Check Help Links Tool

using System;
using System.IO;
using System.Drawing;
using System.Collections;
using System.ComponentModel;
using System.Windows.Forms;
using System.Text.RegularExpressions;
using System.Diagnostics;

using mshtml;

using HtmlHelp;
using HtmlHelp.ChmDecoding;

using Common;

namespace CheckHelpLinks
{
	/// <summary>
	/// Summary description for DlgBrowser.
	/// </summary>
	public class DlgBrowser : System.Windows.Forms.Form
	{
		/// <summary>
		/// Required designer variable.
		/// </summary>
		private System.ComponentModel.Container components = null;

		private System.Windows.Forms.ProgressBar _ProgressBar;
		private System.Windows.Forms.TextBox _TextBoxUrl;
		private AxSHDocVw.AxWebBrowser _WebBrowser;

		private HtmlFileItem _HtmlFile = null;
		private HtmlFileItemCollection _HtmlFiles = null;
		private string _Url = null;
		private bool _AutoClose = true;
		private bool _GetLinks = true;

		private bool _Loading = false;

		public HtmlFileItem HtmlFile { get { return _HtmlFile; } set { _HtmlFile = value; } }
		public HtmlFileItemCollection HtmlFiles { get { return _HtmlFiles; } set { _HtmlFiles = value; } }
		public string Url { get { return _Url; } set { _Url = value; } }
		public bool AutoClose { get { return _AutoClose; } set { _AutoClose = value; } }
		public bool GetLinks { get { return _GetLinks; } set { _GetLinks = value; } }

		public static readonly Regex _Comment = new Regex( "^(?<before>.*)(?<comment><!--.*?-->)(?<after>.*)$", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Compiled );
		public static readonly Regex _Anchor  = new Regex( "<\\s*A\\s[^>]*name\\s*=\\s*(?<quote>['\"])(?<anchor>.*?)\\k<quote>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Compiled );
		public static readonly Regex _Anchor2 = new Regex( "<\\s*A\\s[^>]*name\\s*=\\s*(?<anchor>[^'\"].*?)[\\s>]", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Compiled );
//		public static readonly Regex _HRef    = new Regex( "href\\s*=\\s*['\"](?<url>[^'\">]*)", RegexOptions.IgnoreCase | RegexOptions.Compiled );
		public static readonly Regex _HRef    = new Regex( "<\\s*(?:A|AREA)\\s[^>]*href\\s*=\\s*(?<quote>['\"])(?<url>.*?)\\k<quote>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Compiled );
		public static readonly Regex _HRef2   = new Regex( "<\\s*(?:A|AREA)\\s[^>]*href\\s*=\\s*(?<url>[^'\"].*?)[\\s>]", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Compiled );

		public DlgBrowser()
		{
			InitializeComponent();
		}

		/// <summary>
		/// Clean up any resources being used.
		/// </summary>
		protected override void Dispose( bool disposing )
		{
			if( disposing )
			{
				if(components != null)
				{
					components.Dispose();
				}
			}
			base.Dispose( disposing );
		}

