W3C home > Mailing lists > Public > html-tidy@w3.org > January to March 2003

Using HTML Tidy in C# via Managed C++ Wrapper

From: Bjoern Hoehrmann <derhoermi@gmx.net>
Date: Fri, 24 Jan 2003 03:41:43 +0100
To: tidy-develop@lists.sourceforge.net
Message-ID: <3e3aa1be.144563791@smtp.bjoern.hoehrmann.de>

Hi,

  To take a closer look at the current HTML Tidy Library Code I wrote a
quick and dirty (and probably almost clueless) Managed C++ Wrapper for
the library that enables .NET programmers to use it the basic parsing
functionality. The static member function `ParseHtml` takes a .NET
`String` and returns the HTML document as an System.Xml.XmlDocument
object. You could use it in C# like

  using System;
  using System.Xml;
  
  class Class1
  {
      static void Main(string[] args)
      {
          String s = "<p>Hello World<pre>sample</p>...";
          XmlDocument doc = Tidy.HtmlParser.ParseHtml(s);
          doc.Save(Console.Out);
          Console.WriteLine();
      }
  }

The code lacks for almost anything, no namespace handling, no error
checks, no features like text nodes (!) etc.pp. but basically it seems
to work. Maybe some people like the idea of using HTML Tidy in their
.NET applications and will produce a more sophisticated wrapper, my code
is just a demonstration that one could get such thing working, somehow.

  #using <mscorlib.dll>
  #using <System.Xml.dll>
  #using <System.dll>
  
  #include "tidy.h"
  
  using namespace System;
  using namespace System::Text;
  using namespace System::Xml;
  
  uint tmbstrlen( ctmbstr str )
  {
      uint len = 0;
      while ( *str++ )
          ++len;
      return len;
  }
  
  namespace Tidy
  {
      public __gc class HtmlParser
      {
      public:
          static XmlDocument * ParseHtml(String * s);
      private:
          static void process_tree(TidyNode node,
                                   XmlDocument * doc,
                                   XmlNode * current);
      protected:
      };
  }
  
  XmlDocument * Tidy::HtmlParser::ParseHtml(String * s)
  {
      /* :-( */
      char gbc __gc[] = Encoding::UTF8->GetBytes(s);
      char __pin * ngbc = &gbc[0];
      char * input = (char *)ngbc;
  
      TidyDoc tdoc = tidyCreate();
  
      tidyParseString(tdoc, input);
      tidyCleanAndRepair(tdoc);
      tidyRunDiagnostics(tdoc);
  
      TidyNode root = tidyGetRoot(tdoc);
  
      XmlDocument * doc = new XmlDocument();
      process_tree(root, doc, 0);
  
      tidyRelease( tdoc );
      
      return doc;
  }
  
  void Tidy::HtmlParser::process_tree(TidyNode node,
                                      XmlDocument * doc,
                                      XmlNode * current)
  {
      TidyNode n = node;
  
      if (!doc || !node)
          return;
  
      if (!current)
          current = doc;
  
      while (n)
      {
          int t = tidyNodeGetType(n);
          TidyNode child = tidyGetChild(n);
  
          if (TidyNode_Root == t)
          {
              process_tree(child, doc, current);
          }
          else if (TidyNode_Start == t)
          {
              ctmbstr tname = tidyNodeGetName(n);
              String * name = new String(
                tname, 0, tmbstrlen(tname), Encoding::UTF8);
  
              XmlNode * e = doc->CreateElement(name);
  
              TidyAttr att;
              for (att = tidyAttrFirst(n); att; att = tidyAttrNext(att))
              {
                  ctmbstr tattname = tidyAttrName(att);
                  ctmbstr tattval = tidyAttrValue(att);
                  String * attname = new String(
                    tattname, 0, tmbstrlen(tattname), Encoding::UTF8);
                  String * attval = new String(
                    tattval, 0, tmbstrlen(tattval), Encoding::UTF8);
  
                  XmlAttribute * a = doc->CreateAttribute(attname);
                  a->Value = attval;
                  e->Attributes->Append(a);
              }
  
              current->AppendChild(e);
              process_tree(child, doc, e);
          }
  
          n = tidyGetNext(n);
      }
  }

The C# example should produce output like the following

  <?xml version="1.0" encoding="ibm850"?>
  <html>
    <head>
      <meta content=
       "HTML Tidy for Windows (vers 1st January 2003), see www.w3.org"
       name="generator" />
      <title />
    </head>
    <body>
      <p />
      <pre />
    </body>
  </html>

I had really liked to add text nodes to the XmlDocument, but TidyLib
does not seem to expose any functionality to access comment/procins/
element/... text content...
Received on Thursday, 23 January 2003 21:41:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 April 2012 06:13:53 GMT