import pprint import sgmllib import unittest from test import test_support class EventCollector(sgmllib.SGMLParser): def __init__(self): self.events = [] self.append = self.events.append sgmllib.SGMLParser.__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 unknown_starttag(self, tag, attrs): self.append(("starttag", tag, attrs)) def unknown_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, decl): self.append(("decl", decl)) 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 CDATAEventCollector(EventCollector): def start_cdata(self, attrs): self.append(("starttag", "cdata", attrs)) self.setliteral() class SGMLParserTestCase(unittest.TestCase): collector = EventCollector def get_events(self, source): parser = self.collector() try: for s in source: parser.feed(s) parser.close() except: #self.events = parser.events raise return parser.get_events() def check_events(self, source, expected_events): try: events = self.get_events(source) except: import sys #print >>sys.stderr, pprint.pformat(self.events) raise 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 check_parse_error(self, source): parser = EventCollector() try: parser.feed(source) parser.close() except sgmllib.SGMLParseError: pass else: self.fail("expected SGMLParseError for %r\nReceived:\n%s" % (source, pprint.pformat(parser.get_events()))) def test_doctype_decl_internal(self): inside = """\ DOCTYPE html PUBLIC '-//W3C//DTD HTML 4.01//EN' SYSTEM 'http://www.w3.org/TR/html401/strict.dtd' [ %paramEntity; ]""" self.check_events(["" % inside], [ ("decl", inside), ]) def test_doctype_decl_external(self): inside = "DOCTYPE html PUBLIC '-//W3C//DTD HTML 4.01//EN'" self.check_events("" % inside, [ ("decl", inside), ]) def test_underscore_in_attrname(self): # SF bug #436621 """Make sure attribute names with underscores are accepted""" self.check_events("", [ ("starttag", "a", [("has_under", "has_under"), ("_under", "_under")]), ]) def test_underscore_in_tagname(self): # SF bug #436621 """Make sure tag names with underscores are accepted""" self.check_events("", [ ("starttag", "has_under", []), ("endtag", "has_under"), ]) def test_quotes_in_unquoted_attrs(self): # SF bug #436621 """Be sure quotes in unquoted attributes are made part of the value""" self.check_events("", [ ("starttag", "a", [("href", "foo'bar\"baz")]), ]) def test_xhtml_empty_tag(self): """Handling of XHTML-style empty start tags""" self.check_events("
text", [ ("starttag", "br", []), ("data", "text"), ("starttag", "i", []), ("endtag", "i"), ]) def test_processing_instruction_only(self): self.check_events("", [ ("pi", "processing instruction"), ]) def test_bad_nesting(self): self.check_events("
", [ ("starttag", "a", []), ("starttag", "b", []), ("endtag", "a"), ("endtag", "b"), ]) def test_bare_ampersands(self): self.check_events("this text & contains & ampersands &", [ ("data", "this text & contains & ampersands &"), ]) def test_bare_pointy_brackets(self): self.check_events("this < text > contains < bare>pointy< brackets", [ ("data", "this < text > contains < bare>pointy< brackets"), ]) def test_attr_syntax(self): output = [ ("starttag", "a", [("b", "v"), ("c", "v"), ("d", "v"), ("e", "e")]) ] self.check_events("""""", output) self.check_events("""""", output) self.check_events("""""", output) self.check_events("""""", output) def test_attr_values(self): self.check_events("""""", [("starttag", "a", [("b", "xxx\n\txxx"), ("c", "yyy\t\nyyy"), ("d", "\txyz\n")]) ]) self.check_events("""""", [ ("starttag", "a", [("b", ""), ("c", "")]), ]) def test_attr_funky_names(self): self.check_events("""""", [ ("starttag", "a", [("a.b", "v"), ("c:d", "v"), ("e-f", "v")]), ]) def test_illegal_declarations(self): s = 'abcdef' self.check_events(s, [ ("data", "abc"), ("unknown decl", 'spacer type="block" height="25"'), ("data", "def"), ]) def test_weird_starttags(self): self.check_events("", [ ("starttag", "a", []), ("starttag", "a", []), ]) self.check_events("", [ ("endtag", "a"), ("starttag", "a", []), ]) def test_declaration_junk_chars(self): self.check_parse_error("") def test_get_starttag_text(self): s = """""" self.check_events(s, [ ("starttag", "foobar", [("one", "1"), ("two", "2")]), ]) def test_cdata_content(self): s = (" ¬-an-entity-ref; " " ") self.collector = CDATAEventCollector self.check_events(s, [ ("starttag", "cdata", []), ("data", " ¬-an-entity-ref; "), ("endtag", "cdata"), ("starttag", "notcdata", []), ("data", " "), ("comment", " comment "), ("data", " "), ("endtag", "notcdata"), ]) s = """ """ self.check_events(s, [ ("starttag", "cdata", []), ("data", " "), ("endtag", "cdata"), ]) def test_illegal_declarations(self): s = 'abcdef' self.check_events(s, [ ("data", "abc"), ("unknown decl", 'spacer type="block" height="25"'), ("data", "def"), ]) def test_enumerated_attr_type(self): s = "]>" self.check_events(s, [ ('decl', 'DOCTYPE doc []'), ]) # XXX These tests have been disabled by prefixing their names with # an underscore. The first two exercise outstanding bugs in the # sgmllib module, and the third exhibits questionable behavior # that needs to be carefully considered before changing it. def _test_starttag_end_boundary(self): self.check_events("""""", [("starttag", "a", [("b", "<")])]) self.check_events("""""", [("starttag", "a", [("b", ">")])]) def _test_buffer_artefacts(self): output = [("starttag", "a", [("b", "<")])] self.check_events([""], output) self.check_events([""], output) self.check_events([""], output) self.check_events([""], output) self.check_events([""], output) self.check_events([""], output) output = [("starttag", "a", [("b", ">")])] self.check_events([""], output) self.check_events([""], output) self.check_events([""], output) self.check_events(["'>"], output) self.check_events([""], output) self.check_events([""], output) def _test_starttag_junk_chars(self): self.check_parse_error("<") self.check_parse_error("<>") self.check_parse_error("") self.check_parse_error("") self.check_parse_error("") self.check_parse_error("'") self.check_parse_error("