		#region Windows Form Designer generated code
		/// <summary>
		/// Required method for Designer support - do not modify
		/// the contents of this method with the code editor.
		/// </summary>
		private void InitializeComponent()
		{
			System.Resources.ResourceManager resources = new System.Resources.ResourceManager(typeof(DlgBrowser));
			this._ProgressBar = new System.Windows.Forms.ProgressBar();
			this._TextBoxUrl = new System.Windows.Forms.TextBox();
			this._WebBrowser = new AxSHDocVw.AxWebBrowser();
			((System.ComponentModel.ISupportInitialize)(this._WebBrowser)).BeginInit();
			this.SuspendLayout();
			// 
			// _ProgressBar
			// 
			this._ProgressBar.Anchor = ((System.Windows.Forms.AnchorStyles)(((System.Windows.Forms.AnchorStyles.Top | System.Windows.Forms.AnchorStyles.Left) 
				| System.Windows.Forms.AnchorStyles.Right)));
			this._ProgressBar.Location = new System.Drawing.Point(8, 8);
			this._ProgressBar.Name = "_ProgressBar";
			this._ProgressBar.Size = new System.Drawing.Size(616, 23);
			this._ProgressBar.TabIndex = 0;
			// 
			// _TextBoxUrl
			// 
			this._TextBoxUrl.Anchor = ((System.Windows.Forms.AnchorStyles)(((System.Windows.Forms.AnchorStyles.Top | System.Windows.Forms.AnchorStyles.Left) 
				| System.Windows.Forms.AnchorStyles.Right)));
			this._TextBoxUrl.Location = new System.Drawing.Point(8, 40);
			this._TextBoxUrl.Name = "_TextBoxUrl";
			this._TextBoxUrl.Size = new System.Drawing.Size(616, 20);
			this._TextBoxUrl.TabIndex = 1;
			this._TextBoxUrl.Text = "";
			this._TextBoxUrl.KeyPress += new System.Windows.Forms.KeyPressEventHandler(this._TextBoxUrl_KeyPress);
			// 
			// _WebBrowser
			// 
			this._WebBrowser.Anchor = ((System.Windows.Forms.AnchorStyles)((((System.Windows.Forms.AnchorStyles.Top | System.Windows.Forms.AnchorStyles.Bottom) 
				| System.Windows.Forms.AnchorStyles.Left) 
				| System.Windows.Forms.AnchorStyles.Right)));
			this._WebBrowser.Enabled = true;
			this._WebBrowser.Location = new System.Drawing.Point(8, 72);
			this._WebBrowser.OcxState = ((System.Windows.Forms.AxHost.State)(resources.GetObject("_WebBrowser.OcxState")));
			this._WebBrowser.Size = new System.Drawing.Size(616, 368);
			this._WebBrowser.TabIndex = 2;
			this._WebBrowser.DocumentComplete += new AxSHDocVw.DWebBrowserEvents2_DocumentCompleteEventHandler(this.WebBrowser_DocumentComplete);
			// 
			// DlgBrowser
			// 
			this.AutoScaleBaseSize = new System.Drawing.Size(5, 13);
			this.ClientSize = new System.Drawing.Size(632, 450);
			this.Controls.Add(this._WebBrowser);
			this.Controls.Add(this._TextBoxUrl);
			this.Controls.Add(this._ProgressBar);
			this.Icon = ((System.Drawing.Icon)(resources.GetObject("$this.Icon")));
			this.Name = "DlgBrowser";
			this.ShowInTaskbar = false;
			this.StartPosition = System.Windows.Forms.FormStartPosition.CenterParent;
			this.Text = "Getting Links";
			this.Load += new System.EventHandler(this.DlgBrowser_Load);
			((System.ComponentModel.ISupportInitialize)(this._WebBrowser)).EndInit();
			this.ResumeLayout(false);

		}
		#endregion

//-----------------------------------------------------------------------------

		private void DlgBrowser_Load(object sender, System.EventArgs e)
		{
			User32.PostMessage( Handle, Global.WM_GetLinks, 0, 0 );
		}

		protected override void WndProc( ref Message m )
		{
			if ( m.Msg == Global.WM_GetLinks )
			{
				GetLinksCore();
				return;
			}

			base.WndProc( ref m );
		}

//-----------------------------------------------------------------------------

		private void _TextBoxUrl_KeyPress(object sender, System.Windows.Forms.KeyPressEventArgs e)
		{
			if ( e.KeyChar == '\r' ) Navigate( _TextBoxUrl.Text );
		}

