Mercurial > repos > guerler > springsuite

comparison planemo/lib/python3.7/site-packages/bs4/tests/test_tree.py @ 0:d30785e31577 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload commit 6eee67778febed82ddd413c3ca40b3183a3898f1"
author guerler
date Fri, 31 Jul 2020 00:18:57 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:d30785e31577
1 # -*- coding: utf-8 -*-
2 """Tests for Beautiful Soup's tree traversal methods.
3
4 The tree traversal methods are the main advantage of using Beautiful
5 Soup over just using a parser.
6
7 Different parsers will build different Beautiful Soup trees given the
8 same markup, but all Beautiful Soup trees can be traversed with the
9 methods tested here.
10 """
11
12 from pdb import set_trace
13 import copy
14 import pickle
15 import re
16 import warnings
17 from bs4 import BeautifulSoup
18 from bs4.builder import (
19 builder_registry,
20 HTMLParserTreeBuilder,
21 )
22 from bs4.element import (
23 PY3K,
24 CData,
25 Comment,
26 Declaration,
27 Doctype,
28 Formatter,
29 NavigableString,
30 Script,
31 SoupStrainer,
32 Stylesheet,
33 Tag,
34 TemplateString,
35 )
36 from bs4.testing import (
37 SoupTest,
38 skipIf,
39 )
40 from soupsieve import SelectorSyntaxError
41
42 XML_BUILDER_PRESENT = (builder_registry.lookup("xml") is not None)
43 LXML_PRESENT = (builder_registry.lookup("lxml") is not None)
44
45 class TreeTest(SoupTest):
46
47 def assertSelects(self, tags, should_match):
48 """Make sure that the given tags have the correct text.
49
50 This is used in tests that define a bunch of tags, each
51 containing a single string, and then select certain strings by
52 some mechanism.
53 """
54 self.assertEqual([tag.string for tag in tags], should_match)
55
56 def assertSelectsIDs(self, tags, should_match):
57 """Make sure that the given tags have the correct IDs.
58
59 This is used in tests that define a bunch of tags, each
60 containing a single string, and then select certain strings by
61 some mechanism.
62 """
63 self.assertEqual([tag['id'] for tag in tags], should_match)
64
65
66 class TestFind(TreeTest):
67 """Basic tests of the find() method.
68
69 find() just calls find_all() with limit=1, so it's not tested all
70 that thouroughly here.
71 """
72
73 def test_find_tag(self):
74 soup = self.soup("<a>1</a><b>2</b><a>3</a><b>4</b>")
75 self.assertEqual(soup.find("b").string, "2")
76
77 def test_unicode_text_find(self):
78 soup = self.soup('<h1>Räksmörgås</h1>')
79 self.assertEqual(soup.find(string='Räksmörgås'), 'Räksmörgås')
80
81 def test_unicode_attribute_find(self):
82 soup = self.soup('<h1 id="Räksmörgås">here it is</h1>')
83 str(soup)
84 self.assertEqual("here it is", soup.find(id='Räksmörgås').text)
85
86
87 def test_find_everything(self):
88 """Test an optimization that finds all tags."""
89 soup = self.soup("<a>foo</a><b>bar</b>")
90 self.assertEqual(2, len(soup.find_all()))
91
92 def test_find_everything_with_name(self):
93 """Test an optimization that finds all tags with a given name."""
94 soup = self.soup("<a>foo</a><b>bar</b><a>baz</a>")
95 self.assertEqual(2, len(soup.find_all('a')))
96
97 class TestFindAll(TreeTest):
98 """Basic tests of the find_all() method."""
99
100 def test_find_all_text_nodes(self):
101 """You can search the tree for text nodes."""
102 soup = self.soup("<html>Foo<b>bar</b>\xbb</html>")
103 # Exact match.
104 self.assertEqual(soup.find_all(string="bar"), ["bar"])
105 self.assertEqual(soup.find_all(text="bar"), ["bar"])
106 # Match any of a number of strings.
107 self.assertEqual(
108 soup.find_all(text=["Foo", "bar"]), ["Foo", "bar"])
109 # Match a regular expression.
110 self.assertEqual(soup.find_all(text=re.compile('.*')),
111 ["Foo", "bar", '\xbb'])
112 # Match anything.
113 self.assertEqual(soup.find_all(text=True),
114 ["Foo", "bar", '\xbb'])
115
116 def test_find_all_limit(self):
117 """You can limit the number of items returned by find_all."""
118 soup = self.soup("<a>1</a><a>2</a><a>3</a><a>4</a><a>5</a>")
119 self.assertSelects(soup.find_all('a', limit=3), ["1", "2", "3"])
120 self.assertSelects(soup.find_all('a', limit=1), ["1"])
121 self.assertSelects(
122 soup.find_all('a', limit=10), ["1", "2", "3", "4", "5"])
123
124 # A limit of 0 means no limit.
125 self.assertSelects(
126 soup.find_all('a', limit=0), ["1", "2", "3", "4", "5"])
127
128 def test_calling_a_tag_is_calling_findall(self):
129 soup = self.soup("<a>1</a><b>2<a id='foo'>3</a></b>")
130 self.assertSelects(soup('a', limit=1), ["1"])
131 self.assertSelects(soup.b(id="foo"), ["3"])
132
133 def test_find_all_with_self_referential_data_structure_does_not_cause_infinite_recursion(self):
134 soup = self.soup("<a></a>")
135 # Create a self-referential list.
136 l = []
137 l.append(l)
138
139 # Without special code in _normalize_search_value, this would cause infinite
140 # recursion.
141 self.assertEqual([], soup.find_all(l))
142
143 def test_find_all_resultset(self):
144 """All find_all calls return a ResultSet"""
145 soup = self.soup("<a></a>")
146 result = soup.find_all("a")
147 self.assertTrue(hasattr(result, "source"))
148
149 result = soup.find_all(True)
150 self.assertTrue(hasattr(result, "source"))
151
152 result = soup.find_all(text="foo")
153 self.assertTrue(hasattr(result, "source"))
154
155
156 class TestFindAllBasicNamespaces(TreeTest):
157
158 def test_find_by_namespaced_name(self):
159 soup = self.soup('<mathml:msqrt>4</mathml:msqrt><a svg:fill="red">')
160 self.assertEqual("4", soup.find("mathml:msqrt").string)
161 self.assertEqual("a", soup.find(attrs= { "svg:fill" : "red" }).name)
162
163
164 class TestFindAllByName(TreeTest):
165 """Test ways of finding tags by tag name."""
166
167 def setUp(self):
168 super(TreeTest, self).setUp()
169 self.tree = self.soup("""<a>First tag.</a>
170 <b>Second tag.</b>
171 <c>Third <a>Nested tag.</a> tag.</c>""")
172
173 def test_find_all_by_tag_name(self):
174 # Find all the <a> tags.
175 self.assertSelects(
176 self.tree.find_all('a'), ['First tag.', 'Nested tag.'])
177
178 def test_find_all_by_name_and_text(self):
179 self.assertSelects(
180 self.tree.find_all('a', text='First tag.'), ['First tag.'])
181
182 self.assertSelects(
183 self.tree.find_all('a', text=True), ['First tag.', 'Nested tag.'])
184
185 self.assertSelects(
186 self.tree.find_all('a', text=re.compile("tag")),
187 ['First tag.', 'Nested tag.'])
188
189
190 def test_find_all_on_non_root_element(self):
191 # You can call find_all on any node, not just the root.
192 self.assertSelects(self.tree.c.find_all('a'), ['Nested tag.'])
193
194 def test_calling_element_invokes_find_all(self):
195 self.assertSelects(self.tree('a'), ['First tag.', 'Nested tag.'])
196
197 def test_find_all_by_tag_strainer(self):
198 self.assertSelects(
199 self.tree.find_all(SoupStrainer('a')),
200 ['First tag.', 'Nested tag.'])
201
202 def test_find_all_by_tag_names(self):
203 self.assertSelects(
204 self.tree.find_all(['a', 'b']),
205 ['First tag.', 'Second tag.', 'Nested tag.'])
206
207 def test_find_all_by_tag_dict(self):
208 self.assertSelects(
209 self.tree.find_all({'a' : True, 'b' : True}),
210 ['First tag.', 'Second tag.', 'Nested tag.'])
211
212 def test_find_all_by_tag_re(self):
213 self.assertSelects(
214 self.tree.find_all(re.compile('^[ab]$')),
215 ['First tag.', 'Second tag.', 'Nested tag.'])
216
217 def test_find_all_with_tags_matching_method(self):
218 # You can define an oracle method that determines whether
219 # a tag matches the search.
220 def id_matches_name(tag):
221 return tag.name == tag.get('id')
222
223 tree = self.soup("""<a id="a">Match 1.</a>
224 <a id="1">Does not match.</a>
225 <b id="b">Match 2.</a>""")
226
227 self.assertSelects(
228 tree.find_all(id_matches_name), ["Match 1.", "Match 2."])
229
230 def test_find_with_multi_valued_attribute(self):
231 soup = self.soup(
232 "<div class='a b'>1</div><div class='a c'>2</div><div class='a d'>3</div>"
233 )
234 r1 = soup.find('div', 'a d');
235 r2 = soup.find('div', re.compile(r'a d'));
236 r3, r4 = soup.find_all('div', ['a b', 'a d']);
237 self.assertEqual('3', r1.string)
238 self.assertEqual('3', r2.string)
239 self.assertEqual('1', r3.string)
240 self.assertEqual('3', r4.string)
241
242
243 class TestFindAllByAttribute(TreeTest):
244
245 def test_find_all_by_attribute_name(self):
246 # You can pass in keyword arguments to find_all to search by
247 # attribute.
248 tree = self.soup("""
249 <a id="first">Matching a.</a>
250 <a id="second">
251 Non-matching <b id="first">Matching b.</b>a.
252 </a>""")
253 self.assertSelects(tree.find_all(id='first'),
254 ["Matching a.", "Matching b."])
