1 """Tests for HTMLParser.py."""
2
3 import html.parser
4 import pprint
5 import unittest
6
7 from unittest.mock import patch
8
9
10 class ESC[4;38;5;81mEventCollector(ESC[4;38;5;149mhtmlESC[4;38;5;149m.ESC[4;38;5;149mparserESC[4;38;5;149m.ESC[4;38;5;149mHTMLParser):
11
12 def __init__(self, *args, **kw):
13 self.events = []
14 self.append = self.events.append
15 html.parser.HTMLParser.__init__(self, *args, **kw)
16
17 def get_events(self):
18 # Normalize the list of events so that buffer artefacts don't
19 # separate runs of contiguous characters.
20 L = []
21 prevtype = None
22 for event in self.events:
23 type = event[0]
24 if type == prevtype == "data":
25 L[-1] = ("data", L[-1][1] + event[1])
26 else:
27 L.append(event)
28 prevtype = type
29 self.events = L
30 return L
31
32 # structure markup
33
34 def handle_starttag(self, tag, attrs):
35 self.append(("starttag", tag, attrs))
36
37 def handle_startendtag(self, tag, attrs):
38 self.append(("startendtag", tag, attrs))
39
40 def handle_endtag(self, tag):
41 self.append(("endtag", tag))
42
43 # all other markup
44
45 def handle_comment(self, data):
46 self.append(("comment", data))
47
48 def handle_charref(self, data):
49 self.append(("charref", data))
50
51 def handle_data(self, data):
52 self.append(("data", data))
53
54 def handle_decl(self, data):
55 self.append(("decl", data))
56
57 def handle_entityref(self, data):
58 self.append(("entityref", data))
59
60 def handle_pi(self, data):
61 self.append(("pi", data))
62
63 def unknown_decl(self, decl):
64 self.append(("unknown decl", decl))
65
66
67 class ESC[4;38;5;81mEventCollectorExtra(ESC[4;38;5;149mEventCollector):
68
69 def handle_starttag(self, tag, attrs):
70 EventCollector.handle_starttag(self, tag, attrs)
71 self.append(("starttag_text", self.get_starttag_text()))
72
73
74 class ESC[4;38;5;81mEventCollectorCharrefs(ESC[4;38;5;149mEventCollector):
75
76 def handle_charref(self, data):
77 self.fail('This should never be called with convert_charrefs=True')
78
79 def handle_entityref(self, data):
80 self.fail('This should never be called with convert_charrefs=True')
81
82
83 class ESC[4;38;5;81mTestCaseBase(ESC[4;38;5;149munittestESC[4;38;5;149m.ESC[4;38;5;149mTestCase):
84
85 def get_collector(self):
86 return EventCollector(convert_charrefs=False)
87
88 def _run_check(self, source, expected_events, collector=None):
89 if collector is None:
90 collector = self.get_collector()
91 parser = collector
92 for s in source:
93 parser.feed(s)
94 parser.close()
95 events = parser.get_events()
96 if events != expected_events:
97 self.fail("received events did not match expected events" +
98 "\nSource:\n" + repr(source) +
99 "\nExpected:\n" + pprint.pformat(expected_events) +
100 "\nReceived:\n" + pprint.pformat(events))
101
102 def _run_check_extra(self, source, events):
103 self._run_check(source, events,
104 EventCollectorExtra(convert_charrefs=False))
105
106
107 class ESC[4;38;5;81mHTMLParserTestCase(ESC[4;38;5;149mTestCaseBase):
108
109 def test_processing_instruction_only(self):
110 self._run_check("<?processing instruction>", [
111 ("pi", "processing instruction"),
112 ])
113 self._run_check("<?processing instruction ?>", [
114 ("pi", "processing instruction ?"),
115 ])
116
117 def test_simple_html(self):
118 self._run_check("""
119 <!DOCTYPE html PUBLIC 'foo'>
120 <HTML>&entity; 
121 <!--comment1a
122 -></foo><bar><<?pi?></foo<bar
123 comment1b-->
124 <Img sRc='Bar' isMAP>sample
125 text
126 “
127 <!--comment2a-- --comment2b-->
128 </Html>
129 """, [
130 ("data", "\n"),
131 ("decl", "DOCTYPE html PUBLIC 'foo'"),
132 ("data", "\n"),
133 ("starttag", "html", []),
134 ("entityref", "entity"),
135 ("charref", "32"),
136 ("data", "\n"),
137 ("comment", "comment1a\n-></foo><bar><<?pi?></foo<bar\ncomment1b"),
138 ("data", "\n"),
139 ("starttag", "img", [("src", "Bar"), ("ismap", None)]),
140 ("data", "sample\ntext\n"),
141 ("charref", "x201C"),
142 ("data", "\n"),
143 ("comment", "comment2a-- --comment2b"),
144 ("data", "\n"),
145 ("endtag", "html"),
146 ("data", "\n"),
147 ])
148
149 def test_malformatted_charref(self):
150 self._run_check("<p>&#bad;</p>", [
151 ("starttag", "p", []),
152 ("data", "&#bad;"),
153 ("endtag", "p"),
154 ])
155 # add the [] as a workaround to avoid buffering (see #20288)
156 self._run_check(["<div>&#bad;</div>"], [
157 ("starttag", "div", []),
158 ("data", "&#bad;"),
159 ("endtag", "div"),
160 ])
161
162 def test_unclosed_entityref(self):
163 self._run_check("&entityref foo", [
164 ("entityref", "entityref"),
165 ("data", " foo"),
166 ])
167
168 def test_bad_nesting(self):
169 # Strangely, this *is* supposed to test that overlapping
170 # elements are allowed. HTMLParser is more geared toward
171 # lexing the input that parsing the structure.