//-----------------------------------------------------------------------------

		private void GetLinksCore()
		{
			if ( _HtmlFile != null )
			{
				GetLinksCore( _HtmlFile );
			}

			if ( _HtmlFiles != null )
			{
				_ProgressBar.Maximum = _HtmlFiles.Count;
				_ProgressBar.Value = 0;

				int i = 0;
				foreach ( HtmlFileItem htmlFile in _HtmlFiles.Values )
				{
					_ProgressBar.Value = i++;
					GetLinksCore( htmlFile );
				}
			}

			if ( _Url != null ) Navigate( _Url );

			if ( _AutoClose ) DialogResult = DialogResult.OK;
		}

		private void GetLinksCore( HtmlFileItem htmlFile )
		{
			if ( Global.FastParse )
				GetLinksFast( htmlFile );
			else
				GetLinksSlow( htmlFile );
		}

		private void GetLinksFast( HtmlFileItem htmlFile )
		{
			htmlFile.AnchorsHeader.AnchorItems.Clear();
			htmlFile.Links.Clear();

			string html = htmlFile.Html;
			if ( html.Length == 0 ) return;

			for(;;)
			{
				MatchCollection comment = _Comment.Matches( html );
				if ( comment.Count != 1 ) break;
				string before = comment[ 0 ].Groups[ "before" ].Value;
				string after  = comment[ 0 ].Groups[ "after"  ].Value;
				html = before + after;
			}

			GetAnchorsFast( html, htmlFile );

			if ( _GetLinks ) GetLinksFast( html, htmlFile );
		}

		private void GetAnchorsFast( string html, HtmlFileItem htmlFile )
		{
			MatchCollection matches = null;
			
			// _Anchor
			matches = _Anchor.Matches( html );
			foreach ( Match match in matches )
			{
//				string sMatch = match.Value;
				GroupCollection groups = match.Groups;
				Group group = groups[ "anchor" ];

				string anchor = group.Value;

				AddAnchor( anchor, htmlFile );
			}

			// _Anchor2
			matches = _Anchor2.Matches( html );
			foreach ( Match match in matches )
			{
				GroupCollection groups = match.Groups;
				Group group = groups[ "anchor" ];

				string anchor = group.Value;

				AddAnchor( anchor, htmlFile );
			}
		}

		private void GetLinksFast( string html, HtmlFileItem htmlFile )
		{
			MatchCollection matches = null;

			// _HRef
			matches = _HRef.Matches( html );
			foreach ( Match match in matches )
			{
//				string sMatch = match.Value;
				GroupCollection groups = match.Groups;
				Group group = groups[ "url" ];

				string url = group.Value;

				AddLink( url, htmlFile );
			}

#if true
			// _HRef2
			matches = _HRef2.Matches( html );
			foreach ( Match match in matches )
			{
//				string sMatch = match.Value;
				GroupCollection groups = match.Groups;
				Group group = groups[ "url" ];

				string url = group.Value;

				AddLink( url, htmlFile );
			}
#endif
		}

		private void GetLinksSlow( HtmlFileItem htmlFile )
		{
			_Loading = true;

			Navigate( htmlFile.TopicEntry.URL );

			while ( _Loading ) { Application.DoEvents(); System.Threading.Thread.Sleep( 10 ); }

			htmlFile.AnchorsHeader.AnchorItems.Clear();
			htmlFile.Links.Clear();

			IHTMLDocument2 HTMLDocument = (IHTMLDocument2) _WebBrowser.Document;

			GetAnchorsSlow( HTMLDocument, htmlFile );

			if ( _GetLinks ) GetLinksSlow( HTMLDocument, htmlFile );
		}

		private void GetAnchorsSlow( IHTMLDocument2 HTMLDocument, HtmlFileItem htmlFile )
		{
			IHTMLElementCollection anchors = HTMLDocument.anchors;

			foreach ( object oAnchor in anchors )
			{
				string type = oAnchor.GetType().ToString();

				if ( oAnchor is HTMLAnchorElementClass )
				{
					HTMLAnchorElementClass el = ( HTMLAnchorElementClass ) oAnchor;

					string anchor = el.name;