255
256 def test_find_all_by_utf8_attribute_value(self):
257 peace = "םולש".encode("utf8")
258 data = '<a title="םולש"></a>'.encode("utf8")
259 soup = self.soup(data)
260 self.assertEqual([soup.a], soup.find_all(title=peace))
261 self.assertEqual([soup.a], soup.find_all(title=peace.decode("utf8")))
262 self.assertEqual([soup.a], soup.find_all(title=[peace, "something else"]))
263
264 def test_find_all_by_attribute_dict(self):
265 # You can pass in a dictionary as the argument 'attrs'. This
266 # lets you search for attributes like 'name' (a fixed argument
267 # to find_all) and 'class' (a reserved word in Python.)
268 tree = self.soup("""
269 <a name="name1" class="class1">Name match.</a>
270 <a name="name2" class="class2">Class match.</a>
271 <a name="name3" class="class3">Non-match.</a>
272 <name1>A tag called 'name1'.</name1>
273 """)
274
275 # This doesn't do what you want.
276 self.assertSelects(tree.find_all(name='name1'),
277 ["A tag called 'name1'."])
278 # This does what you want.
279 self.assertSelects(tree.find_all(attrs={'name' : 'name1'}),
280 ["Name match."])
281
282 self.assertSelects(tree.find_all(attrs={'class' : 'class2'}),
283 ["Class match."])
284
285 def test_find_all_by_class(self):
286 tree = self.soup("""
287 <a class="1">Class 1.</a>
288 <a class="2">Class 2.</a>
289 <b class="1">Class 1.</b>
290 <c class="3 4">Class 3 and 4.</c>
291 """)
292
293 # Passing in the class_ keyword argument will search against
294 # the 'class' attribute.
295 self.assertSelects(tree.find_all('a', class_='1'), ['Class 1.'])
296 self.assertSelects(tree.find_all('c', class_='3'), ['Class 3 and 4.'])
297 self.assertSelects(tree.find_all('c', class_='4'), ['Class 3 and 4.'])
298
299 # Passing in a string to 'attrs' will also search the CSS class.
300 self.assertSelects(tree.find_all('a', '1'), ['Class 1.'])
301 self.assertSelects(tree.find_all(attrs='1'), ['Class 1.', 'Class 1.'])
302 self.assertSelects(tree.find_all('c', '3'), ['Class 3 and 4.'])
303 self.assertSelects(tree.find_all('c', '4'), ['Class 3 and 4.'])
304
305 def test_find_by_class_when_multiple_classes_present(self):
306 tree = self.soup("<gar class='foo bar'>Found it</gar>")
307
308 f = tree.find_all("gar", class_=re.compile("o"))
309 self.assertSelects(f, ["Found it"])
310
311 f = tree.find_all("gar", class_=re.compile("a"))
312 self.assertSelects(f, ["Found it"])
313
314 # If the search fails to match the individual strings "foo" and "bar",
315 # it will be tried against the combined string "foo bar".
316 f = tree.find_all("gar", class_=re.compile("o b"))
317 self.assertSelects(f, ["Found it"])
318
319 def test_find_all_with_non_dictionary_for_attrs_finds_by_class(self):
320 soup = self.soup("<a class='bar'>Found it</a>")
321
322 self.assertSelects(soup.find_all("a", re.compile("ba")), ["Found it"])
323
324 def big_attribute_value(value):
325 return len(value) > 3
326
327 self.assertSelects(soup.find_all("a", big_attribute_value), [])
328
329 def small_attribute_value(value):
330 return len(value) <= 3
331
332 self.assertSelects(
333 soup.find_all("a", small_attribute_value), ["Found it"])
334
335 def test_find_all_with_string_for_attrs_finds_multiple_classes(self):
336 soup = self.soup('<a class="foo bar"></a><a class="foo"></a>')
337 a, a2 = soup.find_all("a")
338 self.assertEqual([a, a2], soup.find_all("a", "foo"))
339 self.assertEqual([a], soup.find_all("a", "bar"))
340
341 # If you specify the class as a string that contains a
342 # space, only that specific value will be found.
343 self.assertEqual([a], soup.find_all("a", class_="foo bar"))
344 self.assertEqual([a], soup.find_all("a", "foo bar"))
345 self.assertEqual([], soup.find_all("a", "bar foo"))
346
347 def test_find_all_by_attribute_soupstrainer(self):
348 tree = self.soup("""
349 <a id="first">Match.</a>
350 <a id="second">Non-match.</a>""")
351
352 strainer = SoupStrainer(attrs={'id' : 'first'})
353 self.assertSelects(tree.find_all(strainer), ['Match.'])
354
355 def test_find_all_with_missing_attribute(self):
356 # You can pass in None as the value of an attribute to find_all.
357 # This will match tags that do not have that attribute set.
358 tree = self.soup("""<a id="1">ID present.</a>
359 <a>No ID present.</a>
360 <a id="">ID is empty.</a>""")
361 self.assertSelects(tree.find_all('a', id=None), ["No ID present."])
362
363 def test_find_all_with_defined_attribute(self):
364 # You can pass in None as the value of an attribute to find_all.
365 # This will match tags that have that attribute set to any value.
366 tree = self.soup("""<a id="1">ID present.</a>
367 <a>No ID present.</a>
368 <a id="">ID is empty.</a>""")
369 self.assertSelects(
370 tree.find_all(id=True), ["ID present.", "ID is empty."])
371
372 def test_find_all_with_numeric_attribute(self):
373 # If you search for a number, it's treated as a string.
374 tree = self.soup("""<a id=1>Unquoted attribute.</a>
375 <a id="1">Quoted attribute.</a>""")
376
377 expected = ["Unquoted attribute.", "Quoted attribute."]
378 self.assertSelects(tree.find_all(id=1), expected)
379 self.assertSelects(tree.find_all(id="1"), expected)
380
381 def test_find_all_with_list_attribute_values(self):
382 # You can pass a list of attribute values instead of just one,
383 # and you'll get tags that match any of the values.
384 tree = self.soup("""<a id="1">1</a>
385 <a id="2">2</a>
386 <a id="3">3</a>
387 <a>No ID.</a>""")
388 self.assertSelects(tree.find_all(id=["1", "3", "4"]),
389 ["1", "3"])
390
391 def test_find_all_with_regular_expression_attribute_value(self):
392 # You can pass a regular expression as an attribute value, and
393 # you'll get tags whose values for that attribute match the
394 # regular expression.
395 tree = self.soup("""<a id="a">One a.</a>
396 <a id="aa">Two as.</a>
397 <a id="ab">Mixed as and bs.</a>
398 <a id="b">One b.</a>
399 <a>No ID.</a>""")
400
401 self.assertSelects(tree.find_all(id=re.compile("^a+$")),
402 ["One a.", "Two as."])
403
404 def test_find_by_name_and_containing_string(self):
405 soup = self.soup("<b>foo</b><b>bar</b><a>foo</a>")
406 a = soup.a
407
408 self.assertEqual([a], soup.find_all("a", text="foo"))
409 self.assertEqual([], soup.find_all("a", text="bar"))
410 self.assertEqual([], soup.find_all("a", text="bar"))
411
412 def test_find_by_name_and_containing_string_when_string_is_buried(self):
413 soup = self.soup("<a>foo</a><a><b><c>foo</c></b></a>")
414 self.assertEqual(soup.find_all("a"), soup.find_all("a", text="foo"))
415
416 def test_find_by_attribute_and_containing_string(self):
417 soup = self.soup('<b id="1">foo</b><a id="2">foo</a>')
418 a = soup.a
419
420 self.assertEqual([a], soup.find_all(id=2, text="foo"))
421 self.assertEqual([], soup.find_all(id=1, text="bar"))
422
423
424 class TestSmooth(TreeTest):
425 """Test Tag.smooth."""
426
427 def test_smooth(self):
428 soup = self.soup("<div>a</div>")
429 div = soup.div
430 div.append("b")
431 div.append("c")
432 div.append(Comment("Comment 1"))
433 div.append(Comment("Comment 2"))
434 div.append("d")
435 builder = self.default_builder()
436 span = Tag(soup, builder, 'span')
437 span.append('1')
438 span.append('2')
439 div.append(span)
440
441 # At this point the tree has a bunch of adjacent
442 # NavigableStrings. This is normal, but it has no meaning in
443 # terms of HTML, so we may want to smooth things out for
444 # output.
445
446 # Since the <span> tag has two children, its .string is None.
447 self.assertEqual(None, div.span.string)
448
449 self.assertEqual(7, len(div.contents))
450 div.smooth()
451 self.assertEqual(5, len(div.contents))
452
453 # The three strings at the beginning of div.contents have been
454 # merged into on string.
455 #
456 self.assertEqual('abc', div.contents[0])
457
458 # The call is recursive -- the <span> tag was also smoothed.
459 self.assertEqual('12', div.span.string)
460
461 # The two comments have _not_ been merged, even though
462 # comments are strings. Merging comments would change the
463 # meaning of the HTML.
464 self.assertEqual('Comment 1', div.contents[1])
465 self.assertEqual('Comment 2', div.contents[2])
466
467
468 class TestIndex(TreeTest):
469 """Test Tag.index"""
470 def test_index(self):
471 tree = self.soup("""<div>
472 <a>Identical</a>
473 <b>Not identical</b>
474 <a>Identical</a>
475
476 <c><d>Identical with child</d></c>
477 <b>Also not identical</b>
478 <c><d>Identical with child</d></c>
479 </div>""")
480 div = tree.div
481 for i, element in enumerate(div.contents):
482 self.assertEqual(i, div.index(element))
483 self.assertRaises(ValueError, tree.index, 1)
484
485
486 class TestParentOperations(TreeTest):
487 """Test navigation and searching through an element's parents."""
