// Use this after attempting to find the excerpt element because it destroys the HTML document
private string GetSeparatorExcerpt(IHtmlDocument htmlDocument)
{
if (_separators?.Length > 0)
{
ITreeWalker walker = htmlDocument.CreateTreeWalker(htmlDocument.DocumentElement, FilterSettings.Comment);
IComment comment = (IComment)walker.ToFirst();
while (comment != null && !_separators.Contains(comment.NodeValue.Trim(), StringComparer.OrdinalIgnoreCase))
{
comment = (IComment)walker.ToNext();
}
// Found the first separator
if (comment != null)
{
// Get a clone of the parent element
IElement parent = comment.ParentElement;
if (parent.TagName.Equals("p", StringComparison.OrdinalIgnoreCase))
{
// If we were in a tag inside a paragraph, ascend to the paragraph's parent
parent = parent.ParentElement;
}
// Now remove everything after the separator
walker = htmlDocument.CreateTreeWalker(parent);
bool remove = false;
Stack <INode> removeStack = new Stack <INode>();
INode node = walker.ToFirst();
while (node != null)
{
if (node == comment)
{
remove = true;
}
// Also remove if it's a top-level element that doesn't match the query selector
if (remove ||
(node.Parent == parent &&
node is IElement &&
!string.IsNullOrEmpty(_querySelector) &&
!((IElement)node).Matches(_querySelector)))
{
removeStack.Push(node);
}
node = walker.ToNext();
}
while (removeStack.Count > 0)
{
node = removeStack.Pop();
node.Parent.RemoveChild(node);
}
return(parent.InnerHtml);
}
}
return(null);
}