172 self._run_check("<a><b></a></b>", [
173 ("starttag", "a", []),
174 ("starttag", "b", []),
175 ("endtag", "a"),
176 ("endtag", "b"),
177 ])
178
179 def test_bare_ampersands(self):
180 self._run_check("this text & contains & ampersands &", [
181 ("data", "this text & contains & ampersands &"),
182 ])
183
184 def test_bare_pointy_brackets(self):
185 self._run_check("this < text > contains < bare>pointy< brackets", [
186 ("data", "this < text > contains < bare>pointy< brackets"),
187 ])
188
189 def test_starttag_end_boundary(self):
190 self._run_check("""<a b='<'>""", [("starttag", "a", [("b", "<")])])
191 self._run_check("""<a b='>'>""", [("starttag", "a", [("b", ">")])])
192
193 def test_buffer_artefacts(self):
194 output = [("starttag", "a", [("b", "<")])]
195 self._run_check(["<a b='<'>"], output)
196 self._run_check(["<a ", "b='<'>"], output)
197 self._run_check(["<a b", "='<'>"], output)
198 self._run_check(["<a b=", "'<'>"], output)
199 self._run_check(["<a b='<", "'>"], output)
200 self._run_check(["<a b='<'", ">"], output)
201
202 output = [("starttag", "a", [("b", ">")])]
203 self._run_check(["<a b='>'>"], output)
204 self._run_check(["<a ", "b='>'>"], output)
205 self._run_check(["<a b", "='>'>"], output)
206 self._run_check(["<a b=", "'>'>"], output)
207 self._run_check(["<a b='>", "'>"], output)
208 self._run_check(["<a b='>'", ">"], output)
209
210 output = [("comment", "abc")]
211 self._run_check(["", "<!--abc-->"], output)
212 self._run_check(["<", "!--abc-->"], output)
213 self._run_check(["<!", "--abc-->"], output)
214 self._run_check(["<!-", "-abc-->"], output)
215 self._run_check(["<!--", "abc-->"], output)
216 self._run_check(["<!--a", "bc-->"], output)
217 self._run_check(["<!--ab", "c-->"], output)
218 self._run_check(["<!--abc", "-->"], output)
219 self._run_check(["<!--abc-", "->"], output)
220 self._run_check(["<!--abc--", ">"], output)
221 self._run_check(["<!--abc-->", ""], output)
222
223 def test_valid_doctypes(self):
224 # from http://www.w3.org/QA/2002/04/valid-dtd-list.html
225 dtds = ['HTML', # HTML5 doctype
226 ('HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" '
227 '"http://www.w3.org/TR/html4/strict.dtd"'),
228 ('HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" '
229 '"http://www.w3.org/TR/html4/loose.dtd"'),
230 ('html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" '
231 '"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"'),
232 ('html PUBLIC "-//W3C//DTD XHTML 1.0 Frameset//EN" '
233 '"http://www.w3.org/TR/xhtml1/DTD/xhtml1-frameset.dtd"'),
234 ('math PUBLIC "-//W3C//DTD MathML 2.0//EN" '
235 '"http://www.w3.org/Math/DTD/mathml2/mathml2.dtd"'),
236 ('html PUBLIC "-//W3C//DTD '
237 'XHTML 1.1 plus MathML 2.0 plus SVG 1.1//EN" '
238 '"http://www.w3.org/2002/04/xhtml-math-svg/xhtml-math-svg.dtd"'),
239 ('svg PUBLIC "-//W3C//DTD SVG 1.1//EN" '
240 '"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd"'),
241 'html PUBLIC "-//IETF//DTD HTML 2.0//EN"',
242 'html PUBLIC "-//W3C//DTD HTML 3.2 Final//EN"']
243 for dtd in dtds:
244 self._run_check("<!DOCTYPE %s>" % dtd,