488
489 def setUp(self):
490 super(TestParentOperations, self).setUp()
491 self.tree = self.soup('''<ul id="empty"></ul>
492 <ul id="top">
493 <ul id="middle">
494 <ul id="bottom">
495 <b>Start here</b>
496 </ul>
497 </ul>''')
498 self.start = self.tree.b
499
500
501 def test_parent(self):
502 self.assertEqual(self.start.parent['id'], 'bottom')
503 self.assertEqual(self.start.parent.parent['id'], 'middle')
504 self.assertEqual(self.start.parent.parent.parent['id'], 'top')
505
506 def test_parent_of_top_tag_is_soup_object(self):
507 top_tag = self.tree.contents[0]
508 self.assertEqual(top_tag.parent, self.tree)
509
510 def test_soup_object_has_no_parent(self):
511 self.assertEqual(None, self.tree.parent)
512
513 def test_find_parents(self):
514 self.assertSelectsIDs(
515 self.start.find_parents('ul'), ['bottom', 'middle', 'top'])
516 self.assertSelectsIDs(
517 self.start.find_parents('ul', id="middle"), ['middle'])
518
519 def test_find_parent(self):
520 self.assertEqual(self.start.find_parent('ul')['id'], 'bottom')
521 self.assertEqual(self.start.find_parent('ul', id='top')['id'], 'top')
522
523 def test_parent_of_text_element(self):
524 text = self.tree.find(text="Start here")
525 self.assertEqual(text.parent.name, 'b')
526
527 def test_text_element_find_parent(self):
528 text = self.tree.find(text="Start here")
529 self.assertEqual(text.find_parent('ul')['id'], 'bottom')
530
531 def test_parent_generator(self):
532 parents = [parent['id'] for parent in self.start.parents
533 if parent is not None and 'id' in parent.attrs]
534 self.assertEqual(parents, ['bottom', 'middle', 'top'])
535
536
537 class ProximityTest(TreeTest):
538
539 def setUp(self):
540 super(TreeTest, self).setUp()
541 self.tree = self.soup(
542 '<html id="start"><head></head><body><b id="1">One</b><b id="2">Two</b><b id="3">Three</b></body></html>')
543
544
545 class TestNextOperations(ProximityTest):
546
547 def setUp(self):
548 super(TestNextOperations, self).setUp()
549 self.start = self.tree.b
550
551 def test_next(self):
552 self.assertEqual(self.start.next_element, "One")
553 self.assertEqual(self.start.next_element.next_element['id'], "2")
554
555 def test_next_of_last_item_is_none(self):
556 last = self.tree.find(text="Three")
557 self.assertEqual(last.next_element, None)
558
559 def test_next_of_root_is_none(self):
560 # The document root is outside the next/previous chain.
561 self.assertEqual(self.tree.next_element, None)
562
563 def test_find_all_next(self):
564 self.assertSelects(self.start.find_all_next('b'), ["Two", "Three"])
565 self.start.find_all_next(id=3)
566 self.assertSelects(self.start.find_all_next(id=3), ["Three"])
567
568 def test_find_next(self):
569 self.assertEqual(self.start.find_next('b')['id'], '2')
570 self.assertEqual(self.start.find_next(text="Three"), "Three")
571
572 def test_find_next_for_text_element(self):
573 text = self.tree.find(text="One")
574 self.assertEqual(text.find_next("b").string, "Two")
575 self.assertSelects(text.find_all_next("b"), ["Two", "Three"])
576
577 def test_next_generator(self):
578 start = self.tree.find(text="Two")
579 successors = [node for node in start.next_elements]
580 # There are two successors: the final <b> tag and its text contents.
581 tag, contents = successors
582 self.assertEqual(tag['id'], '3')
583 self.assertEqual(contents, "Three")
584
585 class TestPreviousOperations(ProximityTest):
586
587 def setUp(self):
588 super(TestPreviousOperations, self).setUp()
589 self.end = self.tree.find(text="Three")
590
591 def test_previous(self):
592 self.assertEqual(self.end.previous_element['id'], "3")
593 self.assertEqual(self.end.previous_element.previous_element, "Two")
594
595 def test_previous_of_first_item_is_none(self):
596 first = self.tree.find('html')
597 self.assertEqual(first.previous_element, None)
598
599 def test_previous_of_root_is_none(self):
600 # The document root is outside the next/previous chain.
601 # XXX This is broken!
602 #self.assertEqual(self.tree.previous_element, None)
603 pass
604
605 def test_find_all_previous(self):
606 # The <b> tag containing the "Three" node is the predecessor
607 # of the "Three" node itself, which is why "Three" shows up
608 # here.
609 self.assertSelects(
610 self.end.find_all_previous('b'), ["Three", "Two", "One"])
611 self.assertSelects(self.end.find_all_previous(id=1), ["One"])
612
613 def test_find_previous(self):
614 self.assertEqual(self.end.find_previous('b')['id'], '3')
615 self.assertEqual(self.end.find_previous(text="One"), "One")
616
617 def test_find_previous_for_text_element(self):
618 text = self.tree.find(text="Three")
619 self.assertEqual(text.find_previous("b").string, "Three")
620 self.assertSelects(
621 text.find_all_previous("b"), ["Three", "Two", "One"])
622
623 def test_previous_generator(self):
624 start = self.tree.find(text="One")
625 predecessors = [node for node in start.previous_elements]
626
627 # There are four predecessors: the <b> tag containing "One"
628 # the <body> tag, the <head> tag, and the <html> tag.
629 b, body, head, html = predecessors
630 self.assertEqual(b['id'], '1')
631 self.assertEqual(body.name, "body")
632 self.assertEqual(head.name, "head")
633 self.assertEqual(html.name, "html")
634
635
636 class SiblingTest(TreeTest):
637
638 def setUp(self):
639 super(SiblingTest, self).setUp()
640 markup = '''<html>
641 <span id="1">
642 <span id="1.1"></span>
643 </span>
644 <span id="2">
645 <span id="2.1"></span>
646 </span>
647 <span id="3">
648 <span id="3.1"></span>
649 </span>
650 <span id="4"></span>
651 </html>'''
652 # All that whitespace looks good but makes the tests more
653 # difficult. Get rid of it.
654 markup = re.compile(r"\n\s*").sub("", markup)
655 self.tree = self.soup(markup)
656
657
658 class TestNextSibling(SiblingTest):
659
660 def setUp(self):
661 super(TestNextSibling, self).setUp()
662 self.start = self.tree.find(id="1")
663
664 def test_next_sibling_of_root_is_none(self):
665 self.assertEqual(self.tree.next_sibling, None)
666
667 def test_next_sibling(self):
668 self.assertEqual(self.start.next_sibling['id'], '2')
669 self.assertEqual(self.start.next_sibling.next_sibling['id'], '3')
670
671 # Note the difference between next_sibling and next_element.
672 self.assertEqual(self.start.next_element['id'], '1.1')
673
674 def test_next_sibling_may_not_exist(self):
675 self.assertEqual(self.tree.html.next_sibling, None)
676
677 nested_span = self.tree.find(id="1.1")
678 self.assertEqual(nested_span.next_sibling, None)
679
680 last_span = self.tree.find(id="4")
681 self.assertEqual(last_span.next_sibling, None)
682
683 def test_find_next_sibling(self):
684 self.assertEqual(self.start.find_next_sibling('span')['id'], '2')
685
686 def test_next_siblings(self):
687 self.assertSelectsIDs(self.start.find_next_siblings("span"),
688 ['2', '3', '4'])
689
690 self.assertSelectsIDs(self.start.find_next_siblings(id='3'), ['3'])
691
692 def test_next_sibling_for_text_element(self):
693 soup = self.soup("Foo<b>bar</b>baz")
694 start = soup.find(text="Foo")
695 self.assertEqual(start.next_sibling.name, 'b')
696 self.assertEqual(start.next_sibling.next_sibling, 'baz')
697
698 self.assertSelects(start.find_next_siblings('b'), ['bar'])
699 self.assertEqual(start.find_next_sibling(text="baz"), "baz")
700 self.assertEqual(start.find_next_sibling(text="nonesuch"), None)
701
702
703 class TestPreviousSibling(SiblingTest):
704
705 def setUp(self):
706 super(TestPreviousSibling, self).setUp()
707 self.end = self.tree.find(id="4")
708
709 def test_previous_sibling_of_root_is_none(self):
710 self.assertEqual(self.tree.previous_sibling, None)
711
712 def test_previous_sibling(self):
713 self.assertEqual(self.end.previous_sibling['id'], '3')
714 self.assertEqual(self.end.previous_sibling.previous_sibling['id'], '2')
715
716 # Note the difference between previous_sibling and previous_element.
717 self.assertEqual(self.end.previous_element['id'], '3.1')
718
719 def test_previous_sibling_may_not_exist(self):
720 self.assertEqual(self.tree.html.previous_sibling, None)
721
722 nested_span = self.tree.find(id="1.1")
723 self.assertEqual(nested_span.previous_sibling, None)
724
725 first_span = self.tree.find(id="1")
726 self.assertEqual(first_span.previous_sibling, None)
727
728 def test_find_previous_sibling(self):
729 self.assertEqual(self.end.find_previous_sibling('span')['id'], '3')
730
731 def test_previous_siblings(self):
732 self.assertSelectsIDs(self.end.find_previous_siblings("span"),
733 ['3', '2', '1'])
734
735 self.assertSelectsIDs(self.end.find_previous_siblings(id='1'), ['1'])
736
737 def test_previous_sibling_for_text_element(self):
738 soup = self.soup("Foo<b>bar</b>baz")
739 start = soup.find(text="baz")
740 self.assertEqual(start.previous_sibling.name, 'b')
741 self.assertEqual(start.previous_sibling.previous_sibling, 'Foo')
742
743 self.assertSelects(start.find_previous_siblings('b'), ['bar'])
744 self.assertEqual(start.find_previous_sibling(text="Foo"), "Foo")
745 self.assertEqual(start.find_previous_sibling(text="nonesuch"), None)
746
747
748 class TestTag(SoupTest):
749
750 # Test various methods of Tag.
