public DocHTML(HTMLDocument doc)
{
mDoc = doc;
mDoc2 = (IHTMLDocument2)mDoc;
mDoc3 = (IHTMLDocument3)mDoc;
mDoc4 = (IHTMLDocument4)mDoc;
mDoc5 = (IHTMLDocument5)mDoc;
}
public DocHTML(HTMLDocument doc)
{
this.mDoc = doc;
this.mDoc2 = (IHTMLDocument2)this.mDoc;
this.mDoc3 = (IHTMLDocument3)this.mDoc;
this.mDoc4 = (IHTMLDocument4)this.mDoc;
this.mDoc5 = (IHTMLDocument5)this.mDoc;
}
public DocHTML(HTMLDocument doc)
{
mDoc = doc;
mDoc2 = (IHTMLDocument2)mDoc;
mDoc3 = (IHTMLDocument3)mDoc;
mDoc4 = (IHTMLDocument4)mDoc;
mDoc5 = (IHTMLDocument5)mDoc;
}
public DocumentAdapter( object document )
{
//_documentUri = documentUri;
_raw = document;
_document = document as IHTMLDocument;
_document2 = document as IHTMLDocument2;
_document3 = document as IHTMLDocument3;
_document4 = document as IHTMLDocument4;
_document5 = document as IHTMLDocument5;
}
public DocumentAdapter(object document)
{
//_documentUri = documentUri;
_raw = document;
_document = document as IHTMLDocument;
_document2 = document as IHTMLDocument2;
_document3 = document as IHTMLDocument3;
_document4 = document as IHTMLDocument4;
_document5 = document as IHTMLDocument5;
}
private static object getDocumentAttribute(IHTMLDocument2 theHTMLDocument, string theAttributeName)
{
IHTMLDocument5 doc5 = (IHTMLDocument5)theHTMLDocument;
IHTMLDocument3 doc3 = (IHTMLDocument3)theHTMLDocument;
//compatibility mode affects how height is computed
if ((doc3.documentElement != null) && (!doc5.compatMode.Equals("BackCompat")))
{
return(doc3.documentElement.getAttribute(theAttributeName, 0));
}
else
{
return(theHTMLDocument.body.getAttribute(theAttributeName, 0));
}
}
//************************************************************
//* Functions of get new donators *
//************************************************************
private string GetChatContent()
{
string HTMLContent = null;
try
{
IntPtr Message = RegisterWindowMessage("WM_HTML_GETOBJECT");
IntPtr MessageTimeOut = SendMessageTimeout(PotPlayer.MainChatRoom, Message);
IHTMLDocument5 HTMLDocument = GetHTMLDocument(MessageTimeOut);
HTMLContent = ((HTMLDocumentClass)HTMLDocument).body.innerHTML;
}
catch (Exception e)
{
File.WriteAllText(@".\GetChatContentError.txt", e.ToString());
}
return(HTMLContent);
}
public IHTMLDocument5 GetHTMLDocument(IntPtr ChatRoomHandle)
{
IHTMLDocument5 Result = null;
try
{
Result = (IHTMLDocument5)ObjectFromLresult(ChatRoomHandle, typeof(IHTMLDocument5).GUID, IntPtr.Zero);
}
catch (COMException)
{
}
catch (InvalidCastException)
{
return(null);
}
return(Result);
}
private List <IntPtr> Get_PotPlayer_ChatRooms(List <IntPtr> IESs)
{
List <IntPtr> Result = new List <IntPtr>();
IntPtr Message = IntPtr.Zero;
IntPtr MessageTimeOut = IntPtr.Zero;
IHTMLDocument5 HTMLDocument = null;
string HTMLContent = null;
foreach (IntPtr IES in IESs)
{
Message = RegisterWindowMessage("WM_HTML_GETOBJECT");
MessageTimeOut = SendMessageTimeout(IES, Message);
HTMLDocument = GetHTMLDocument(MessageTimeOut);
if (HTMLDocument != null && HTMLDocument.GetType().Name == "HTMLDocumentClass" && ((HTMLDocumentClass)HTMLDocument).url.Contains("blank"))
{
HTMLContent = ((HTMLDocumentClass)HTMLDocument).body.innerHTML;
if (HTMLContent != null && HTMLContent.Contains("wrap_chat"))
{
Result.Add(IES);
}
}
}
return(Result);
}
/// <summary>
/// Private helper method for the constructors
/// </summary>
/// <param name="document2">IHTMLDocument2</param>
/// <param name="contentWindow">WindowDetails</param>
private void Init(IHTMLDocument2 document2, WindowDetails contentWindow)
{
this.document2 = document2;
this.contentWindow = contentWindow;
this.document3 = document2 as IHTMLDocument3;
// Check what access method is needed for the document
IHTMLDocument5 document5 = (IHTMLDocument5)document2;
