mirrors/cpython
1
0
Fork 0
mirror of https://github.com/python/cpython.git synced 2025-07-08 03:45:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 8

Enhanced the test for DOCTYPE declarations, added a test for dealing with

Browse source 
broken declaration-like things.
This commit is contained in:
Fred Drake 2001-09-04 15:13:04 +00:00
parent 68eac2b574
commit c20a698932
1 changed files with 23 additions and 14 deletions
Download patch file Download diff file Expand all files Collapse all files

37
Lib/test/test_htmlparser.py
View file

@ -60,6 +60,9 @@ class EventCollector(HTMLParser.HTMLParser):
def handle_pi(self, data):
self.append(("pi", data))
def unknown_decl(self, decl):
self.append(("unknown decl", decl))
class EventCollectorExtra(EventCollector):
@ -70,24 +73,16 @@ class EventCollectorExtra(EventCollector):
class TestCaseBase(unittest.TestCase):
# Constant pieces of source and events
prologue = ""
epilogue = ""
initial_events = []
final_events = []
def _run_check(self, source, events, collector=EventCollector):
def _run_check(self, source, expected_events, collector=EventCollector):
parser = collector()
parser.feed(self.prologue)
for s in source:
parser.feed(s)
for c in self.epilogue:
parser.feed(c)
parser.close()
events = parser.get_events()
self.assertEqual(events,
self.initial_events + events + self.final_events,
"got events:\n" + pprint.pformat(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)
@ -144,7 +139,13 @@ text
DOCTYPE html [
<!ELEMENT html - O EMPTY>
<!ATTLIST html
version CDATA #IMPLIED '4.0'>
version CDATA #IMPLIED
profile CDATA 'DublinCore'>
<!NOTATION datatype SYSTEM 'http://xml.python.org/notations/python-module'>
<!ENTITY myEntity 'internal parsed entity'>
<!ENTITY anEntity SYSTEM 'http://xml.python.org/entities/something.xml'>
<!ENTITY % paramEntity 'name|name|name'>
%paramEntity;
<!-- comment -->
]"""
self._run_check("<!%s>" % inside, [
@ -201,6 +202,14 @@ DOCTYPE html [
("starttag", "a", [("a.b", "v"), ("c:d", "v"), ("e-f", "v")]),
])
def test_illegal_declarations(self):
s = 'abc<!spacer type="block" height="25">def'
self._run_check(s, [
("data", "abc"),
("unknown decl", 'spacer type="block" height="25"'),
("data", "def"),
])
def test_starttag_end_boundary(self):
self._run_check("""<a b='<'>""", [("starttag", "a", [("b", "<")])])
self._run_check("""<a b='>'>""", [("starttag", "a", [("b", ">")])])