Mercurial > cpython-lru_cache-weakref
diff Lib/test/test_htmlparser.py @ 73159:41d41776aa6d 3.2
#13273: fix a bug that prevented HTMLParser to properly detect some tags when strict=False.
| author | Ezio Melotti <ezio.melotti@gmail.com> |
|---|---|
| date | Fri, 28 Oct 2011 13:21:09 +0300 |
| parents | 9896fc2a8167 |
| children | a6f2244b251f |
line wrap: on
line diff
--- a/Lib/test/test_htmlparser.py Fri Oct 28 12:32:53 2011 +0300 +++ b/Lib/test/test_htmlparser.py Fri Oct 28 13:21:09 2011 +0300 @@ -373,6 +373,39 @@ [('action', 'bogus|&#()value')])], collector = self.collector) + def test_issue13273(self): + html = ('<div style="" ><b>The <a href="some_url">rain</a> ' + '<br /> in <span>Spain</span></b></div>') + expected = [ + ('starttag', 'div', [('style', '')]), + ('starttag', 'b', []), + ('data', 'The '), + ('starttag', 'a', [('href', 'some_url')]), + ('data', 'rain'), + ('endtag', 'a'), + ('data', ' '), + ('startendtag', 'br', []), + ('data', ' in '), + ('starttag', 'span', []), + ('data', 'Spain'), + ('endtag', 'span'), + ('endtag', 'b'), + ('endtag', 'div') + ] + self._run_check(html, expected, collector=self.collector) + + def test_issue13273_2(self): + html = '<div style="", foo = "bar" ><b>The <a href="some_url">rain</a>' + expected = [ + ('starttag', 'div', [('style', ''), ('foo', 'bar')]), + ('starttag', 'b', []), + ('data', 'The '), + ('starttag', 'a', [('href', 'some_url')]), + ('data', 'rain'), + ('endtag', 'a'), + ] + self._run_check(html, expected, collector=self.collector) + def test_unescape_function(self): p = html.parser.HTMLParser() self.assertEqual(p.unescape('&#bad;'),'&#bad;')