245 [('decl', 'DOCTYPE ' + dtd)])
246
247 def test_startendtag(self):
248 self._run_check("<p/>", [
249 ("startendtag", "p", []),
250 ])
251 self._run_check("<p></p>", [
252 ("starttag", "p", []),
253 ("endtag", "p"),
254 ])
255 self._run_check("<p><img src='foo' /></p>", [
256 ("starttag", "p", []),
257 ("startendtag", "img", [("src", "foo")]),
258 ("endtag", "p"),
259 ])
260
261 def test_get_starttag_text(self):
262 s = """<foo:bar \n one="1"\ttwo=2 >"""
263 self._run_check_extra(s, [
264 ("starttag", "foo:bar", [("one", "1"), ("two", "2")]),
265 ("starttag_text", s)])
266
267 def test_cdata_content(self):
268 contents = [
269 '<!-- not a comment --> ¬-an-entity-ref;',
270 "<not a='start tag'>",
271 '<a href="" /> <p> <span></span>',
272 'foo = "</scr" + "ipt>";',
273 'foo = "</SCRIPT" + ">";',
274 'foo = <\n/script> ',
275 '<!-- document.write("</scr" + "ipt>"); -->',
276 ('\n//<![CDATA[\n'
277 'document.write(\'<s\'+\'cript type="text/javascript" '
278 'src="http://www.example.org/r=\'+new '
279 'Date().getTime()+\'"><\\/s\'+\'cript>\');\n//]]>'),
280 '\n<!-- //\nvar foo = 3.14;\n// -->\n',
281 'foo = "</sty" + "le>";',
282 '<!-- \u2603 -->',
283 # these two should be invalid according to the HTML 5 spec,
284 # section 8.1.2.2
285 #'foo = </\nscript>',
286 #'foo = </ script>',
287 ]
288 elements = ['script', 'style', 'SCRIPT', 'STYLE', 'Script', 'Style']
289 for content in contents:
290 for element in elements:
291 element_lower = element.lower()
292 s = '<{element}>{content}</{element}>'.format(element=element,
293 content=content)
294 self._run_check(s, [("starttag", element_lower, []),
295 ("data", content),
296 ("endtag", element_lower)])
297
298 def test_cdata_with_closing_tags(self):
299 # see issue #13358
300 # make sure that HTMLParser calls handle_data only once for each CDATA.
301 # The normal event collector normalizes the events in get_events,
302 # so we override it to return the original list of events.
303 class ESC[4;38;5;81mCollector(ESC[4;38;5;149mEventCollector):
304 def get_events(self):
305 return self.events
306
307 content = """<!-- not a comment --> ¬-an-entity-ref;
308 <a href="" /> </p><p> <span></span></style>
309 '</script' + '>'"""
310 for element in [' script', 'script ', ' script ',
311 '\nscript', 'script\n', '\nscript\n']:
312 element_lower = element.lower().strip()
313 s = '<script>{content}</{element}>'.format(element=element,
314 content=content)
315 self._run_check(s, [("starttag", element_lower, []),
316 ("data", content),
317 ("endtag", element_lower)],
318 collector=Collector(convert_charrefs=False))
319
320 def test_comments(self):
321 html = ("<!-- I'm a valid comment -->"
322 '<!--me too!-->'
323 '<!------>'
324 '<!---->'
325 '<!----I have many hyphens---->'
326 '<!-- I have a > in the middle -->'
327 '<!-- and I have -- in the middle! -->')
328 expected = [('comment', " I'm a valid comment "),
329 ('comment', 'me too!'),
330 ('comment', '--'),
331 ('comment', ''),
332 ('comment', '--I have many hyphens--'),
333 ('comment', ' I have a > in the middle '),
