public void testHeaderContents()
{
// h* tags (h1 .. h9) in browsers can handle any internal content other than other h*. which is not per any
// spec, which defines them as containing phrasing content only. so, reality over theory.
Document doc = NSoupClient.Parse("<h1>Hello <div>There</div> now</h1> <h2>More <h3>Content</h3></h2>");
Assert.AreEqual("<h1>Hello <div>There</div> now</h1> <h2>More </h2><h3>Content</h3>", TextUtil.StripNewLines(doc.Body.Html()));
}
public void handlesMisnestedTagsBP()
{
// whatwg: <b><p></b></p>
string h = "<b>1<p>2</b>3</p>";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("<b>1</b>\n<p><b>2</b>3</p>", doc.Body.Html());
}
public void findsCharsetInMalformedMeta()
{
string h = "<meta http-equiv=Content-Type content=text/html; charset=gb2312>";
// example cited for reason of html5's <meta charset> element
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("gb2312", doc.Select("meta").Attr("charset"));
}
public void normalisesDocument()
{
string h = "<!doctype html>One<html>Two<head>Three<link></head>Four<body>Five </body>Six </html>Seven ";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("<!DOCTYPE html><html><head></head><body>OneTwoThree<link />FourFive Six Seven </body></html>",
TextUtil.StripNewLines(doc.Html()));
}
public void ignoresContentAfterFrameset()
{
string h = "<html><head><title>One</title></head><frameset><frame /><frame /></frameset><table></table></html>";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("<html><head><title>One</title></head><frameset><frame /><frame /></frameset></html>", TextUtil.StripNewLines(doc.Html()));
// no body, no table. No crash!
}
public void reconstructFormattingElements()
{
// tests attributes and multi b
string h = "<p><b class=one>One <i>Two <b>Three</p><p>Hello</p>";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("<p><b class=\"one\">One <i>Two <b>Three</b></i></b></p>\n<p><b class=\"one\"><i><b>Hello</b></i></b></p>", doc.Body.Html());
}
public void dropsUnterminatedAttribute()
{
// NSoupClient used to parse this to <p id="foo">, but whatwg, webkit will drop.
string h1 = "<p id=\"foo";
Document doc = NSoupClient.Parse(h1);
Assert.AreEqual("", doc.Text());
}
public void handlesSolidusAtAttributeEnd()
{
// this test makes sure [<a href=/>link</a>] is parsed as [<a href="/">link</a>], not [<a href="" /><a>link</a>]
string h = "<a href=/>link</a>";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("<a href=\"/\">link</a>", doc.Body.Html());
}
public void handlesMultiClosingBody()
{
string h = "<body><p>Hello</body><p>there</p></body></body></html><p>now";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual(3, doc.Select("p").Count);
Assert.AreEqual(3, doc.Body.Children.Count);
}
public void handlesUnknownNamespaceTags()
{
// note that the first foo:bar should not really be allowed to be self closing, if parsed in html mode.
string h = "<foo:bar id='1' /><abc:def id=2>Foo<p>Hello</p></abc:def><foo:bar>There</foo:bar>";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("<foo:bar id=\"1\" /><abc:def id=\"2\">Foo<p>Hello</p></abc:def><foo:bar>There</foo:bar>", TextUtil.StripNewLines(doc.Body.Html()));
}
public void traverse()
{
Document doc = NSoupClient.Parse("<div><p>Hello</p></div><div>There</div>");
StringBuilder accum = new StringBuilder();
doc.Select("div").First.Traverse(new TestNodeVisitor(accum));
Assert.AreEqual("<div><p><#text></#text></p></div>", accum.ToString());
}
public void handlesTextArea()
{
Document doc = NSoupClient.Parse("<textarea>Hello</textarea>");
Elements els = doc.Select("textarea");
Assert.AreEqual("Hello", els.Text);
Assert.AreEqual("Hello", els.Val());
}
public void handlesMisnestedTagsBI()
{
// whatwg: <b><i></b></i>
string h = "<p>1<b>2<i>3</b>4</i>5</p>";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("<p>1<b>2<i>3</i></b><i>4</i>5</p>", doc.Body.Html());
// adoption agency on </b>, reconstruction of formatters on 4.
