"""Tests for HTMLParser.py.""" import html.parser import pprint import unittest from test import support class EventCollector(html.parser.HTMLParser): def __init__(self, *args, **kw): self.events = [] self.append = self.events.append html.parser.HTMLParser.__init__(self, *args, **kw) def get_events(self): # Normalize the list of events so that buffer artefacts don't # separate runs of contiguous characters. L = [] prevtype = None for event in self.events: type = event[0] if type == prevtype == "data": L[-1] = ("data", L[-1][1] + event[1]) else: L.append(event) prevtype = type self.events = L return L # structure markup def handle_starttag(self, tag, attrs): self.append(("starttag", tag, attrs)) def handle_startendtag(self, tag, attrs): self.append(("startendtag", tag, attrs)) def handle_endtag(self, tag): self.append(("endtag", tag)) # all other markup def handle_comment(self, data): self.append(("comment", data)) def handle_charref(self, data): self.append(("charref", data)) def handle_data(self, data): self.append(("data", data)) def handle_decl(self, data): self.append(("decl", data)) def handle_entityref(self, data): self.append(("entityref", data)) def handle_pi(self, data): self.append(("pi", data)) def unknown_decl(self, decl): self.append(("unknown decl", decl)) class EventCollectorExtra(EventCollector): def handle_starttag(self, tag, attrs): EventCollector.handle_starttag(self, tag, attrs) self.append(("starttag_text", self.get_starttag_text())) class TestCaseBase(unittest.TestCase): def get_collector(self): raise NotImplementedError def _run_check(self, source, expected_events, collector=None): if collector is None: collector = self.get_collector() parser = collector for s in source: parser.feed(s) parser.close() events = parser.get_events() if events != expected_events: self.fail("received events did not match expected events\n" "Expected:\n" + pprint.pformat(expected_events) + "\nReceived:\n" + pprint.pformat(events)) def _run_check_extra(self, source, events): self._run_check(source, events, EventCollectorExtra()) def _parse_error(self, source): def parse(source=source): parser = html.parser.HTMLParser() parser.feed(source) parser.close() self.assertRaises(html.parser.HTMLParseError, parse) class HTMLParserStrictTestCase(TestCaseBase): def get_collector(self): return EventCollector(strict=True) def test_processing_instruction_only(self): self._run_check("", [ ("pi", "processing instruction"), ]) self._run_check("", [ ("pi", "processing instruction ?"), ]) def test_simple_html(self): self._run_check(""" &entity;

sample text “ """, [ ("data", "\n"), ("decl", "DOCTYPE html PUBLIC 'foo'"), ("data", "\n"), ("starttag", "html", []), ("entityref", "entity"), ("charref", "32"), ("data", "\n"), ("comment", "comment1a\n-><&#bad;

", [ ("starttag", "p", []), ("startendtag", "img", [("src", "foo")]), ("endtag", "p"), ]) def test_get_starttag_text(self): s = """""" self._run_check_extra(s, [ ("starttag", "foo:bar", [("one", "1"), ("two", "2")]), ("starttag_text", s)]) def test_cdata_content(self): contents = [ ' ¬-an-entity-ref;', "", '

', 'foo = "";', 'foo = "";', 'foo = <\n/script> ', '', ('\n//<\\/s\'+\'cript>\');\n//]]>'), '\n\n', 'foo = "";', '', # these two should be invalid according to the HTML 5 spec, # section 8.1.2.2 #'foo = ', #'foo = ', ] elements = ['script', 'style', 'SCRIPT', 'STYLE', 'Script', 'Style'] for content in contents: for element in elements: element_lower = element.lower() s = '<{element}>{content}'.format(element=element, content=content) self._run_check(s, [("starttag", element_lower, []), ("data", content), ("endtag", element_lower)]) def test_entityrefs_in_attributes(self): self._run_check("", [("starttag", "html", [("foo", "\u20AC&aa&unsupported;")])]) class HTMLParserTolerantTestCase(HTMLParserStrictTestCase): def get_collector(self): return EventCollector(strict=False) def test_tolerant_parsing(self): self._run_check('te>>xt&a<\n' '" "" "
" "- software-and-i" "- library
") expected = [ ('starttag', 'html', []), ('starttag', 'body', [('bgcolor', 'd0ca90'), ('text', '181008')]), ('starttag', 'table', [('cellspacing', '0'), ('cellpadding', '1'), ('width', '100%')]), ('starttag', 'tr', []), ('starttag', 'td', [('align', 'left')]), ('starttag', 'font', [('size', '-1')]), ('data', '- '), ('starttag', 'a', [('href', '/rabota/')]), ('starttag', 'span', [('class', 'en')]), ('data', ' software-and-i'), ('endtag', 'span'), ('endtag', 'a'), ('data', '- '), ('starttag', 'a', [('href', '/1/')]), ('starttag', 'span', [('class', 'en')]), ('data', ' library'), ('endtag', 'span'), ('endtag', 'a'), ('endtag', 'table') ] self._run_check(html, expected) def test_comma_between_attributes(self): self._run_check('

', [ ('starttag', 'form', [('action', '/xxx.php?a=1&b=2&'), ('method', 'post')])]) def test_weird_chars_in_unquoted_attribute_values(self): self._run_check('', [ ('starttag', 'form', [('action', 'bogus|&#()value')])]) def test_correct_detection_of_start_tags(self): # see #13273 html = ('

The rain ' '
in Spain

') 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) html = '

The rain' expected = [ ('starttag', 'div', [('style', ''), ('foo', 'bar')]), ('starttag', 'b', []), ('data', 'The '), ('starttag', 'a', [('href', 'some_url')]), ('data', 'rain'), ('endtag', 'a'), ] self._run_check(html, expected) def test_unescape_function(self): p = html.parser.HTMLParser() self.assertEqual(p.unescape('&#bad;'),'&#bad;') self.assertEqual(p.unescape('&'),'&') # see #12888 self.assertEqual(p.unescape('{ ' * 1050), '{ ' * 1050) def test_main(): support.run_unittest(HTMLParserStrictTestCase, HTMLParserTolerantTestCase) if __name__ == "__main__": test_main()