334 ('comment', ' and I have -- in the middle! ')]
335 self._run_check(html, expected)
336
337 def test_condcoms(self):
338 html = ('<!--[if IE & !(lte IE 8)]>aren\'t<![endif]-->'
339 '<!--[if IE 8]>condcoms<![endif]-->'
340 '<!--[if lte IE 7]>pretty?<![endif]-->')
341 expected = [('comment', "[if IE & !(lte IE 8)]>aren't<![endif]"),
342 ('comment', '[if IE 8]>condcoms<![endif]'),
343 ('comment', '[if lte IE 7]>pretty?<![endif]')]
344 self._run_check(html, expected)
345
346 def test_convert_charrefs(self):
347 # default value for convert_charrefs is now True
348 collector = lambda: EventCollectorCharrefs()
349 self.assertTrue(collector().convert_charrefs)
350 charrefs = ['"', '"', '"', '"', '"', '"']
351 # check charrefs in the middle of the text/attributes
352 expected = [('starttag', 'a', [('href', 'foo"zar')]),
353 ('data', 'a"z'), ('endtag', 'a')]
354 for charref in charrefs:
355 self._run_check('<a href="foo{0}zar">a{0}z</a>'.format(charref),
356 expected, collector=collector())
357 # check charrefs at the beginning/end of the text/attributes
358 expected = [('data', '"'),
359 ('starttag', 'a', [('x', '"'), ('y', '"X'), ('z', 'X"')]),
360 ('data', '"'), ('endtag', 'a'), ('data', '"')]
361 for charref in charrefs:
362 self._run_check('{0}<a x="{0}" y="{0}X" z="X{0}">'
363 '{0}</a>{0}'.format(charref),
364 expected, collector=collector())
365 # check charrefs in <script>/<style> elements
366 for charref in charrefs:
367 text = 'X'.join([charref]*3)
368 expected = [('data', '"'),
369 ('starttag', 'script', []), ('data', text),
370 ('endtag', 'script'), ('data', '"'),
371 ('starttag', 'style', []), ('data', text),
372 ('endtag', 'style'), ('data', '"')]
373 self._run_check('{1}<script>{0}</script>{1}'
374 '<style>{0}</style>{1}'.format(text, charref),
375 expected, collector=collector())
376 # check truncated charrefs at the end of the file
377 html = '&quo &# &#x'
378 for x in range(1, len(html)):
379 self._run_check(html[:x], [('data', html[:x])],
380 collector=collector())
381 # check a string with no charrefs
382 self._run_check('no charrefs here', [('data', 'no charrefs here')],
383 collector=collector())
384
385 # the remaining tests were for the "tolerant" parser (which is now
386 # the default), and check various kind of broken markup
387 def test_tolerant_parsing(self):
388 self._run_check('<html <html>te>>xt&a<<bc</a></html>\n'
389 '<img src="URL><//img></html</html>', [
390 ('starttag', 'html', [('<html', None)]),
391 ('data', 'te>>xt'),
392 ('entityref', 'a'),
393 ('data', '<'),
394 ('starttag', 'bc<', [('a', None)]),
395 ('endtag', 'html'),
396 ('data', '\n<img src="URL>'),
397 ('comment', '/img'),
398 ('endtag', 'html<')])
399
400 def test_starttag_junk_chars(self):
401 self._run_check("</>", [])
402 self._run_check("</$>", [('comment', '$')])
403 self._run_check("</", [('data', '</')])
404 self._run_check("</a", [('data', '</a')])
405 self._run_check("<a<a>", [('starttag', 'a<a', [])])
406 self._run_check("</a<a>", [('endtag', 'a<a')])
407 self._run_check("<!", [('data', '<!')])