//compatibility mode affects how height is computed
isDTD = false;
try {
if ((document3.documentElement != null) && (!document5.compatMode.Equals("BackCompat")))
{
isDTD = true;
}
} catch (Exception ex) {
LOG.Error("Error checking the compatibility mode:");
LOG.Error(ex);
}
Rectangle clientRectangle = contentWindow.WindowRectangle;
try {
IHTMLWindow2 window2 = (IHTMLWindow2)document2.parentWindow;
//IHTMLWindow3 window3 = (IHTMLWindow3)document2.parentWindow;
IHTMLScreen2 screen2 = (IHTMLScreen2)window2.screen;
IHTMLScreen screen = window2.screen;
if (parent != null)
{
// Copy parent values
zoomLevelX = parent.zoomLevelX;
zoomLevelY = parent.zoomLevelY;
viewportRectangle = parent.viewportRectangle;
}
else
{
//DisableScrollbars(document2);
// Calculate zoom level
zoomLevelX = (double)screen2.deviceXDPI / (double)screen2.logicalXDPI;
zoomLevelY = (double)screen2.deviceYDPI / (double)screen2.logicalYDPI;
// Calculate the viewport rectangle, needed if there is a frame around the html window
LOG.DebugFormat("Screen {0}x{1}", ScaleX(screen.width), ScaleY(screen.height));
//LOG.DebugFormat("Screen location {0},{1}", window3.screenLeft, window3.screenTop);
LOG.DebugFormat("Window rectangle {0}", clientRectangle);
LOG.DebugFormat("Client size {0}x{1}", ClientWidth, ClientHeight);
int diffX = clientRectangle.Width - ClientWidth;
int diffY = clientRectangle.Height - ClientHeight;
// If there is a border around the inner window, the diff == 4
// If there is a border AND a scrollbar the diff == 20
if ((diffX == 4 || diffX >= 20) && (diffY == 4 || diffY >= 20))
{
Point viewportOffset = new Point(2, 2);
Size viewportSize = new Size(ClientWidth, ClientHeight);
viewportRectangle = new Rectangle(viewportOffset, viewportSize);
LOG.DebugFormat("viewportRect {0}", viewportRectangle);
}
}
LOG.DebugFormat("Zoomlevel {0}, {1}", zoomLevelX, zoomLevelY);
} catch (Exception e) {
LOG.Warn("Can't get certain properties for documents, using default. Due to: ", e);
}
try {
LOG.DebugFormat("Calculated location {0} for {1}", startLocation, document2.title);
if (name == null)
{
name = document2.title;
}
} catch (Exception e) {
LOG.Warn("Problem while trying to get document title!", e);
}
try {
url = document2.url;
} catch (Exception e) {
LOG.Warn("Problem while trying to get document url!", e);
}
sourceLocation = new Point(ScaleX((int)startLocation.X), ScaleY((int)startLocation.Y));
destinationLocation = new Point(ScaleX((int)startLocation.X), ScaleY((int)startLocation.Y));
if (parent != null)
{
return;
}
try {
IHTMLFramesCollection2 frameCollection = (IHTMLFramesCollection2)document2.frames;
for (int frame = 0; frame < frameCollection.length; frame++)
{
try {
IHTMLWindow2 frameWindow = frameCollection.item(frame);
DocumentContainer frameData = new DocumentContainer(frameWindow, contentWindow, this);
// check if frame is hidden
if (!frameData.isHidden)
{
LOG.DebugFormat("Creating DocumentContainer for Frame {0} found in window with rectangle {1}", frameData.name, frameData.SourceRectangle);
frames.Add(frameData);
}
else
{
LOG.DebugFormat("Skipping frame {0}", frameData.Name);
}
} catch (Exception e) {
LOG.Warn("Problem while trying to get information from a frame, skipping the frame!", e);
}
}
} catch (Exception ex) {
LOG.Warn("Problem while trying to get the frames, skipping!", ex);
}
try {
// Correct iframe locations
foreach (IHTMLElement frameElement in document3.getElementsByTagName("IFRAME"))
{
try {
CorrectFrameLocations(frameElement);
} catch (Exception e) {
LOG.Warn("Problem while trying to get information from an iframe, skipping the frame!", e);
}
}
} catch (Exception ex) {
LOG.Warn("Problem while trying to get the iframes, skipping!", ex);
}
}
/// <summary>
/// Public constructor.