751
752 def test__should_pretty_print(self):
753 # Test the rules about when a tag should be pretty-printed.
754 tag = self.soup("").new_tag("a_tag")
755
756 # No list of whitespace-preserving tags -> pretty-print
757 tag._preserve_whitespace_tags = None
758 self.assertEqual(True, tag._should_pretty_print(0))
759
760 # List exists but tag is not on the list -> pretty-print
761 tag.preserve_whitespace_tags = ["some_other_tag"]
762 self.assertEqual(True, tag._should_pretty_print(1))
763
764 # Indent level is None -> don't pretty-print
765 self.assertEqual(False, tag._should_pretty_print(None))
766
767 # Tag is on the whitespace-preserving list -> don't pretty-print
768 tag.preserve_whitespace_tags = ["some_other_tag", "a_tag"]
769 self.assertEqual(False, tag._should_pretty_print(1))
770
771
772 class TestTagCreation(SoupTest):
773 """Test the ability to create new tags."""
774 def test_new_tag(self):
775 soup = self.soup("")
776 new_tag = soup.new_tag("foo", bar="baz", attrs={"name": "a name"})
777 self.assertTrue(isinstance(new_tag, Tag))
778 self.assertEqual("foo", new_tag.name)
779 self.assertEqual(dict(bar="baz", name="a name"), new_tag.attrs)
780 self.assertEqual(None, new_tag.parent)
781
782 def test_tag_inherits_self_closing_rules_from_builder(self):
783 if XML_BUILDER_PRESENT:
784 xml_soup = BeautifulSoup("", "lxml-xml")
785 xml_br = xml_soup.new_tag("br")
786 xml_p = xml_soup.new_tag("p")
787
788 # Both the <br> and <p> tag are empty-element, just because
789 # they have no contents.
790 self.assertEqual(b"<br/>", xml_br.encode())
791 self.assertEqual(b"<p/>", xml_p.encode())
792
793 html_soup = BeautifulSoup("", "html.parser")
794 html_br = html_soup.new_tag("br")
795 html_p = html_soup.new_tag("p")
796
797 # The HTML builder users HTML's rules about which tags are
798 # empty-element tags, and the new tags reflect these rules.
799 self.assertEqual(b"<br/>", html_br.encode())
800 self.assertEqual(b"<p></p>", html_p.encode())
801
802 def test_new_string_creates_navigablestring(self):
803 soup = self.soup("")
804 s = soup.new_string("foo")
805 self.assertEqual("foo", s)
806 self.assertTrue(isinstance(s, NavigableString))
807
808 def test_new_string_can_create_navigablestring_subclass(self):
809 soup = self.soup("")
810 s = soup.new_string("foo", Comment)
811 self.assertEqual("foo", s)
812 self.assertTrue(isinstance(s, Comment))
813
814 class TestTreeModification(SoupTest):
815
816 def test_attribute_modification(self):
817 soup = self.soup('<a id="1"></a>')
818 soup.a['id'] = 2
819 self.assertEqual(soup.decode(), self.document_for('<a id="2"></a>'))
820 del(soup.a['id'])
821 self.assertEqual(soup.decode(), self.document_for('<a></a>'))
822 soup.a['id2'] = 'foo'
823 self.assertEqual(soup.decode(), self.document_for('<a id2="foo"></a>'))
824
825 def test_new_tag_creation(self):
826 builder = builder_registry.lookup('html')()
827 soup = self.soup("<body></body>", builder=builder)
828 a = Tag(soup, builder, 'a')
829 ol = Tag(soup, builder, 'ol')
830 a['href'] = 'http://foo.com/'
831 soup.body.insert(0, a)
832 soup.body.insert(1, ol)
833 self.assertEqual(
834 soup.body.encode(),
835 b'<body><a href="http://foo.com/"></a><ol></ol></body>')
836
837 def test_append_to_contents_moves_tag(self):
838 doc = """<p id="1">Don't leave me <b>here</b>.</p>
839 <p id="2">Don\'t leave!</p>"""
840 soup = self.soup(doc)
841 second_para = soup.find(id='2')
842 bold = soup.b
843
844 # Move the <b> tag to the end of the second paragraph.
845 soup.find(id='2').append(soup.b)
846
847 # The <b> tag is now a child of the second paragraph.
848 self.assertEqual(bold.parent, second_para)
849
850 self.assertEqual(
851 soup.decode(), self.document_for(
852 '<p id="1">Don\'t leave me .</p>\n'
853 '<p id="2">Don\'t leave!<b>here</b></p>'))
854
855 def test_replace_with_returns_thing_that_was_replaced(self):
856 text = "<a></a><b><c></c></b>"
857 soup = self.soup(text)
858 a = soup.a
859 new_a = a.replace_with(soup.c)
860 self.assertEqual(a, new_a)
861
862 def test_unwrap_returns_thing_that_was_replaced(self):
863 text = "<a><b></b><c></c></a>"
864 soup = self.soup(text)
865 a = soup.a
866 new_a = a.unwrap()
867 self.assertEqual(a, new_a)
868
869 def test_replace_with_and_unwrap_give_useful_exception_when_tag_has_no_parent(self):
870 soup = self.soup("<a><b>Foo</b></a><c>Bar</c>")
871 a = soup.a
872 a.extract()
873 self.assertEqual(None, a.parent)
874 self.assertRaises(ValueError, a.unwrap)
875 self.assertRaises(ValueError, a.replace_with, soup.c)
876
877 def test_replace_tag_with_itself(self):
878 text = "<a><b></b><c>Foo<d></d></c></a><a><e></e></a>"
879 soup = self.soup(text)
880 c = soup.c
881 soup.c.replace_with(c)
882 self.assertEqual(soup.decode(), self.document_for(text))
883
884 def test_replace_tag_with_its_parent_raises_exception(self):
885 text = "<a><b></b></a>"
886 soup = self.soup(text)
887 self.assertRaises(ValueError, soup.b.replace_with, soup.a)
888
889 def test_insert_tag_into_itself_raises_exception(self):
890 text = "<a><b></b></a>"
891 soup = self.soup(text)
892 self.assertRaises(ValueError, soup.a.insert, 0, soup.a)
893
894 def test_insert_beautifulsoup_object_inserts_children(self):
895 """Inserting one BeautifulSoup object into another actually inserts all
896 of its children -- you'll never combine BeautifulSoup objects.
897 """
898 soup = self.soup("<p>And now, a word:</p><p>And we're back.</p>")
899
900 text = "<p>p2</p><p>p3</p>"
901 to_insert = self.soup(text)
902 soup.insert(1, to_insert)
903
904 for i in soup.descendants:
905 assert not isinstance(i, BeautifulSoup)
906
907 p1, p2, p3, p4 = list(soup.children)
908 self.assertEqual("And now, a word:", p1.string)
909 self.assertEqual("p2", p2.string)
910 self.assertEqual("p3", p3.string)
911 self.assertEqual("And we're back.", p4.string)
912
913
914 def test_replace_with_maintains_next_element_throughout(self):
915 soup = self.soup('<p><a>one</a><b>three</b></p>')
916 a = soup.a
917 b = a.contents[0]
918 # Make it so the <a> tag has two text children.
919 a.insert(1, "two")
920
921 # Now replace each one with the empty string.
922 left, right = a.contents
923 left.replaceWith('')
924 right.replaceWith('')
925
926 # The <b> tag is still connected to the tree.
927 self.assertEqual("three", soup.b.string)
928
929 def test_replace_final_node(self):
930 soup = self.soup("<b>Argh!</b>")
931 soup.find(text="Argh!").replace_with("Hooray!")
932 new_text = soup.find(text="Hooray!")
933 b = soup.b
934 self.assertEqual(new_text.previous_element, b)
935 self.assertEqual(new_text.parent, b)
936 self.assertEqual(new_text.previous_element.next_element, new_text)
937 self.assertEqual(new_text.next_element, None)
938
939 def test_consecutive_text_nodes(self):
940 # A builder should never create two consecutive text nodes,
941 # but if you insert one next to another, Beautiful Soup will
942 # handle it correctly.
943 soup = self.soup("<a><b>Argh!</b><c></c></a>")
944 soup.b.insert(1, "Hooray!")
945
946 self.assertEqual(
947 soup.decode(), self.document_for(
948 "<a><b>Argh!Hooray!</b><c></c></a>"))
949
950 new_text = soup.find(text="Hooray!")
951 self.assertEqual(new_text.previous_element, "Argh!")
952 self.assertEqual(new_text.previous_element.next_element, new_text)
953
954 self.assertEqual(new_text.previous_sibling, "Argh!")
955 self.assertEqual(new_text.previous_sibling.next_sibling, new_text)
956
957 self.assertEqual(new_text.next_sibling, None)
958 self.assertEqual(new_text.next_element, soup.c)
959
960 def test_insert_string(self):
961 soup = self.soup("<a></a>")
962 soup.a.insert(0, "bar")
963 soup.a.insert(0, "foo")
964 # The string were added to the tag.
965 self.assertEqual(["foo", "bar"], soup.a.contents)
966 # And they were converted to NavigableStrings.
967 self.assertEqual(soup.a.contents[0].next_element, "bar")
968
969 def test_insert_tag(self):
970 builder = self.default_builder()
971 soup = self.soup(
972 "<a><b>Find</b><c>lady!</c><d></d></a>", builder=builder)
973 magic_tag = Tag(soup, builder, 'magictag')
974 magic_tag.insert(0, "the")
975 soup.a.insert(1, magic_tag)
976
977 self.assertEqual(
978 soup.decode(), self.document_for(
979 "<a><b>Find</b><magictag>the</magictag><c>lady!</c><d></d></a>"))
980
981 # Make sure all the relationships are hooked up correctly.