408 self._run_check("<a", [('data', '<a')])
409 self._run_check("<a foo='bar'", [('data', "<a foo='bar'")])
410 self._run_check("<a foo='bar", [('data', "<a foo='bar")])
411 self._run_check("<a foo='>'", [('data', "<a foo='>'")])
412 self._run_check("<a foo='>", [('data', "<a foo='>")])
413 self._run_check("<a$>", [('starttag', 'a$', [])])
414 self._run_check("<a$b>", [('starttag', 'a$b', [])])
415 self._run_check("<a$b/>", [('startendtag', 'a$b', [])])
416 self._run_check("<a$b >", [('starttag', 'a$b', [])])
417 self._run_check("<a$b />", [('startendtag', 'a$b', [])])
418
419 def test_slashes_in_starttag(self):
420 self._run_check('<a foo="var"/>', [('startendtag', 'a', [('foo', 'var')])])
421 html = ('<img width=902 height=250px '
422 'src="/sites/default/files/images/homepage/foo.jpg" '
423 '/*what am I doing here*/ />')
424 expected = [(
425 'startendtag', 'img',
426 [('width', '902'), ('height', '250px'),
427 ('src', '/sites/default/files/images/homepage/foo.jpg'),
428 ('*what', None), ('am', None), ('i', None),
429 ('doing', None), ('here*', None)]
430 )]
431 self._run_check(html, expected)
432 html = ('<a / /foo/ / /=/ / /bar/ / />'
433 '<a / /foo/ / /=/ / /bar/ / >')
434 expected = [
435 ('startendtag', 'a', [('foo', None), ('=', None), ('bar', None)]),
436 ('starttag', 'a', [('foo', None), ('=', None), ('bar', None)])
437 ]
438 self._run_check(html, expected)
439 #see issue #14538
440 html = ('<meta><meta / ><meta // ><meta / / >'
441 '<meta/><meta /><meta //><meta//>')
442 expected = [
443 ('starttag', 'meta', []), ('starttag', 'meta', []),
444 ('starttag', 'meta', []), ('starttag', 'meta', []),
445 ('startendtag', 'meta', []), ('startendtag', 'meta', []),
446 ('startendtag', 'meta', []), ('startendtag', 'meta', []),
447 ]
448 self._run_check(html, expected)
449
450 def test_declaration_junk_chars(self):
451 self._run_check("<!DOCTYPE foo $ >", [('decl', 'DOCTYPE foo $ ')])
452
453 def test_illegal_declarations(self):
454 self._run_check('<!spacer type="block" height="25">',
455 [('comment', 'spacer type="block" height="25"')])
456
457 def test_invalid_end_tags(self):
458 # A collection of broken end tags. <br> is used as separator.
459 # see http://www.w3.org/TR/html5/tokenization.html#end-tag-open-state
460 # and #13993
461 html = ('<br></label</p><br></div end tmAd-leaderBoard><br></<h4><br>'
462 '</li class="unit"><br></li\r\n\t\t\t\t\t\t</ul><br></><br>')
463 expected = [('starttag', 'br', []),
464 # < is part of the name, / is discarded, p is an attribute
465 ('endtag', 'label<'),
466 ('starttag', 'br', []),
467 # text and attributes are discarded
468 ('endtag', 'div'),
469 ('starttag', 'br', []),
470 # comment because the first char after </ is not a-zA-Z
471 ('comment', '<h4'),
472 ('starttag', 'br', []),
473 # attributes are discarded
474 ('endtag', 'li'),
475 ('starttag', 'br', []),
476 # everything till ul (included) is discarded
477 ('endtag', 'li'),
478 ('starttag', 'br', []),
479 # </> is ignored
480 ('starttag', 'br', [])]
481 self._run_check(html, expected)
482
483 def test_broken_invalid_end_tag(self):
484 # This is technically wrong (the "> shouldn't be included in the 'data')
485 # but is probably not worth fixing it (in addition to all the cases of
486 # the previous test, it would require a full attribute parsing).
487 # see #13993
488 html = '<b>This</b attr=">"> confuses the parser'
489 expected = [('starttag', 'b', []),
490 ('data', 'This'),
491 ('endtag', 'b'),
492 ('data', '"> confuses the parser')]
493 self._run_check(html, expected)
494
495 def test_correct_detection_of_start_tags(self):
496 # see #13273
497 html = ('<div style="" ><b>The <a href="some_url">rain</a> '
498 '<br /> in <span>Spain</span></b></div>')
499 expected = [
500 ('starttag', 'div', [('style', '')]),
501 ('starttag', 'b', []),
502 ('data', 'The '),
503 ('starttag', 'a', [('href', 'some_url')]),
504 ('data', 'rain'),
505 ('endtag', 'a'),
506 ('data', ' '),
507 ('startendtag', 'br', []),
508 ('data', ' in '),
509 ('starttag', 'span', []),
510 ('data', 'Spain'),
511 ('endtag', 'span'),
512 ('endtag', 'b'),
513 ('endtag', 'div')
514 ]
515 self._run_check(html, expected)
516
517 html = '<div style="", foo = "bar" ><b>The <a href="some_url">rain</a>'
518 expected = [
519 ('starttag', 'div', [('style', ''), (',', None), ('foo', 'bar')]),
520 ('starttag', 'b', []),
521 ('data', 'The '),
522 ('starttag', 'a', [('href', 'some_url')]),
523 ('data', 'rain'),
524 ('endtag', 'a'),
525 ]
526 self._run_check(html, expected)
527
528 def test_EOF_in_charref(self):
529 # see #17802
530 # This test checks that the UnboundLocalError reported in the issue
531 # is not raised, however I'm not sure the returned values are correct.
