From 0c9e690dedf720c7c34cc2433f0ccd03f7eb2a85 Mon Sep 17 00:00:00 2001 From: Leonard Richardson Date: Fri, 18 Feb 2011 14:24:42 -0500 Subject: Ported tests of bad markup that were lying around the TODO. --- tests/test_html5lib.py | 54 ++++++++++++++++++++++++++++++++++++++++++++------ tests/test_lxml.py | 49 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 97 insertions(+), 6 deletions(-) (limited to 'tests') diff --git a/tests/test_html5lib.py b/tests/test_html5lib.py index 5aeac76..3045b02 100644 --- a/tests/test_html5lib.py +++ b/tests/test_html5lib.py @@ -131,14 +131,56 @@ class TestHTML5BuilderInvalidMarkup(TestLXMLBuilderInvalidMarkup): self.assertEquals(comment, 'b Sacr\xe9 bleu!""" - soup = self.soup(isolatin) + def test_document_starts_with_bogus_declaration(self): + soup = self.soup('a') + # 'Foo' becomes a comment that appears before the HTML. + comment = soup.contents[0] + self.assertTrue(isinstance(comment, Comment)) + self.assertEquals(comment, 'Foo') - utf8 = isolatin.replace("ISO-Latin-1".encode(), "utf-8".encode()) - utf8 = utf8.replace("\xe9", "\xc3\xa9") + self.assertEquals(self.find(text="a") == "a") - #print soup + def test_attribute_value_was_closed_by_subsequent_tag(self): + markup = """baz""" + soup = self.soup(markup) + # The string between the first and second quotes was interpreted + # as the value of the 'href' attribute. + self.assertEquals(soup.a['href'], 'foo,

a

') + # The declaration becomes a comment. + comment = soup.contents[0] + self.assertTrue(isinstance(comment, Comment)) + self.assertEquals(comment, ' Foo ') + self.assertEquals(soup.p.string, 'a') + + def test_document_ends_with_incomplete_declaration(self): + soup = self.soup('

a<Hello>") + # Compare html5lib, which completes the entity. + self.assertEquals(soup.p.string, "") + + def test_nonexistent_entity(self): + soup = self.soup("

foo&#bar;baz

") + self.assertEquals(soup.p.string, "foo&#bar;baz") + + # Compare a real entity. + soup = self.soup("

foodbaz

") + self.assertEquals(soup.p.string, "foodbaz") class TestHTML5LibEncodingConversion(TestLXMLBuilderEncodingConversion): diff --git a/tests/test_lxml.py b/tests/test_lxml.py index 11ef15a..85c6a1b 100644 --- a/tests/test_lxml.py +++ b/tests/test_lxml.py @@ -376,6 +376,55 @@ class TestLXMLBuilderInvalidMarkup(SoupTest): markup = "

" self.assertSoupEquals(markup, "

") + def test_attribute_value_never_got_closed(self): + markup = ' and blah and blah") + + def test_attribute_value_was_closed_by_subsequent_tag(self): + markup = """baz""" + soup = self.soup(markup) + # The string between the first and second quotes was interpreted + # as the value of the 'href' attribute. + self.assertEquals(soup.a['href'], 'foo, ') + self.assertEquals(soup.a['b'], '') + + def test_nonexistent_entity(self): + soup = self.soup("

foo&#bar;baz

") + self.assertEquals(soup.p.string, "foobar;baz") + + # Compare a real entity. + soup = self.soup("

foodbaz

") + self.assertEquals(soup.p.string, "foodbaz") + + # Also compare html5lib, which preserves the &# before the + # entity name. + + def test_entity_was_not_finished(self): + soup = self.soup("

<Hello>") + # Compare html5lib, which completes the entity. + self.assertEquals(soup.p.string, "a

a

') + # The declaration is ignored altogether. + self.assertEquals(soup.encode(), "

a

") + class TestLXMLBuilderEncodingConversion(SoupTest): # Test Beautiful Soup's ability to decode and encode from various -- cgit v1.2.3