}
public void handlesBaseWithoutHref()
{
string h = "<head><base target='_blank'></head><body><a href=/foo>Test</a></body>";
Document doc = NSoupClient.Parse(h, "http://example.com/");
Element a = doc.Select("a").First;
Assert.AreEqual("/foo", a.Attr("href"));
Assert.AreEqual("http://example.com/foo", a.Attr("abs:href"));
}
public void handlesFrames()
{
string h = "<html><head><script></script><noscript></noscript></head><frameset><frame src=foo></frame><frame src=foo></frameset></html>";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("<html><head><script></script><noscript></noscript></head><frameset><frame src=\"foo\" /><frame src=\"foo\" /></frameset></html>",
TextUtil.StripNewLines(doc.Html()));
// no body auto vivification
}
public void discardsNakedTds()
{
// NSoupClient used to make this into an implicit table; but browsers make it into a text run
string h = "<td>Hello<td><p>There<p>now";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("Hello<p>There</p><p>now</p>", TextUtil.StripNewLines(doc.Body.Html()));
// <tbody> is introduced if no implicitly creating table, but allows tr to be directly under table
}
public void testClonesDeclarations()
{
Document doc = NSoupClient.Parse("<!DOCTYPE html><html><head><title>Doctype test");
Document clone = (Document)doc.Clone();
Assert.AreEqual(doc.Html(), clone.Html());
Assert.AreEqual("<!DOCTYPE html><html><head><title>Doctype test</title></head><body></body></html>",
TextUtil.StripNewLines(clone.Html()));
}
public void handlesWhitespaceInoDocType()
{
string html = "<!DOCTYPE html\r\n" +
" PUBLIC \"-//W3C//DTD XHTML 1.0 Transitional//EN\"\r\n" +
" \"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd\">";
Document doc = NSoupClient.Parse(html);
Assert.AreEqual("<!DOCTYPE html PUBLIC \"-//W3C//DTD XHTML 1.0 Transitional//EN\" \"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd\">", doc.ChildNodes[0].OuterHtml());
}
public void createsStructureFromBodySnippet()
{
// the bar baz stuff naturally goes into the body, but the 'foo' goes into root, and the normalisation routine
// needs to move into the start of the body
string html = "foo <b>bar</b> baz";
Document doc = NSoupClient.Parse(html);
Assert.AreEqual("foo bar baz", doc.Text());
}
public void testSupplyParserToDataStream()
{
using (Stream input = getFile("Test.htmltests.xml-test.xml"))
{
Document doc = NSoupClient.Parse(input, null, "http://foo.com", NSoup.Parse.Parser.XmlParser());
Assert.AreEqual("<doc><val>One<val>Two</val>Three</val></doc>",
TextUtil.StripNewLines(doc.Html()));
}
}
public void relaxedBaseEntityMatchAndStrictExtendedMatch()
{
// extended entities need a ; at the end to match, base does not
string html = "& " ® &icy &hopf и 𝕙";
Document doc = NSoupClient.Parse(html);
doc.OutputSettings().EscapeMode = Entities.EscapeMode.Extended; // modifies output only to clarify test
Assert.AreEqual(StringUtil.NormaliseWhitespace("& " ® &icy &hopf и 𝕙"), doc.Body.Html());
}
public void doesNotFindShortestMatchingEntity()
{
// previous behaviour was to identify a possible entity, then chomp down the string until a match was found.
// (as defined in html5.) However in practise that lead to spurious matches against the author's intent.
string html = "One &clubsuite; ♣";
Document doc = NSoupClient.Parse(html);
Assert.AreEqual(StringUtil.NormaliseWhitespace("One &clubsuite; ג™£"), doc.Body.Html());
}
public void letterDigitEntities()
{
string html = "<p>¹²³¼½¾</p>";
Document doc = NSoupClient.Parse(html);
Element p = doc.Select("p").First;
Assert.AreEqual("¹²³¼½¾", p.Html());
Assert.AreEqual("¹²³¼½¾", p.Text());
}
public void handlesUnexpectedMarkupInTables()
{
// whatwg - tests markers in active formatting (if they didn't work, would get in in table)
// also tests foster parenting
string h = "<table><b><tr><td>aaa</td></tr>bbb</table>ccc";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual("<b></b><b>bbb</b><table><tbody><tr><td>aaa</td></tr></tbody></table><b>ccc</b>", TextUtil.StripNewLines(doc.Body.Html()));
}
public void handlesSpanInTBody()
{
// test for bug 64
string h = "<table><tbody><span class='1'><tr><td>One</td></tr><tr><td>Two</td></tr></span></tbody></table>";
Document doc = NSoupClient.Parse(h);
Assert.AreEqual(doc.Select("span").First.Children.Count, 0); // the span gets closed
Assert.AreEqual(doc.Select("table").Count, 1); // only one table
}
public void handlesEscapedData()
{
string html = "<div title='Surf & Turf'>Reef & Beef</div>";
Document doc = NSoupClient.Parse(html);
Element div = doc.GetElementsByTag("div")[0];
Assert.AreEqual("Surf & Turf", div.Attr("title"));
Assert.AreEqual("Reef & Beef", div.Text());
}
public void wrapDiv()
{
string h = "<p><b>This</b> is <b>jsoup</b>.</p> <p>How do you like it?</p>";
Document doc = NSoupClient.Parse(h);
doc.Select("p").Wrap("<div></div>");
Assert.AreEqual("<div><p><b>This</b> is <b>jsoup</b>.</p></div> <div><p>How do you like it?</p></div>",
TextUtil.StripNewLines(doc.Body.Html()));
}
public String Down(int index)
{
String temp = WebHelper.GetHtmlCodeByWebClientWithGzip(this.Items[index].Url, "utf-8");
Document doc = NSoupClient.Parse(temp);
Element et = doc.Select("a#fn_pc_download").First;
String down = et.Attr("_href");
return(down);
}
public void handlesNewLinesAndReturns()
{
string html = "<a\r\nfoo='bar\r\nqux'\r\nbar\r\n=\r\ntwo>One</a>";
Element el = NSoupClient.Parse(html).Select("a").First;
Assert.AreEqual(2, el.Attributes.Count);
Assert.AreEqual("bar\r\nqux", el.Attr("foo")); // currently preserves newlines in quoted attributes. todo confirm if should.
Assert.AreEqual("two", el.Attr("bar"));
}
public void parsesUnterminatedTextarea()
{
// don't parse right to end, but break on <p>
Document doc = NSoupClient.Parse("<body><p><textarea>one<p>two");
Element t = doc.Select("textarea").First;
Assert.AreEqual("one", t.Text());
Assert.AreEqual("two", doc.Select("p")[1].Text());
}