982 b_tag = soup.b
983 self.assertEqual(b_tag.next_sibling, magic_tag)
984 self.assertEqual(magic_tag.previous_sibling, b_tag)
985
986 find = b_tag.find(text="Find")
987 self.assertEqual(find.next_element, magic_tag)
988 self.assertEqual(magic_tag.previous_element, find)
989
990 c_tag = soup.c
991 self.assertEqual(magic_tag.next_sibling, c_tag)
992 self.assertEqual(c_tag.previous_sibling, magic_tag)
993
994 the = magic_tag.find(text="the")
995 self.assertEqual(the.parent, magic_tag)
996 self.assertEqual(the.next_element, c_tag)
997 self.assertEqual(c_tag.previous_element, the)
998
999 def test_append_child_thats_already_at_the_end(self):
1000 data = "<a><b></b></a>"
1001 soup = self.soup(data)
1002 soup.a.append(soup.b)
1003 self.assertEqual(data, soup.decode())
1004
1005 def test_extend(self):
1006 data = "<a><b><c><d><e><f><g></g></f></e></d></c></b></a>"
1007 soup = self.soup(data)
1008 l = [soup.g, soup.f, soup.e, soup.d, soup.c, soup.b]
1009 soup.a.extend(l)
1010 self.assertEqual("<a><g></g><f></f><e></e><d></d><c></c><b></b></a>", soup.decode())
1011
1012 def test_move_tag_to_beginning_of_parent(self):
1013 data = "<a><b></b><c></c><d></d></a>"
1014 soup = self.soup(data)
1015 soup.a.insert(0, soup.d)
1016 self.assertEqual("<a><d></d><b></b><c></c></a>", soup.decode())
1017
1018 def test_insert_works_on_empty_element_tag(self):
1019 # This is a little strange, since most HTML parsers don't allow
1020 # markup like this to come through. But in general, we don't
1021 # know what the parser would or wouldn't have allowed, so
1022 # I'm letting this succeed for now.
1023 soup = self.soup("<br/>")
1024 soup.br.insert(1, "Contents")
1025 self.assertEqual(str(soup.br), "<br>Contents</br>")
1026
1027 def test_insert_before(self):
1028 soup = self.soup("<a>foo</a><b>bar</b>")
1029 soup.b.insert_before("BAZ")
1030 soup.a.insert_before("QUUX")
1031 self.assertEqual(
1032 soup.decode(), self.document_for("QUUX<a>foo</a>BAZ<b>bar</b>"))
1033
1034 soup.a.insert_before(soup.b)
1035 self.assertEqual(
1036 soup.decode(), self.document_for("QUUX<b>bar</b><a>foo</a>BAZ"))
1037
1038 # Can't insert an element before itself.
1039 b = soup.b
1040 self.assertRaises(ValueError, b.insert_before, b)
1041
1042 # Can't insert before if an element has no parent.
1043 b.extract()
1044 self.assertRaises(ValueError, b.insert_before, "nope")
1045
1046 # Can insert an identical element
1047 soup = self.soup("<a>")
1048 soup.a.insert_before(soup.new_tag("a"))
1049
1050 def test_insert_multiple_before(self):
1051 soup = self.soup("<a>foo</a><b>bar</b>")
1052 soup.b.insert_before("BAZ", " ", "QUUX")
1053 soup.a.insert_before("QUUX", " ", "BAZ")
1054 self.assertEqual(
1055 soup.decode(), self.document_for("QUUX BAZ<a>foo</a>BAZ QUUX<b>bar</b>"))
1056
1057 soup.a.insert_before(soup.b, "FOO")
1058 self.assertEqual(
1059 soup.decode(), self.document_for("QUUX BAZ<b>bar</b>FOO<a>foo</a>BAZ QUUX"))
1060
1061 def test_insert_after(self):
1062 soup = self.soup("<a>foo</a><b>bar</b>")
1063 soup.b.insert_after("BAZ")
1064 soup.a.insert_after("QUUX")
1065 self.assertEqual(
1066 soup.decode(), self.document_for("<a>foo</a>QUUX<b>bar</b>BAZ"))
1067 soup.b.insert_after(soup.a)
1068 self.assertEqual(
1069 soup.decode(), self.document_for("QUUX<b>bar</b><a>foo</a>BAZ"))
1070
1071 # Can't insert an element after itself.
1072 b = soup.b
1073 self.assertRaises(ValueError, b.insert_after, b)
1074
1075 # Can't insert after if an element has no parent.
1076 b.extract()
1077 self.assertRaises(ValueError, b.insert_after, "nope")
1078
1079 # Can insert an identical element
1080 soup = self.soup("<a>")
1081 soup.a.insert_before(soup.new_tag("a"))
1082
1083 def test_insert_multiple_after(self):
1084 soup = self.soup("<a>foo</a><b>bar</b>")
1085 soup.b.insert_after("BAZ", " ", "QUUX")
1086 soup.a.insert_after("QUUX", " ", "BAZ")
1087 self.assertEqual(
1088 soup.decode(), self.document_for("<a>foo</a>QUUX BAZ<b>bar</b>BAZ QUUX"))
1089 soup.b.insert_after(soup.a, "FOO ")
1090 self.assertEqual(
1091 soup.decode(), self.document_for("QUUX BAZ<b>bar</b><a>foo</a>FOO BAZ QUUX"))
1092
1093 def test_insert_after_raises_exception_if_after_has_no_meaning(self):
1094 soup = self.soup("")
1095 tag = soup.new_tag("a")
1096 string = soup.new_string("")
1097 self.assertRaises(ValueError, string.insert_after, tag)
1098 self.assertRaises(NotImplementedError, soup.insert_after, tag)
1099 self.assertRaises(ValueError, tag.insert_after, tag)
1100
1101 def test_insert_before_raises_notimplementederror_if_before_has_no_meaning(self):
1102 soup = self.soup("")
1103 tag = soup.new_tag("a")
1104 string = soup.new_string("")
1105 self.assertRaises(ValueError, string.insert_before, tag)
1106 self.assertRaises(NotImplementedError, soup.insert_before, tag)
1107 self.assertRaises(ValueError, tag.insert_before, tag)
1108
1109 def test_replace_with(self):
1110 soup = self.soup(
1111 "<p>There's <b>no</b> business like <b>show</b> business</p>")
1112 no, show = soup.find_all('b')
1113 show.replace_with(no)
1114 self.assertEqual(
1115 soup.decode(),
1116 self.document_for(
1117 "<p>There's business like <b>no</b> business</p>"))
1118
1119 self.assertEqual(show.parent, None)
1120 self.assertEqual(no.parent, soup.p)
1121 self.assertEqual(no.next_element, "no")
1122 self.assertEqual(no.next_sibling, " business")
1123
1124 def test_replace_first_child(self):
1125 data = "<a><b></b><c></c></a>"
1126 soup = self.soup(data)
1127 soup.b.replace_with(soup.c)
1128 self.assertEqual("<a><c></c></a>", soup.decode())
1129
1130 def test_replace_last_child(self):
1131 data = "<a><b></b><c></c></a>"
1132 soup = self.soup(data)
1133 soup.c.replace_with(soup.b)
1134 self.assertEqual("<a><b></b></a>", soup.decode())
1135
1136 def test_nested_tag_replace_with(self):
1137 soup = self.soup(
1138 """<a>We<b>reserve<c>the</c><d>right</d></b></a><e>to<f>refuse</f><g>service</g></e>""")
1139
1140 # Replace the entire <b> tag and its contents ("reserve the
1141 # right") with the <f> tag ("refuse").
1142 remove_tag = soup.b
1143 move_tag = soup.f
1144 remove_tag.replace_with(move_tag)
1145
1146 self.assertEqual(
1147 soup.decode(), self.document_for(
1148 "<a>We<f>refuse</f></a><e>to<g>service</g></e>"))
1149
1150 # The <b> tag is now an orphan.
1151 self.assertEqual(remove_tag.parent, None)
1152 self.assertEqual(remove_tag.find(text="right").next_element, None)
1153 self.assertEqual(remove_tag.previous_element, None)
1154 self.assertEqual(remove_tag.next_sibling, None)
1155 self.assertEqual(remove_tag.previous_sibling, None)
1156
1157 # The <f> tag is now connected to the <a> tag.
1158 self.assertEqual(move_tag.parent, soup.a)
1159 self.assertEqual(move_tag.previous_element, "We")
1160 self.assertEqual(move_tag.next_element.next_element, soup.e)
1161 self.assertEqual(move_tag.next_sibling, None)
1162
1163 # The gap where the <f> tag used to be has been mended, and
1164 # the word "to" is now connected to the <g> tag.
1165 to_text = soup.find(text="to")
1166 g_tag = soup.g
1167 self.assertEqual(to_text.next_element, g_tag)
1168 self.assertEqual(to_text.next_sibling, g_tag)
1169 self.assertEqual(g_tag.previous_element, to_text)
1170 self.assertEqual(g_tag.previous_sibling, to_text)
1171
1172 def test_unwrap(self):
1173 tree = self.soup("""
1174 <p>Unneeded <em>formatting</em> is unneeded</p>
1175 """)
1176 tree.em.unwrap()
1177 self.assertEqual(tree.em, None)
1178 self.assertEqual(tree.p.text, "Unneeded formatting is unneeded")
1179
1180 def test_wrap(self):
1181 soup = self.soup("I wish I was bold.")
1182 value = soup.string.wrap(soup.new_tag("b"))
1183 self.assertEqual(value.decode(), "<b>I wish I was bold.</b>")
1184 self.assertEqual(
1185 soup.decode(), self.document_for("<b>I wish I was bold.</b>"))
1186
1187 def test_wrap_extracts_tag_from_elsewhere(self):
1188 soup = self.soup("<b></b>I wish I was bold.")
1189 soup.b.next_sibling.wrap(soup.b)
1190 self.assertEqual(
1191 soup.decode(), self.document_for("<b>I wish I was bold.</b>"))
1192
1193 def test_wrap_puts_new_contents_at_the_end(self):
1194 soup = self.soup("<b>I like being bold.</b>I wish I was bold.")