532 # Maybe HTMLParser should use self.unescape for these
533 data = [
534 ('a&', [('data', 'a&')]),
535 ('a&b', [('data', 'ab')]),
536 ('a&b ', [('data', 'a'), ('entityref', 'b'), ('data', ' ')]),
537 ('a&b;', [('data', 'a'), ('entityref', 'b')]),
538 ]
539 for html, expected in data:
540 self._run_check(html, expected)
541
542 def test_broken_comments(self):
543 html = ('<! not really a comment >'
544 '<! not a comment either -->'
545 '<! -- close enough -->'
546 '<!><!<-- this was an empty comment>'
547 '<!!! another bogus comment !!!>')
548 expected = [
549 ('comment', ' not really a comment '),
550 ('comment', ' not a comment either --'),
551 ('comment', ' -- close enough --'),
552 ('comment', ''),
553 ('comment', '<-- this was an empty comment'),
554 ('comment', '!! another bogus comment !!!'),
555 ]
556 self._run_check(html, expected)
557
558 def test_broken_condcoms(self):
559 # these condcoms are missing the '--' after '<!' and before the '>'
560 html = ('<![if !(IE)]>broken condcom<![endif]>'
561 '<![if ! IE]><link href="favicon.tiff"/><![endif]>'
562 '<![if !IE 6]><img src="firefox.png" /><![endif]>'
563 '<![if !ie 6]><b>foo</b><![endif]>'
564 '<![if (!IE)|(lt IE 9)]><img src="mammoth.bmp" /><![endif]>')
565 # According to the HTML5 specs sections "8.2.4.44 Bogus comment state"
566 # and "8.2.4.45 Markup declaration open state", comment tokens should
567 # be emitted instead of 'unknown decl', but calling unknown_decl
568 # provides more flexibility.
569 # See also Lib/_markupbase.py:parse_declaration
570 expected = [
571 ('unknown decl', 'if !(IE)'),
572 ('data', 'broken condcom'),
573 ('unknown decl', 'endif'),
574 ('unknown decl', 'if ! IE'),
575 ('startendtag', 'link', [('href', 'favicon.tiff')]),
576 ('unknown decl', 'endif'),
577 ('unknown decl', 'if !IE 6'),
578 ('startendtag', 'img', [('src', 'firefox.png')]),
579 ('unknown decl', 'endif'),
580 ('unknown decl', 'if !ie 6'),
581 ('starttag', 'b', []),
582 ('data', 'foo'),
583 ('endtag', 'b'),
584 ('unknown decl', 'endif'),
585 ('unknown decl', 'if (!IE)|(lt IE 9)'),
586 ('startendtag', 'img', [('src', 'mammoth.bmp')]),
587 ('unknown decl', 'endif')
588 ]
589 self._run_check(html, expected)
590
591 def test_convert_charrefs_dropped_text(self):
592 # #23144: make sure that all the events are triggered when
593 # convert_charrefs is True, even if we don't call .close()
594 parser = EventCollector(convert_charrefs=True)
595 # before the fix, bar & baz was missing
596 parser.feed("foo <a>link</a> bar & baz")
597 self.assertEqual(
598 parser.get_events(),
599 [('data', 'foo '), ('starttag', 'a', []), ('data', 'link'),
600 ('endtag', 'a'), ('data', ' bar & baz')]
601 )
602
603
604 class ESC[4;38;5;81mAttributesTestCase(ESC[4;38;5;149mTestCaseBase):
605
606 def test_attr_syntax(self):
607 output = [
608 ("starttag", "a", [("b", "v"), ("c", "v"), ("d", "v"), ("e", None)])
609 ]
610 self._run_check("""<a b='v' c="v" d=v e>""", output)
611 self._run_check("""<a b = 'v' c = "v" d = v e>""", output)
612 self._run_check("""<a\nb\n=\n'v'\nc\n=\n"v"\nd\n=\nv\ne>""", output)
613 self._run_check("""<a\tb\t=\t'v'\tc\t=\t"v"\td\t=\tv\te>""", output)
614
615 def test_attr_values(self):
616 self._run_check("""<a b='xxx\n\txxx' c="yyy\t\nyyy" d='\txyz\n'>""",
617 [("starttag", "a", [("b", "xxx\n\txxx"),
618 ("c", "yyy\t\nyyy"),
619 ("d", "\txyz\n")])])
620 self._run_check("""<a b='' c="">""",
621 [("starttag", "a", [("b", ""), ("c", "")])])
622 # Regression test for SF patch #669683.
