public void TreeMisnestedTagsHeisenbergWithFurthest()
{
//8.2.8.2 Misnested tags: <b><p></b></p>
var doc = DocumentBuilder.Html(@"<b>1<p>2</b>3</p>");
var tree = new HTMLHtmlElement()
.AppendChild(new HTMLHeadElement()).ParentNode
.AppendChild(new HTMLBodyElement())
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new TextNode("1")).ParentNode.ParentNode
.AppendChild(new HTMLParagraphElement())
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new TextNode("2")).ParentNode.ParentNode
.AppendChild(new TextNode("3")).ParentNode.ParentNode.ParentNode;
Assert.AreEqual(tree.ToHtml(), doc.DocumentElement.ToHtml());
}
public void TreeNonConformingTable()
{
//8.2.5.4.7 The "in body" insertion mode - "In the non-conforming ..."
var doc = DocumentBuilder.Html(@"<a href=""a"">a<table><a href=""b"">b</table>x");
var tree = new HTMLHtmlElement()
.AppendChild(new HTMLHeadElement()).ParentNode
.AppendChild(new HTMLBodyElement())
.AppendChild(new HTMLAnchorElement().SetAttribute("href", "a"))
.AppendChild(new TextNode("a")).ParentNode
.AppendChild(new HTMLAnchorElement().SetAttribute("href", "b"))
.AppendChild(new TextNode("b")).ParentNode.ParentNode
.AppendChild(new HTMLTableElement()).ParentNode.ParentNode
.AppendChild(new HTMLAnchorElement().SetAttribute("href", "b"))
.AppendChild(new TextNode("x")).ParentNode.ParentNode.ParentNode;
Assert.AreEqual(tree.ToHtml(), doc.DocumentElement.ToHtml());
}
public void TreeOneTextNodeTableBeforeABCD()
{
//One Text node before the table, containing "ABCD"
var doc = DocumentBuilder.Html(@"A<table>B<tr>C</tr>D</table>");
var tree = new HTMLHtmlElement()
.AppendChild(new HTMLHeadElement()).ParentNode
.AppendChild(new HTMLBodyElement())
.AppendChild(new TextNode("ABCD")).ParentNode
.AppendChild(new HTMLTableElement())
.AppendChild(new HTMLTableSectionElement())
.AppendChild(new HTMLTableRowElement()).ParentNode.ParentNode.ParentNode.ParentNode;
Assert.AreEqual(tree.ToHtml(), doc.DocumentElement.ToHtml());
}
public void TreeUnexpectedTableMarkup()
{
//8.2.8.3 Unexpected markup in tables
var doc = DocumentBuilder.Html(@"<table><b><tr><td>aaa</td></tr>bbb</table>ccc");
var tree = new HTMLHtmlElement()
.AppendChild(new HTMLHeadElement()).ParentNode
.AppendChild(new HTMLBodyElement())
.AppendChild(new HTMLElement() { NodeName = "b" }).ParentNode
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new TextNode("bbb")).ParentNode.ParentNode
.AppendChild(new HTMLTableElement())
.AppendChild(new HTMLTableSectionElement())
.AppendChild(new HTMLTableRowElement())
.AppendChild(new HTMLTableCellElement())
.AppendChild(new TextNode("aaa")).ParentNode.ParentNode.ParentNode.ParentNode.ParentNode
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new TextNode("ccc")).ParentNode.ParentNode.ParentNode;
Assert.AreEqual(tree.ToHtml(), doc.DocumentElement.ToHtml());
}
public void TreeOneTextNodeTableBeforeAspaceBC()
{
//One Text node before the table, containing "A BC" (A-space-B-C), and one Text node inside the table (as a child of a tbody) with a single space character.
var doc = DocumentBuilder.Html(@"A<table><tr> B</tr> </em>C</table>");
var tree = new HTMLHtmlElement()
.AppendChild(new HTMLHeadElement()).ParentNode
.AppendChild(new HTMLBodyElement())
.AppendChild(new TextNode("A BC")).ParentNode
.AppendChild(new HTMLTableElement())
.AppendChild(new HTMLTableSectionElement())
.AppendChild(new HTMLTableRowElement()).ParentNode
.AppendChild(new TextNode(" ")).ParentNode.ParentNode.ParentNode.ParentNode;
Assert.AreEqual(tree.ToHtml(), doc.DocumentElement.ToHtml());
}
public void TreeUnclosedFormattingElements()
{
//8.2.8.6 Unclosed formatting elements
var doc = DocumentBuilder.Html(@"<!DOCTYPE html>
<p><b class=x><b class=x><b><b class=x><b class=x><b>X<p>X<p><b><b class=x><b>X<p></b></b></b></b></b></b>X");
var tree = new HTMLHtmlElement()
.AppendChild(new HTMLHeadElement()).ParentNode
.AppendChild(new HTMLBodyElement())
.AppendChild(new HTMLParagraphElement())
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new TextNode("X")).ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode
.AppendChild(new HTMLParagraphElement())
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new TextNode("X")).ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode
.AppendChild(new HTMLParagraphElement())
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new HTMLElement() { NodeName = "b" }.SetAttribute("class", "x"))
.AppendChild(new HTMLElement() { NodeName = "b" })
.AppendChild(new TextNode("X")).ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode.ParentNode
.AppendChild(new HTMLParagraphElement())
.AppendChild(new TextNode("X")).ParentNode.ParentNode.ParentNode;
Assert.AreEqual(tree.ToHtml(), doc.DocumentElement.ToHtml());
}
public void HtmlHasRightHeadElement()
{
var doc = new HTMLDocument();
var root = new HTMLHtmlElement();
doc.AppendChild(root);
var head = new HTMLHeadElement();
root.AppendChild(head);
Assert.AreEqual(head, doc.Head);
}
public void HtmlHasRightBodyElement()
{
var doc = new HTMLDocument();
var root = new HTMLHtmlElement();
doc.AppendChild(root);
var body = new HTMLBodyElement();
root.AppendChild(body);
Assert.AreEqual(body, doc.Body);
}
/// <summary>
/// Adds the root element (html) to the document.
