public string CleanHtml(IDataMap map, string itemPath, string html)
{
if (String.IsNullOrEmpty(html))
{
return(html);
}
var document = new HtmlDocument();
document.LoadHtml(html);
HtmlNodeCollection tryGetNodes = document.DocumentNode.SelectNodes("./*|./text()");
if (tryGetNodes == null || !tryGetNodes.Any())
{
return(html);
}
var nodes = new Queue <HtmlNode>(tryGetNodes);
int i = 0;
while (nodes.Count > 0)
{
var node = nodes.Dequeue();
var nodeName = node.Name.ToLower();
var parentNode = node.ParentNode;
var childNodes = node.SelectNodes("./*|./text()");
if (childNodes != null)
{
foreach (var child in childNodes)
{
nodes.Enqueue(child);
}
}
if (UnwantedTags.Any(tag => tag == nodeName))
{
// if this node is one to remove
if (childNodes != null)
{
// make sure children are added back
foreach (var child in childNodes)
{
parentNode.InsertBefore(child, node);
}
}
parentNode.RemoveChild(node);
}
else if (node.HasAttributes)
{
// if it's not being removed
foreach (string s in UnwantedAttributes) // remove unwanted attributes
{
node.Attributes.Remove(s);
}
//replace images
if (nodeName.Equals("img"))
{
// see if it exists
string imgSrc = node.Attributes["src"].Value;
MediaItem newImg = HandleImage(map, MediaParentItem, itemPath, imgSrc);
if (newImg != null)
{
string newSrc = string.Format("-/media/{0}.ashx", newImg.ID.ToShortID().ToString());
// replace the node with sitecore tag
node.SetAttributeValue("src", newSrc);
}
}
}
i++;
}
return(document.DocumentNode.InnerHtml);
}
private void HandleNextNode(Queue <HtmlNode> nodes, IDataMap map, string itemPath)
{
var node = nodes.Dequeue();
var nodeName = node.Name.ToLower();
var parentNode = node.ParentNode;
var childNodes = node.SelectNodes("./*|./text()");
if (childNodes != null)
{
foreach (var child in childNodes)
{
nodes.Enqueue(child);
}
}
if (UnwantedTags.Any(tag => tag == nodeName))
{
// if this node is one to remove
if (childNodes != null)
{
// make sure children are added back
foreach (var child in childNodes)
{
parentNode.InsertBefore(child, node);
}
}
parentNode.RemoveChild(node);
}
else if (node.HasAttributes)
{
// if it's not being removed
foreach (string s in UnwantedAttributes) // remove unwanted attributes
{
node.Attributes.Remove(s);
}
//replace images
if (nodeName.Equals("img") || nodeName.Equals("script"))
{
// see if it exists
string imgSrc = node.Attributes.Contains("src") ? node.Attributes["src"].Value : string.Empty;
if (!string.IsNullOrEmpty(imgSrc))
{
MediaItem newImg = HandleMedia(map, itemPath, imgSrc);
if (newImg != null)
{
string newSrc = string.Format("-/media/{0}.ashx", newImg.ID.ToShortID());
// replace the node with sitecore tag
node.SetAttributeValue("src", newSrc);
}
}
}
else if (nodeName.Equals("a") || nodeName.Equals("link"))
{
if (nodeName.Equals("a") && node.Attributes.Contains("target"))
{
string target = node.Attributes.Contains("target") ? node.Attributes["target"].Value : string.Empty;
if (target.Equals("_blank", StringComparison.InvariantCultureIgnoreCase))
{
node.SetAttributeValue("rel", "noopener noreferrer");
}
}
// see if it exists
string linkHref = node.Attributes.Contains("href") ? node.Attributes["href"].Value : string.Empty;
if (!string.IsNullOrEmpty(linkHref))
{
MediaItem newImg = HandleMedia(map, itemPath, linkHref);
if (newImg != null)
{
string newHref = string.Format("-/media/{0}.ashx", newImg.ID.ToShortID());
// replace the node with sitecore tag
node.SetAttributeValue("href", newHref);
}
}
}
}
else
{
//Keep node as is
}
}
private void HandleNextNode(Queue <HtmlNode> nodes, IDataMap map, string itemPath, Item importRow)
{
var node = nodes.Dequeue();
var nodeName = node.Name.ToLower();
var parentNode = node.ParentNode;
var childNodes = node.SelectNodes("./*|./text()");
if (childNodes != null)
{
foreach (var child in childNodes)
{
nodes.Enqueue(child);
}
}
if (UnwantedTags.Any(tag => tag == nodeName))
{
// if this node is one to remove
if (childNodes != null)
{
// make sure children are added back
foreach (var child in childNodes)
{
parentNode.InsertBefore(child, node);
}
}
parentNode.RemoveChild(node);
}
else if (node.HasAttributes)
{
// if it's not being removed
foreach (string s in UnwantedAttributes) // remove unwanted attributes
{
node.Attributes.Remove(s);
}
//replace images
if (nodeName.Equals("img"))
{
// see if it exists
string imgSrc = node.Attributes["src"].Value;
DynamicLink dynamicLink;
if (!DynamicLink.TryParse(imgSrc, out dynamicLink))
{
return;
}
MediaItem mediaItem = importRow.Database.GetItem(dynamicLink.ItemId, dynamicLink.Language ?? map.ImportToLanguage);
var mediaParentItem = BuildMediaPath(map.ToDB, mediaItem.InnerItem.Paths.ParentPath);
MediaItem newImg = HandleMediaItem(map, mediaParentItem, itemPath, mediaItem);
if (newImg != null)
{
string newSrc = string.Format("-/media/{0}.ashx", newImg.ID.ToShortID().ToString());
// replace the node with sitecore tag
node.SetAttributeValue("src", newSrc);
}
}
}
else if (ReplaceTags.ContainsKey(nodeName))
{
// Replace tag
node.Name = ReplaceTags[nodeName];
}
else
{
//Keep node as is
}
}