623 self._run_check("<e a=rgb(1,2,3)>",
624 [("starttag", "e", [("a", "rgb(1,2,3)")])])
625 # Regression test for SF bug #921657.
626 self._run_check(
627 "<a href=mailto:xyz@example.com>",
628 [("starttag", "a", [("href", "mailto:xyz@example.com")])])
629
630 def test_attr_nonascii(self):
631 # see issue 7311
632 self._run_check(
633 "<img src=/foo/bar.png alt=\u4e2d\u6587>",
634 [("starttag", "img", [("src", "/foo/bar.png"),
635 ("alt", "\u4e2d\u6587")])])
636 self._run_check(
637 "<a title='\u30c6\u30b9\u30c8' href='\u30c6\u30b9\u30c8.html'>",
638 [("starttag", "a", [("title", "\u30c6\u30b9\u30c8"),
639 ("href", "\u30c6\u30b9\u30c8.html")])])
640 self._run_check(
641 '<a title="\u30c6\u30b9\u30c8" href="\u30c6\u30b9\u30c8.html">',
642 [("starttag", "a", [("title", "\u30c6\u30b9\u30c8"),
643 ("href", "\u30c6\u30b9\u30c8.html")])])
644
645 def test_attr_entity_replacement(self):
646 self._run_check(
647 "<a b='&><"''>",
648 [("starttag", "a", [("b", "&><\"'")])])
649
650 def test_attr_funky_names(self):
651 self._run_check(
652 "<a a.b='v' c:d=v e-f=v>",
653 [("starttag", "a", [("a.b", "v"), ("c:d", "v"), ("e-f", "v")])])
654
655 def test_entityrefs_in_attributes(self):
656 self._run_check(
657 "<html foo='€&aa&unsupported;'>",
658 [("starttag", "html", [("foo", "\u20AC&aa&unsupported;")])])
659
660
661 def test_attr_funky_names2(self):
662 self._run_check(
663 r"<a $><b $=%><c \=/>",
664 [("starttag", "a", [("$", None)]),
665 ("starttag", "b", [("$", "%")]),
666 ("starttag", "c", [("\\", "/")])])
667
668 def test_entities_in_attribute_value(self):
669 # see #1200313
670 for entity in ['&', '&', '&', '&']:
671 self._run_check('<a href="%s">' % entity,
672 [("starttag", "a", [("href", "&")])])
673 self._run_check("<a href='%s'>" % entity,
674 [("starttag", "a", [("href", "&")])])
675 self._run_check("<a href=%s>" % entity,
676 [("starttag", "a", [("href", "&")])])
677
678 def test_malformed_attributes(self):
679 # see #13357
680 html = (
681 "<a href=test'style='color:red;bad1'>test - bad1</a>"
682 "<a href=test'+style='color:red;ba2'>test - bad2</a>"
683 "<a href=test' style='color:red;bad3'>test - bad3</a>"
684 "<a href = test' style='color:red;bad4' >test - bad4</a>"
685 )
686 expected = [
687 ('starttag', 'a', [('href', "test'style='color:red;bad1'")]),
688 ('data', 'test - bad1'), ('endtag', 'a'),
689 ('starttag', 'a', [('href', "test'+style='color:red;ba2'")]),
690 ('data', 'test - bad2'), ('endtag', 'a'),
691 ('starttag', 'a', [('href', "test'\xa0style='color:red;bad3'")]),
692 ('data', 'test - bad3'), ('endtag', 'a'),
693 ('starttag', 'a', [('href', "test'\xa0style='color:red;bad4'")]),
694 ('data', 'test - bad4'), ('endtag', 'a')
695 ]
696 self._run_check(html, expected)
697
698 def test_malformed_adjacent_attributes(self):
699 # see #12629
700 self._run_check('<x><y z=""o"" /></x>',
701 [('starttag', 'x', []),
702 ('startendtag', 'y', [('z', ''), ('o""', None)]),
703 ('endtag', 'x')])
704 self._run_check('<x><y z="""" /></x>',
705 [('starttag', 'x', []),
706 ('startendtag', 'y', [('z', ''), ('""', None)]),
707 ('endtag', 'x')])
708
709 # see #755670 for the following 3 tests
710 def test_adjacent_attributes(self):
711 self._run_check('<a width="100%"cellspacing=0>',
712 [("starttag", "a",
713 [("width", "100%"), ("cellspacing","0")])])
714
715 self._run_check('<a id="foo"class="bar">',
716 [("starttag", "a",
717 [("id", "foo"), ("class","bar")])])
718
719 def test_missing_attribute_value(self):
720 self._run_check('<a v=>',
721 [("starttag", "a", [("v", "")])])
722
723 def test_javascript_attribute_value(self):