1195 soup.b.next_sibling.wrap(soup.b)
1196 self.assertEqual(2, len(soup.b.contents))
1197 self.assertEqual(
1198 soup.decode(), self.document_for(
1199 "<b>I like being bold.I wish I was bold.</b>"))
1200
1201 def test_extract(self):
1202 soup = self.soup(
1203 '<html><body>Some content. <div id="nav">Nav crap</div> More content.</body></html>')
1204
1205 self.assertEqual(len(soup.body.contents), 3)
1206 extracted = soup.find(id="nav").extract()
1207
1208 self.assertEqual(
1209 soup.decode(), "<html><body>Some content. More content.</body></html>")
1210 self.assertEqual(extracted.decode(), '<div id="nav">Nav crap</div>')
1211
1212 # The extracted tag is now an orphan.
1213 self.assertEqual(len(soup.body.contents), 2)
1214 self.assertEqual(extracted.parent, None)
1215 self.assertEqual(extracted.previous_element, None)
1216 self.assertEqual(extracted.next_element.next_element, None)
1217
1218 # The gap where the extracted tag used to be has been mended.
1219 content_1 = soup.find(text="Some content. ")
1220 content_2 = soup.find(text=" More content.")
1221 self.assertEqual(content_1.next_element, content_2)
1222 self.assertEqual(content_1.next_sibling, content_2)
1223 self.assertEqual(content_2.previous_element, content_1)
1224 self.assertEqual(content_2.previous_sibling, content_1)
1225
1226 def test_extract_distinguishes_between_identical_strings(self):
1227 soup = self.soup("<a>foo</a><b>bar</b>")
1228 foo_1 = soup.a.string
1229 bar_1 = soup.b.string
1230 foo_2 = soup.new_string("foo")
1231 bar_2 = soup.new_string("bar")
1232 soup.a.append(foo_2)
1233 soup.b.append(bar_2)
1234
1235 # Now there are two identical strings in the <a> tag, and two
1236 # in the <b> tag. Let's remove the first "foo" and the second
1237 # "bar".
1238 foo_1.extract()
1239 bar_2.extract()
1240 self.assertEqual(foo_2, soup.a.string)
1241 self.assertEqual(bar_2, soup.b.string)
1242
1243 def test_extract_multiples_of_same_tag(self):
1244 soup = self.soup("""
1245 <html>
1246 <head>
1247 <script>foo</script>
1248 </head>
1249 <body>
1250 <script>bar</script>
1251 <a></a>
1252 </body>
1253 <script>baz</script>
1254 </html>""")
1255 [soup.script.extract() for i in soup.find_all("script")]
1256 self.assertEqual("<body>\n\n<a></a>\n</body>", str(soup.body))
1257
1258
1259 def test_extract_works_when_element_is_surrounded_by_identical_strings(self):
1260 soup = self.soup(
1261 '<html>\n'
1262 '<body>hi</body>\n'
1263 '</html>')
1264 soup.find('body').extract()
1265 self.assertEqual(None, soup.find('body'))
1266
1267
1268 def test_clear(self):
1269 """Tag.clear()"""
1270 soup = self.soup("<p><a>String <em>Italicized</em></a> and another</p>")
1271 # clear using extract()
1272 a = soup.a
1273 soup.p.clear()
1274 self.assertEqual(len(soup.p.contents), 0)
1275 self.assertTrue(hasattr(a, "contents"))
1276
1277 # clear using decompose()
1278 em = a.em
1279 a.clear(decompose=True)
1280 self.assertEqual(0, len(em.contents))
1281
1282
1283 def test_decompose(self):
1284 # Test PageElement.decompose() and PageElement.decomposed
1285 soup = self.soup("<p><a>String <em>Italicized</em></a></p><p>Another para</p>")
1286 p1, p2 = soup.find_all('p')
1287 a = p1.a
1288 text = p1.em.string
1289 for i in [p1, p2, a, text]:
1290 self.assertEqual(False, i.decomposed)
1291
1292 # This sets p1 and everything beneath it to decomposed.
1293 p1.decompose()
1294 for i in [p1, a, text]:
1295 self.assertEqual(True, i.decomposed)
1296 # p2 is unaffected.
1297 self.assertEqual(False, p2.decomposed)
1298
1299 def test_string_set(self):
1300 """Tag.string = 'string'"""
1301 soup = self.soup("<a></a> <b><c></c></b>")
1302 soup.a.string = "foo"
1303 self.assertEqual(soup.a.contents, ["foo"])
1304 soup.b.string = "bar"
1305 self.assertEqual(soup.b.contents, ["bar"])
1306
1307 def test_string_set_does_not_affect_original_string(self):
1308 soup = self.soup("<a><b>foo</b><c>bar</c>")
1309 soup.b.string = soup.c.string
1310 self.assertEqual(soup.a.encode(), b"<a><b>bar</b><c>bar</c></a>")
1311
1312 def test_set_string_preserves_class_of_string(self):
1313 soup = self.soup("<a></a>")
1314 cdata = CData("foo")
1315 soup.a.string = cdata
1316 self.assertTrue(isinstance(soup.a.string, CData))
1317
1318 class TestElementObjects(SoupTest):
1319 """Test various features of element objects."""
1320
1321 def test_len(self):
1322 """The length of an element is its number of children."""
1323 soup = self.soup("<top>1<b>2</b>3</top>")
1324
1325 # The BeautifulSoup object itself contains one element: the
1326 # <top> tag.
1327 self.assertEqual(len(soup.contents), 1)
1328 self.assertEqual(len(soup), 1)
1329
1330 # The <top> tag contains three elements: the text node "1", the
1331 # <b> tag, and the text node "3".
1332 self.assertEqual(len(soup.top), 3)
1333 self.assertEqual(len(soup.top.contents), 3)
1334
1335 def test_member_access_invokes_find(self):
1336 """Accessing a Python member .foo invokes find('foo')"""
1337 soup = self.soup('<b><i></i></b>')
1338 self.assertEqual(soup.b, soup.find('b'))
1339 self.assertEqual(soup.b.i, soup.find('b').find('i'))
1340 self.assertEqual(soup.a, None)
1341
1342 def test_deprecated_member_access(self):
1343 soup = self.soup('<b><i></i></b>')
1344 with warnings.catch_warnings(record=True) as w:
1345 tag = soup.bTag
1346 self.assertEqual(soup.b, tag)
1347 self.assertEqual(
1348 '.bTag is deprecated, use .find("b") instead. If you really were looking for a tag called bTag, use .find("bTag")',
1349 str(w[0].message))
1350
1351 def test_has_attr(self):
1352 """has_attr() checks for the presence of an attribute.
1353
1354 Please note note: has_attr() is different from
1355 __in__. has_attr() checks the tag's attributes and __in__
1356 checks the tag's chidlren.
1357 """
1358 soup = self.soup("<foo attr='bar'>")
1359 self.assertTrue(soup.foo.has_attr('attr'))
1360 self.assertFalse(soup.foo.has_attr('attr2'))
1361
1362
1363 def test_attributes_come_out_in_alphabetical_order(self):
1364 markup = '<b a="1" z="5" m="3" f="2" y="4"></b>'
1365 self.assertSoupEquals(markup, '<b a="1" f="2" m="3" y="4" z="5"></b>')
1366
1367 def test_string(self):
1368 # A tag that contains only a text node makes that node
1369 # available as .string.
1370 soup = self.soup("<b>foo</b>")
1371 self.assertEqual(soup.b.string, 'foo')
1372
1373 def test_empty_tag_has_no_string(self):
1374 # A tag with no children has no .stirng.
1375 soup = self.soup("<b></b>")
1376 self.assertEqual(soup.b.string, None)
1377
1378 def test_tag_with_multiple_children_has_no_string(self):
1379 # A tag with no children has no .string.
1380 soup = self.soup("<a>foo<b></b><b></b></b>")
1381 self.assertEqual(soup.b.string, None)
1382
1383 soup = self.soup("<a>foo<b></b>bar</b>")
1384 self.assertEqual(soup.b.string, None)
1385
1386 # Even if all the children are strings, due to trickery,
1387 # it won't work--but this would be a good optimization.
1388 soup = self.soup("<a>foo</b>")
1389 soup.a.insert(1, "bar")
1390 self.assertEqual(soup.a.string, None)
1391
1392 def test_tag_with_recursive_string_has_string(self):
1393 # A tag with a single child which has a .string inherits that
1394 # .string.
1395 soup = self.soup("<a><b>foo</b></a>")
1396 self.assertEqual(soup.a.string, "foo")
1397 self.assertEqual(soup.string, "foo")
1398
1399 def test_lack_of_string(self):
1400 """Only a tag containing a single text node has a .string."""
1401 soup = self.soup("<b>f<i>e</i>o</b>")
1402 self.assertFalse(soup.b.string)
1403
1404 soup = self.soup("<b></b>")
1405 self.assertFalse(soup.b.string)
1406
1407 def test_all_text(self):
1408 """Tag.text and Tag.get_text(sep=u"") -> all child text, concatenated"""
1409 soup = self.soup("<a>a<b>r</b> <r> t </r></a>")
1410 self.assertEqual(soup.a.text, "ar t ")
1411 self.assertEqual(soup.a.get_text(strip=True), "art")
1412 self.assertEqual(soup.a.get_text(","), "a,r, , t ")
1413 self.assertEqual(soup.a.get_text(",", strip=True), "a,r,t")
1414
1415 def test_get_text_ignores_special_string_containers(self):
1416 soup = self.soup("foo<!--IGNORE-->bar")
1417 self.assertEqual(soup.get_text(), "foobar")
1418
1419 self.assertEqual(
1420 soup.get_text(types=(NavigableString, Comment)), "fooIGNOREbar")
1421 self.assertEqual(
1422 soup.get_text(types=None), "fooIGNOREbar")
1423
1424 soup = self.soup("foo<style>CSS</style><script>Javascript</script>bar")
1425 self.assertEqual(soup.get_text(), "foobar")
1426
1427 def test_all_strings_ignores_special_string_containers(self):
1428 soup = self.soup("foo<!--IGNORE-->bar")
1429 self.assertEqual(['foo', 'bar'], list(soup.strings))
1430
1431 soup = self.soup("foo<style>CSS</style><script>Javascript</script>bar")
1432 self.assertEqual(['foo', 'bar'], list(soup.strings))
1433
1434
1435 class TestCDAtaListAttributes(SoupTest):
1436
1437 """Testing cdata-list attributes like 'class'.