/// </summary>
/// <param name="token">The token which started this process.</param>
void AddRoot(HtmlToken token)
{
var element = new HTMLHtmlElement();
doc.AppendChild(element);
SetupElement(element, token, false);
open.Add(element);
tokenizer.AcceptsCharacterData = !element.IsInHtml;
}
/// <summary>
/// Switches to the fragment algorithm with the specified context element.
/// </summary>
/// <param name="context">The context element where the algorithm is applied to.</param>
public void SwitchToFragment(Node context)
{
if (started)
throw new InvalidOperationException("Fragment mode has to be activated before running the parser!");
switch (context.NodeName)
{
case HTMLTitleElement.Tag:
case HTMLTextAreaElement.Tag:
{
tokenizer.Switch(HtmlParseMode.RCData);
break;
}
case HTMLStyleElement.Tag:
case HTMLSemanticElement.XmpTag:
case HTMLIFrameElement.Tag:
case HTMLNoElement.NoEmbedTag:
case HTMLNoElement.NoFramesTag:
{
tokenizer.Switch(HtmlParseMode.Rawtext);
break;
}
case HTMLScriptElement.Tag:
{
tokenizer.Switch(HtmlParseMode.Script);
break;
}
case HTMLNoElement.NoScriptTag:
{
if (doc.IsScripting)
tokenizer.Switch(HtmlParseMode.Rawtext);
break;
}
case HTMLSemanticElement.PlaintextTag:
{
tokenizer.Switch(HtmlParseMode.Plaintext);
break;
}
}
var root = new HTMLHtmlElement();
doc.AppendChild(root);
open.Add(root);
Reset(context);
fragmentContext = context;
tokenizer.AcceptsCharacterData = !AdjustedCurrentNode.IsInHtml;
do
{
if (context is HTMLFormElement)
{
form = (HTMLFormElement)context;
break;
}
context = context.ParentNode;
}
while (context != null);
}
/// <summary>
/// See 8.2.5.4.2 The "before html" insertion mode.
/// </summary>
/// <param name="token">The passed token.</param>
void BeforeHtml(HtmlToken token)
{
if (token.Type == HtmlTokenType.DOCTYPE)
{
RaiseErrorOccurred(ErrorCode.DoctypeTagInappropriate);
}
else if (token.Type == HtmlTokenType.Comment)
{
AddComment(doc, token);
}
else if (token.Type == HtmlTokenType.StartTag && ((HtmlTagToken)token).Name == HTMLHtmlElement.Tag)
{
var element = new HTMLHtmlElement();
AddElementToDocument(element, token);
//TODO
//If the Document is being loaded as part of navigation of a browsing context, then:
// if the newly created element has a manifest attribute whose value is not the empty string,
// then resolve the value of that attribute to an absolute URL, relative to the newly created element,
// and if that is successful, run the application cache selection algorithm with the result of applying
// the URL serializer algorithm to the resulting parsed URL with the exclude fragment flag set;
// otherwise, if there is no such attribute, or its value is the empty string, or resolving its value fails,
// run the application cache selection algorithm with no manifest. The algorithm must be passed the Document object.
insert = HtmlTreeMode.BeforeHead;
}
else if (token.Type == HtmlTokenType.EndTag && !(((HtmlTagToken)token).Name.IsHtmlBodyOrBreakRowElement(true)))
{
RaiseErrorOccurred(ErrorCode.TagCannotEndHere);
}
else if(!token.IsIgnorable)
{
var element = new HTMLHtmlElement();
AddElementToDocument(element, HtmlToken.OpenTag(HTMLHtmlElement.Tag));
//TODO
//If the Document is being loaded as part of navigation of a browsing context, then:
// run the application cache selection algorithm with no manifest, passing it the Document object.
insert = HtmlTreeMode.BeforeHead;
BeforeHead(token);
}
}