724 self._run_check("<a href=javascript:popup('/popup/help.html')>",
725 [("starttag", "a",
726 [("href", "javascript:popup('/popup/help.html')")])])
727
728 def test_end_tag_in_attribute_value(self):
729 # see #1745761
730 self._run_check("<a href='http://www.example.org/\">;'>spam</a>",
731 [("starttag", "a",
732 [("href", "http://www.example.org/\">;")]),
733 ("data", "spam"), ("endtag", "a")])
734
735 def test_with_unquoted_attributes(self):
736 # see #12008
737 html = ("<html><body bgcolor=d0ca90 text='181008'>"
738 "<table cellspacing=0 cellpadding=1 width=100% ><tr>"
739 "<td align=left><font size=-1>"
740 "- <a href=/rabota/><span class=en> software-and-i</span></a>"
741 "- <a href='/1/'><span class=en> library</span></a></table>")
742 expected = [
743 ('starttag', 'html', []),
744 ('starttag', 'body', [('bgcolor', 'd0ca90'), ('text', '181008')]),
745 ('starttag', 'table',
746 [('cellspacing', '0'), ('cellpadding', '1'), ('width', '100%')]),
747 ('starttag', 'tr', []),
748 ('starttag', 'td', [('align', 'left')]),
749 ('starttag', 'font', [('size', '-1')]),
750 ('data', '- '), ('starttag', 'a', [('href', '/rabota/')]),
751 ('starttag', 'span', [('class', 'en')]), ('data', ' software-and-i'),
752 ('endtag', 'span'), ('endtag', 'a'),
753 ('data', '- '), ('starttag', 'a', [('href', '/1/')]),
754 ('starttag', 'span', [('class', 'en')]), ('data', ' library'),
755 ('endtag', 'span'), ('endtag', 'a'), ('endtag', 'table')
756 ]
757 self._run_check(html, expected)
758
759 def test_comma_between_attributes(self):
760 # see bpo 41478
761 # HTMLParser preserves duplicate attributes, leaving the task of
762 # removing duplicate attributes to a conformant html tree builder
763 html = ('<div class=bar,baz=asd>' # between attrs (unquoted)
764 '<div class="bar",baz="asd">' # between attrs (quoted)
765 '<div class=bar, baz=asd,>' # after values (unquoted)
766 '<div class="bar", baz="asd",>' # after values (quoted)
767 '<div class="bar",>' # one comma values (quoted)
768 '<div class=,bar baz=,asd>' # before values (unquoted)
769 '<div class=,"bar" baz=,"asd">' # before values (quoted)
770 '<div ,class=bar ,baz=asd>' # before names
771 '<div class,="bar" baz,="asd">' # after names
772 )
773 expected = [
774 ('starttag', 'div', [('class', 'bar,baz=asd'),]),
775 ('starttag', 'div', [('class', 'bar'), (',baz', 'asd')]),
776 ('starttag', 'div', [('class', 'bar,'), ('baz', 'asd,')]),
777 ('starttag', 'div', [('class', 'bar'), (',', None),
778 ('baz', 'asd'), (',', None)]),
779 ('starttag', 'div', [('class', 'bar'), (',', None)]),
780 ('starttag', 'div', [('class', ',bar'), ('baz', ',asd')]),
781 ('starttag', 'div', [('class', ',"bar"'), ('baz', ',"asd"')]),
782 ('starttag', 'div', [(',class', 'bar'), (',baz', 'asd')]),
783 ('starttag', 'div', [('class,', 'bar'), ('baz,', 'asd')]),
784 ]
785 self._run_check(html, expected)
786
787 def test_weird_chars_in_unquoted_attribute_values(self):
788 self._run_check('<form action=bogus|&#()value>', [
789 ('starttag', 'form',
790 [('action', 'bogus|&#()value')])])
791
792
793 class ESC[4;38;5;81mTestInheritance(ESC[4;38;5;149munittestESC[4;38;5;149m.ESC[4;38;5;149mTestCase):
794
795 @patch("_markupbase.ParserBase.__init__")
796 @patch("_markupbase.ParserBase.reset")
797 def test_base_class_methods_called(self, super_reset_method, super_init_method):
798 with patch('_markupbase.ParserBase') as parser_base:
799 EventCollector()
800 super_init_method.assert_called_once()
801 super_reset_method.assert_called_once()
802
803
804 if __name__ == "__main__":
805 unittest.main()