1438 """
1439 def test_single_value_becomes_list(self):
1440 soup = self.soup("<a class='foo'>")
1441 self.assertEqual(["foo"],soup.a['class'])
1442
1443 def test_multiple_values_becomes_list(self):
1444 soup = self.soup("<a class='foo bar'>")
1445 self.assertEqual(["foo", "bar"], soup.a['class'])
1446
1447 def test_multiple_values_separated_by_weird_whitespace(self):
1448 soup = self.soup("<a class='foo\tbar\nbaz'>")
1449 self.assertEqual(["foo", "bar", "baz"],soup.a['class'])
1450
1451 def test_attributes_joined_into_string_on_output(self):
1452 soup = self.soup("<a class='foo\tbar'>")
1453 self.assertEqual(b'<a class="foo bar"></a>', soup.a.encode())
1454
1455 def test_get_attribute_list(self):
1456 soup = self.soup("<a id='abc def'>")
1457 self.assertEqual(['abc def'], soup.a.get_attribute_list('id'))
1458
1459 def test_accept_charset(self):
1460 soup = self.soup('<form accept-charset="ISO-8859-1 UTF-8">')
1461 self.assertEqual(['ISO-8859-1', 'UTF-8'], soup.form['accept-charset'])
1462
1463 def test_cdata_attribute_applying_only_to_one_tag(self):
1464 data = '<a accept-charset="ISO-8859-1 UTF-8"></a>'
1465 soup = self.soup(data)
1466 # We saw in another test that accept-charset is a cdata-list
1467 # attribute for the <form> tag. But it's not a cdata-list
1468 # attribute for any other tag.
1469 self.assertEqual('ISO-8859-1 UTF-8', soup.a['accept-charset'])
1470
1471 def test_string_has_immutable_name_property(self):
1472 string = self.soup("s").string
1473 self.assertEqual(None, string.name)
1474 def t():
1475 string.name = 'foo'
1476 self.assertRaises(AttributeError, t)
1477
1478 class TestPersistence(SoupTest):
1479 "Testing features like pickle and deepcopy."
1480
1481 def setUp(self):
1482 super(TestPersistence, self).setUp()
1483 self.page = """<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"
1484 "http://www.w3.org/TR/REC-html40/transitional.dtd">
1485 <html>
1486 <head>
1487 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
1488 <title>Beautiful Soup: We called him Tortoise because he taught us.</title>
1489 <link rev="made" href="mailto:leonardr@segfault.org">
1490 <meta name="Description" content="Beautiful Soup: an HTML parser optimized for screen-scraping.">
1491 <meta name="generator" content="Markov Approximation 1.4 (module: leonardr)">
1492 <meta name="author" content="Leonard Richardson">
1493 </head>
1494 <body>
1495 <a href="foo">foo</a>
1496 <a href="foo"><b>bar</b></a>
1497 </body>
1498 </html>"""
1499 self.tree = self.soup(self.page)
1500
1501 def test_pickle_and_unpickle_identity(self):
1502 # Pickling a tree, then unpickling it, yields a tree identical
1503 # to the original.
1504 dumped = pickle.dumps(self.tree, 2)
1505 loaded = pickle.loads(dumped)
1506 self.assertEqual(loaded.__class__, BeautifulSoup)
1507 self.assertEqual(loaded.decode(), self.tree.decode())
1508
1509 def test_deepcopy_identity(self):
1510 # Making a deepcopy of a tree yields an identical tree.
1511 copied = copy.deepcopy(self.tree)
1512 self.assertEqual(copied.decode(), self.tree.decode())
1513
1514 def test_copy_preserves_encoding(self):
1515 soup = BeautifulSoup(b'<p>&nbsp;</p>', 'html.parser')
1516 encoding = soup.original_encoding
1517 copy = soup.__copy__()
1518 self.assertEqual("<p> </p>", str(copy))
1519 self.assertEqual(encoding, copy.original_encoding)
1520
1521 def test_copy_preserves_builder_information(self):
1522
1523 tag = self.soup('<p></p>').p
1524
1525 # Simulate a tag obtained from a source file.
1526 tag.sourceline = 10
1527 tag.sourcepos = 33
1528
1529 copied = tag.__copy__()
1530
1531 # The TreeBuilder object is no longer availble, but information
1532 # obtained from it gets copied over to the new Tag object.
1533 self.assertEqual(tag.sourceline, copied.sourceline)
1534 self.assertEqual(tag.sourcepos, copied.sourcepos)
1535 self.assertEqual(
1536 tag.can_be_empty_element, copied.can_be_empty_element
1537 )
1538 self.assertEqual(
1539 tag.cdata_list_attributes, copied.cdata_list_attributes
1540 )
1541 self.assertEqual(
1542 tag.preserve_whitespace_tags, copied.preserve_whitespace_tags
1543 )
1544
1545
1546 def test_unicode_pickle(self):
1547 # A tree containing Unicode characters can be pickled.
1548 html = "<b>\N{SNOWMAN}</b>"
1549 soup = self.soup(html)
1550 dumped = pickle.dumps(soup, pickle.HIGHEST_PROTOCOL)
1551 loaded = pickle.loads(dumped)
1552 self.assertEqual(loaded.decode(), soup.decode())
1553
1554 def test_copy_navigablestring_is_not_attached_to_tree(self):
1555 html = "<b>Foo<a></a></b><b>Bar</b>"
1556 soup = self.soup(html)
1557 s1 = soup.find(string="Foo")
1558 s2 = copy.copy(s1)
1559 self.assertEqual(s1, s2)
1560 self.assertEqual(None, s2.parent)
1561 self.assertEqual(None, s2.next_element)
1562 self.assertNotEqual(None, s1.next_sibling)
1563 self.assertEqual(None, s2.next_sibling)
1564 self.assertEqual(None, s2.previous_element)
1565
1566 def test_copy_navigablestring_subclass_has_same_type(self):
1567 html = "<b><!--Foo--></b>"
1568 soup = self.soup(html)
1569 s1 = soup.string
1570 s2 = copy.copy(s1)
1571 self.assertEqual(s1, s2)
1572 self.assertTrue(isinstance(s2, Comment))
1573
1574 def test_copy_entire_soup(self):
1575 html = "<div><b>Foo<a></a></b><b>Bar</b></div>end"
1576 soup = self.soup(html)
1577 soup_copy = copy.copy(soup)
1578 self.assertEqual(soup, soup_copy)
1579
1580 def test_copy_tag_copies_contents(self):
1581 html = "<div><b>Foo<a></a></b><b>Bar</b></div>end"
1582 soup = self.soup(html)
1583 div = soup.div
1584 div_copy = copy.copy(div)
1585
1586 # The two tags look the same, and evaluate to equal.
1587 self.assertEqual(str(div), str(div_copy))
1588 self.assertEqual(div, div_copy)
1589
1590 # But they're not the same object.
1591 self.assertFalse(div is div_copy)
1592
1593 # And they don't have the same relation to the parse tree. The
1594 # copy is not associated with a parse tree at all.
1595 self.assertEqual(None, div_copy.parent)
1596 self.assertEqual(None, div_copy.previous_element)
1597 self.assertEqual(None, div_copy.find(string='Bar').next_element)
1598 self.assertNotEqual(None, div.find(string='Bar').next_element)
1599
1600 class TestSubstitutions(SoupTest):
1601
1602 def test_default_formatter_is_minimal(self):
1603 markup = "<b>&lt;&lt;Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!&gt;&gt;</b>"
1604 soup = self.soup(markup)
1605 decoded = soup.decode(formatter="minimal")
1606 # The < is converted back into &lt; but the e-with-acute is left alone.
1607 self.assertEqual(
1608 decoded,
1609 self.document_for(
1610 "<b>&lt;&lt;Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!&gt;&gt;</b>"))
1611
1612 def test_formatter_html(self):
1613 markup = "<br><b>&lt;&lt;Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!&gt;&gt;</b>"
1614 soup = self.soup(markup)
1615 decoded = soup.decode(formatter="html")
1616 self.assertEqual(
1617 decoded,
1618 self.document_for("<br/><b>&lt;&lt;Sacr&eacute; bleu!&gt;&gt;</b>"))
1619
1620 def test_formatter_html5(self):
1621 markup = "<br><b>&lt;&lt;Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!&gt;&gt;</b>"
1622 soup = self.soup(markup)
1623 decoded = soup.decode(formatter="html5")
1624 self.assertEqual(
1625 decoded,
1626 self.document_for("<br><b>&lt;&lt;Sacr&eacute; bleu!&gt;&gt;</b>"))
1627
1628 def test_formatter_minimal(self):
1629 markup = "<b>&lt;&lt;Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!&gt;&gt;</b>"
1630 soup = self.soup(markup)
1631 decoded = soup.decode(formatter="minimal")
1632 # The < is converted back into &lt; but the e-with-acute is left alone.
1633 self.assertEqual(
1634 decoded,
1635 self.document_for(
1636 "<b>&lt;&lt;Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!&gt;&gt;</b>"))
1637
1638 def test_formatter_null(self):
1639 markup = "<b>&lt;&lt;Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!&gt;&gt;</b>"
1640 soup = self.soup(markup)
1641 decoded = soup.decode(formatter=None)
1642 # Neither the angle brackets nor the e-with-acute are converted.
1643 # This is not valid HTML, but it's what the user wanted.
