

with

[/section]

as part of your cleanup. Then you can use your existing regular expression.