					AddAnchor( anchor, htmlFile );

					continue;
				}

				Debug.WriteLine( "Unknown anchor type: '" + oAnchor.GetType() + "' in '" + htmlFile.TopicEntry.URL + "'." );
				Debug.Assert( false );
			}
		}

		private void GetLinksSlow( IHTMLDocument2 HTMLDocument, HtmlFileItem htmlFile )
		{
			IHTMLElementCollection links = HTMLDocument.links;

			foreach ( object oLink in links )
			{
				if ( oLink is HTMLAnchorElementClass )
				{
					HTMLAnchorElementClass el = ( HTMLAnchorElementClass ) oLink;

					LinkItem link = new HtmlLinkItem( htmlFile, el.href );

					htmlFile.Links.Add( link );

					continue;
				}

				if ( oLink is HTMLAreaElementClass )
				{
					HTMLAreaElementClass el = ( HTMLAreaElementClass ) oLink;

					LinkItem link = new HtmlLinkItem( htmlFile, el.href );

					htmlFile.Links.Add( link );

					continue;
				}

				Debug.WriteLine( "Unknown link type: '" + oLink.GetType() + "' in '" + htmlFile.TopicEntry.URL + "'." );
				Debug.Assert( false );
			}
		}

		private void WebBrowser_DocumentComplete(object sender, AxSHDocVw.DWebBrowserEvents2_DocumentCompleteEvent e)
		{
			_Loading = false;
		}

//-----------------------------------------------------------------------------

		private void AddAnchor( string anchor, HtmlFileItem htmlFile )
		{
			if ( anchor == null ) { Debug.Assert( false ); return; }

			AnchorItem existing = htmlFile.AnchorsHeader.AnchorItems[ anchor ] as AnchorItem;
			if ( existing != null )
			{
				existing.State = State.Broken;
				htmlFile.AnchorsHeader.State = State.Broken;
				return;
			}

			string url =
				"ms-its:" + htmlFile.Parent.ChmFile.ChmFilePath + "::/" +
				htmlFile.LinkFile + "#" + anchor;

			AnchorItem item = new AnchorItem( htmlFile.AnchorsHeader, url );

			htmlFile.AnchorsHeader.AnchorItems.Add( anchor, item );
		}

		private void AddLink( string url, HtmlFileItem htmlFile )
		{
			if ( Global.IgnoreCss && Global.IsCss( url ) ) return;

			if ( url.Length > 0 )
				if ( url[ 0 ] == '#' )
				{
					url = htmlFile.LinkFile + url;
				}
				else
					if ( url[ 0 ] != '/' )
				{
					int i2 = url.IndexOf( ':' );
					if ( i2 < 0 )
					{
						string dir = Path.GetDirectoryName( htmlFile.LinkFile );
						if ( dir.Length > 0 ) url = dir + "/" + url;
					}
				}

			int i = url.IndexOf( ':' );
			if ( i < 0 || i == 1 )
			{
				string s = "ms-its:" + htmlFile.Parent.ChmFile.ChmFilePath + "::";
				if ( url.Length > 0 && url[ 0 ] != '/' ) s += "/";
				s += url;
				url = s;
			}

//			if ( Global.IsHelp( url ) && ! Global.IsHtml( url ) )
//				continue;

			LinkItem link = new HtmlLinkItem( htmlFile, url );

			htmlFile.Links.Add( link );
		}

//-----------------------------------------------------------------------------

		private void Navigate( string url )
		{
			_TextBoxUrl.Text = url;

			object Zero = 0;
			object EmptyString = String.Empty;
			_WebBrowser.Navigate( url,
				ref Zero, ref EmptyString, ref EmptyString, ref EmptyString);

//			MessageBox.Show( url );
		}

//-----------------------------------------------------------------------------
	}

}