"""Tests for HTMLParser.py.""" import HTMLParser import pprint import unittest from test import test_support class EventCollector(HTMLParser.HTMLParser): def __init__(self): self.events = [] self.append = self.events.append HTMLParser.HTMLParser.__init__(self) 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 _run_check(self, source, expected_events, collector=EventCollector): 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 = HTMLParser.HTMLParser() parser.feed(source) parser.close() self.assertRaises(HTMLParser.HTMLParseError, parse) class HTMLParserTestCase(TestCaseBase): 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->< %paramEntity; ]""" self._run_check("" % inside, [ ("decl", inside), ]) def test_bad_nesting(self): # Strangely, this *is* supposed to test that overlapping # elements are allowed. HTMLParser is more geared toward # lexing the input that parsing the structure. self._run_check("", [ ("starttag", "a", []), ("starttag", "b", []), ("endtag", "a"), ("endtag", "b"), ]) def test_bare_ampersands(self): self._run_check("this text & contains & ampersands &", [ ("data", "this text & contains & ampersands &"), ]) def test_bare_pointy_brackets(self): self._run_check("this < text > contains < bare>pointy< brackets", [ ("data", "this < text > contains < bare>pointy< brackets"), ]) def test_illegal_declarations(self): self._parse_error('') def test_starttag_end_boundary(self): self._run_check("""""", [("starttag", "a", [("b", "<")])]) self._run_check("""""", [("starttag", "a", [("b", ">")])]) def test_buffer_artefacts(self): output = [("starttag", "a", [("b", "<")])] self._run_check([""], output) self._run_check([""], output) self._run_check([""], output) self._run_check([""], output) self._run_check([""], output) self._run_check([""], output) output = [("starttag", "a", [("b", ">")])] self._run_check([""], output) self._run_check([""], output) self._run_check([""], output) self._run_check(["'>"], output) self._run_check([""], output) self._run_check([""], output) output = [("comment", "abc")] self._run_check(["", ""], output) self._run_check(["<", "!--abc-->"], output) self._run_check([""], output) self._run_check([""], output) self._run_check([""], output) self._run_check([""], output) self._run_check([""], output) self._run_check([""], output) self._run_check(["", ""], output) def test_starttag_junk_chars(self): self._parse_error("") self._parse_error("") self._parse_error("") self._parse_error("") self._parse_error("'") self._parse_error("

", [ ("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 = "";', u'', # 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 = u'<{element}>{content}'.format(element=element, content=content) self._run_check(s, [("starttag", element_lower, []), ("data", content), ("endtag", element_lower)]) def test_malformatted_charref(self): self._run_check("

&#bad;

", [ ("starttag", "p", []), ("data", "&#bad;"), ("endtag", "p"), ]) def test_unescape_function(self): parser = HTMLParser.HTMLParser() self.assertEqual(parser.unescape('&#bad;'),'&#bad;') self.assertEqual(parser.unescape('&'),'&') class AttributesTestCase(TestCaseBase): def test_attr_syntax(self): output = [ ("starttag", "a", [("b", "v"), ("c", "v"), ("d", "v"), ("e", None)]) ] self._run_check("""""", output) self._run_check("""""", output) self._run_check("""""", output) self._run_check("""""", output) def test_attr_values(self): self._run_check("""""", [("starttag", "a", [("b", "xxx\n\txxx"), ("c", "yyy\t\nyyy"), ("d", "\txyz\n")])]) self._run_check("""""", [("starttag", "a", [("b", ""), ("c", "")])]) # Regression test for SF patch #669683. self._run_check("", [("starttag", "e", [("a", "rgb(1,2,3)")])]) # Regression test for SF bug #921657. self._run_check( "", [("starttag", "a", [("href", "mailto:xyz@example.com")])]) def test_attr_nonascii(self): # see issue 7311 self._run_check( u"\u4e2d\u6587", [("starttag", "img", [("src", "/foo/bar.png"), ("alt", u"\u4e2d\u6587")])]) self._run_check( u"", [("starttag", "a", [("title", u"\u30c6\u30b9\u30c8"), ("href", u"\u30c6\u30b9\u30c8.html")])]) self._run_check( u'', [("starttag", "a", [("title", u"\u30c6\u30b9\u30c8"), ("href", u"\u30c6\u30b9\u30c8.html")])]) def test_attr_entity_replacement(self): self._run_check( "", [("starttag", "a", [("b", "&><\"'")])]) def test_attr_funky_names(self): self._run_check( "", [("starttag", "a", [("a.b", "v"), ("c:d", "v"), ("e-f", "v")])]) self._run_check( "", [("starttag", "a", [("$", None)]), ("starttag", "b", [("$", "%")]), ("starttag", "c", [("\\", "/")])]) def test_entityrefs_in_attributes(self): self._run_check( "", [("starttag", "html", [("foo", u"\u20AC&aa&unsupported;")])]) def test_entities_in_attribute_value(self): # see #1200313 for entity in ['&', '&', '&', '&']: self._run_check('' % entity, [("starttag", "a", [("href", "&")])]) self._run_check("" % entity, [("starttag", "a", [("href", "&")])]) self._run_check("" % entity, [("starttag", "a", [("href", "&")])]) def test_malformed_attributes(self): # see #13357 html = ( "test - bad1" "test - bad2" "test - bad3" "test - bad4" ) expected = [ ('starttag', 'a', [('href', "test'style='color:red;bad1'")]), ('data', 'test - bad1'), ('endtag', 'a'), ('starttag', 'a', [('href', "test'+style='color:red;ba2'")]), ('data', 'test - bad2'), ('endtag', 'a'), ('starttag', 'a', [('href', u"test'\xa0style='color:red;bad3'")]), ('data', 'test - bad3'), ('endtag', 'a'), ('starttag', 'a', [('href', u"test'\xa0style='color:red;bad4'")]), ('data', 'test - bad4'), ('endtag', 'a') ] self._run_check(html, expected) def test_malformed_adjacent_attributes(self): # see #12629 self._run_check('', [('starttag', 'x', []), ('startendtag', 'y', [('z', ''), ('o""', None)]), ('endtag', 'x')]) self._run_check('', [('starttag', 'x', []), ('startendtag', 'y', [('z', ''), ('""', None)]), ('endtag', 'x')]) # see #755670 for the following 3 tests def test_adjacent_attributes(self): self._run_check('', [("starttag", "a", [("width", "100%"), ("cellspacing","0")])]) self._run_check('', [("starttag", "a", [("id", "foo"), ("class","bar")])]) def test_missing_attribute_value(self): self._run_check('', [("starttag", "a", [("v", "")])]) def test_javascript_attribute_value(self): self._run_check("", [("starttag", "a", [("href", "javascript:popup('/popup/help.html')")])]) def test_end_tag_in_attribute_value(self): # see #1745761 self._run_check("spam", [("starttag", "a", [("href", "http://www.example.org/\">;")]), ("data", "spam"), ("endtag", "a")]) def test_main(): test_support.run_unittest(HTMLParserTestCase, AttributesTestCase) if __name__ == "__main__": test_main()