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