/// </summary>
/// <param name="html">Html string to be processed.</param>
public HtmlProcessor(string html)
{
if (string.IsNullOrEmpty(html))
{
throw new ArgumentException("HTML string can not be null or empty.", "html");
}
this.documentSource = html;
this.doc2 = new HTMLDocumentClass();
this.doc2.write(new object[] { html });
//while (this.doc2.readyState != "complete" && this.doc2.readyState != "interactive")
//{
// System.Windows.Forms.Application.DoEvents();
//}
this.doc3 = (IHTMLDocument3)this.doc2;
this.doc4 = (IHTMLDocument4)this.doc2;
this.doc5 = (IHTMLDocument5)this.doc2;
this.InnerText = (this.doc2.body == null) ? null : this.doc2.body.innerText;
this.Title = this.doc2.title;
this._elements = new List <DomElement>();
IHTMLElementCollection collection = this.doc2.all;
foreach (var item in collection)
{
var element = (IHTMLElement)item;
if (element != null)
{
DomElement domElement = new DomElement
{
Class = element.className,
InnerHtml = element.innerHTML.TrimSafe(),
OuterHtml = element.outerHTML.TrimSafe(),
InnerText = element.innerText.TrimSafe(),
Id = element.id,
TagName = element.tagName.ToLowerInvariant()
};
domElement.Attributes = new NameValueCollection();
if (element.outerHTML.IsNullOrEmpty())
{
continue;
}
//@"<INPUT onblur=google&&google.fade&&google.fade() class=lst title='Google Search' value=TESTING maxLength=2048 size=55 name=q autocomplete='off' init='true'/>";
string node = element.outerHTML.Substring(0, element.outerHTML.IndexOf(">") + 1);
string pattern = @"
(?:<)(?<Tag>[^\s/>]+) # Extract the tag name.
(?![/>]) # Stop if /> is found
# -- Extract Attributes Key Value Pairs --
((?:\s+) # One to many spaces start the attribute
(?<Key>[^=]+) # Name/key of the attribute
(?:=) # Equals sign needs to be matched, but not captured.
(?([\x22\x27]) # If quotes are found
(?:[\x22\x27])
(?<Value>[^\x22\x27]+) # Place the value into named Capture
(?:[\x22\x27])
| # Else no quotes
(?<Value>[^\s/>]*) # Place the value into named Capture
)
)+ # -- One to many attributes found!";
//http://omegacoder.com/?p=512
var attr = (from Match mt in Regex.Matches(node, pattern, RegexOptions.IgnorePatternWhitespace)
select new
{
Name = mt.Groups["Tag"],
Attrs = (from cpKey in mt.Groups["Key"].Captures.Cast <Capture>().Select((a, i) => new { a.Value, i })
join cpValue in mt.Groups["Value"].Captures.Cast <Capture>().Select((b, i) => new { b.Value, i }) on cpKey.i equals cpValue.i
select new KeyValuePair <string, string>(cpKey.Value, cpValue.Value)).ToDictionary(kvp => kvp.Key, kvp => kvp.Value)
}).FirstOrDefault();
if (attr != null)
{
var attributes = attr.Attrs;
foreach (KeyValuePair <string, string> kvp in attributes)
{
if (!string.IsNullOrEmpty(kvp.Value))
{
domElement.Attributes.Add(kvp.Key.ToLowerInvariant(), kvp.Value);
}
}
}
this._elements.Add(domElement);
}
}
}