1644 self.assertEqual(decoded,
1645 self.document_for("<b><<Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!>></b>"))
1646
1647 def test_formatter_custom(self):
1648 markup = "<b>&lt;foo&gt;</b><b>bar</b><br/>"
1649 soup = self.soup(markup)
1650 decoded = soup.decode(formatter = lambda x: x.upper())
1651 # Instead of normal entity conversion code, the custom
1652 # callable is called on every string.
1653 self.assertEqual(
1654 decoded,
1655 self.document_for("<b><FOO></b><b>BAR</b><br/>"))
1656
1657 def test_formatter_is_run_on_attribute_values(self):
1658 markup = '<a href="http://a.com?a=b&c=é">e</a>'
1659 soup = self.soup(markup)
1660 a = soup.a
1661
1662 expect_minimal = '<a href="http://a.com?a=b&amp;c=é">e</a>'
1663
1664 self.assertEqual(expect_minimal, a.decode())
1665 self.assertEqual(expect_minimal, a.decode(formatter="minimal"))
1666
1667 expect_html = '<a href="http://a.com?a=b&amp;c=&eacute;">e</a>'
1668 self.assertEqual(expect_html, a.decode(formatter="html"))
1669
1670 self.assertEqual(markup, a.decode(formatter=None))
1671 expect_upper = '<a href="HTTP://A.COM?A=B&C=É">E</a>'
1672 self.assertEqual(expect_upper, a.decode(formatter=lambda x: x.upper()))
1673
1674 def test_formatter_skips_script_tag_for_html_documents(self):
1675 doc = """
1676 <script type="text/javascript">
1677 console.log("< < hey > > ");
1678 </script>
1679 """
1680 encoded = BeautifulSoup(doc, 'html.parser').encode()
1681 self.assertTrue(b"< < hey > >" in encoded)
1682
1683 def test_formatter_skips_style_tag_for_html_documents(self):
1684 doc = """
1685 <style type="text/css">
1686 console.log("< < hey > > ");
1687 </style>
1688 """
1689 encoded = BeautifulSoup(doc, 'html.parser').encode()
1690 self.assertTrue(b"< < hey > >" in encoded)
1691
1692 def test_prettify_leaves_preformatted_text_alone(self):
1693 soup = self.soup("<div> foo <pre> \tbar\n \n </pre> baz <textarea> eee\nfff\t</textarea></div>")
1694 # Everything outside the <pre> tag is reformatted, but everything
1695 # inside is left alone.
1696 self.assertEqual(
1697 '<div>\n foo\n <pre> \tbar\n \n </pre>\n baz\n <textarea> eee\nfff\t</textarea>\n</div>',
1698 soup.div.prettify())
1699
1700 def test_prettify_accepts_formatter_function(self):
1701 soup = BeautifulSoup("<html><body>foo</body></html>", 'html.parser')
1702 pretty = soup.prettify(formatter = lambda x: x.upper())
1703 self.assertTrue("FOO" in pretty)
1704
1705 def test_prettify_outputs_unicode_by_default(self):
1706 soup = self.soup("<a></a>")
1707 self.assertEqual(str, type(soup.prettify()))
1708
1709 def test_prettify_can_encode_data(self):
1710 soup = self.soup("<a></a>")
1711 self.assertEqual(bytes, type(soup.prettify("utf-8")))
1712
1713 def test_html_entity_substitution_off_by_default(self):
1714 markup = "<b>Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!</b>"
1715 soup = self.soup(markup)
1716 encoded = soup.b.encode("utf-8")
1717 self.assertEqual(encoded, markup.encode('utf-8'))
1718
1719 def test_encoding_substitution(self):
1720 # Here's the <meta> tag saying that a document is
1721 # encoded in Shift-JIS.
1722 meta_tag = ('<meta content="text/html; charset=x-sjis" '
1723 'http-equiv="Content-type"/>')
1724 soup = self.soup(meta_tag)
1725
1726 # Parse the document, and the charset apprears unchanged.
1727 self.assertEqual(soup.meta['content'], 'text/html; charset=x-sjis')
1728
1729 # Encode the document into some encoding, and the encoding is
1730 # substituted into the meta tag.
1731 utf_8 = soup.encode("utf-8")
1732 self.assertTrue(b"charset=utf-8" in utf_8)
1733
1734 euc_jp = soup.encode("euc_jp")
1735 self.assertTrue(b"charset=euc_jp" in euc_jp)
1736
1737 shift_jis = soup.encode("shift-jis")
1738 self.assertTrue(b"charset=shift-jis" in shift_jis)
1739
1740 utf_16_u = soup.encode("utf-16").decode("utf-16")
1741 self.assertTrue("charset=utf-16" in utf_16_u)
1742
1743 def test_encoding_substitution_doesnt_happen_if_tag_is_strained(self):
1744 markup = ('<head><meta content="text/html; charset=x-sjis" '
1745 'http-equiv="Content-type"/></head><pre>foo</pre>')
1746
1747 # Beautiful Soup used to try to rewrite the meta tag even if the
1748 # meta tag got filtered out by the strainer. This test makes
1749 # sure that doesn't happen.
1750 strainer = SoupStrainer('pre')
1751 soup = self.soup(markup, parse_only=strainer)
1752 self.assertEqual(soup.contents[0].name, 'pre')
1753
1754 class TestEncoding(SoupTest):
1755 """Test the ability to encode objects into strings."""
1756
1757 def test_unicode_string_can_be_encoded(self):
1758 html = "<b>\N{SNOWMAN}</b>"
1759 soup = self.soup(html)
1760 self.assertEqual(soup.b.string.encode("utf-8"),
1761 "\N{SNOWMAN}".encode("utf-8"))
1762
1763 def test_tag_containing_unicode_string_can_be_encoded(self):
1764 html = "<b>\N{SNOWMAN}</b>"
1765 soup = self.soup(html)
1766 self.assertEqual(
1767 soup.b.encode("utf-8"), html.encode("utf-8"))
1768
1769 def test_encoding_substitutes_unrecognized_characters_by_default(self):
1770 html = "<b>\N{SNOWMAN}</b>"
1771 soup = self.soup(html)
1772 self.assertEqual(soup.b.encode("ascii"), b"<b>&#9731;</b>")
1773
1774 def test_encoding_can_be_made_strict(self):
1775 html = "<b>\N{SNOWMAN}</b>"
1776 soup = self.soup(html)
1777 self.assertRaises(
1778 UnicodeEncodeError, soup.encode, "ascii", errors="strict")
1779
1780 def test_decode_contents(self):
1781 html = "<b>\N{SNOWMAN}</b>"
1782 soup = self.soup(html)
1783 self.assertEqual("\N{SNOWMAN}", soup.b.decode_contents())
1784
1785 def test_encode_contents(self):
1786 html = "<b>\N{SNOWMAN}</b>"
1787 soup = self.soup(html)
1788 self.assertEqual(
1789 "\N{SNOWMAN}".encode("utf8"), soup.b.encode_contents(
1790 encoding="utf8"))
1791
1792 def test_deprecated_renderContents(self):
1793 html = "<b>\N{SNOWMAN}</b>"
1794 soup = self.soup(html)
1795 self.assertEqual(
1796 "\N{SNOWMAN}".encode("utf8"), soup.b.renderContents())
1797
1798 def test_repr(self):
1799 html = "<b>\N{SNOWMAN}</b>"
1800 soup = self.soup(html)
1801 if PY3K:
1802 self.assertEqual(html, repr(soup))
1803 else:
1804 self.assertEqual(b'<b>\\u2603</b>', repr(soup))
1805
1806 class TestFormatter(SoupTest):
1807
1808 def test_default_attributes(self):
1809 # Test the default behavior of Formatter.attributes().
1810 formatter = Formatter()
1811 tag = Tag(name="tag")
1812 tag['b'] = 1
1813 tag['a'] = 2
1814
1815 # Attributes come out sorted by name. In Python 3, attributes
1816 # normally come out of a dictionary in the order they were
1817 # added.
1818 self.assertEqual([('a', 2), ('b', 1)], formatter.attributes(tag))
1819
1820 # This works even if Tag.attrs is None, though this shouldn't
1821 # normally happen.
1822 tag.attrs = None
1823 self.assertEqual([], formatter.attributes(tag))
1824
1825 def test_sort_attributes(self):
1826 # Test the ability to override Formatter.attributes() to,
1827 # e.g., disable the normal sorting of attributes.
1828 class UnsortedFormatter(Formatter):
1829 def attributes(self, tag):
1830 self.called_with = tag
1831 for k, v in sorted(tag.attrs.items()):
1832 if k == 'ignore':
1833 continue
1834 yield k,v
1835
1836 soup = self.soup('<p cval="1" aval="2" ignore="ignored"></p>')
1837 formatter = UnsortedFormatter()
1838 decoded = soup.decode(formatter=formatter)
1839
1840 # attributes() was called on the <p> tag. It filtered out one
1841 # attribute and sorted the other two.
1842 self.assertEqual(formatter.called_with, soup.p)
1843 self.assertEqual('<p aval="2" cval="1"></p>', decoded)
1844
1845
1846 class TestNavigableStringSubclasses(SoupTest):
1847
1848 def test_cdata(self):
1849 # None of the current builders turn CDATA sections into CData
1850 # objects, but you can create them manually.
1851 soup = self.soup("")
1852 cdata = CData("foo")
1853 soup.insert(1, cdata)
1854 self.assertEqual(str(soup), "<![CDATA[foo]]>")
1855 self.assertEqual(soup.find(text="foo"), "foo")
1856 self.assertEqual(soup.contents[0], "foo")
1857
1858 def test_cdata_is_never_formatted(self):
1859 """Text inside a CData object is passed into the formatter.
1860
1861 But the return value is ignored.
1862 """
1863
1864 self.count = 0
1865 def increment(*args):
1866 self.count += 1
1867 return "BITTER FAILURE"
1868
1869 soup = self.soup("")
1870 cdata = CData("<><><>")
1871 soup.insert(1, cdata)
1872 self.assertEqual(
1873 b"<![CDATA[<><><>]]>", soup.encode(formatter=increment))
1874 self.assertEqual(1, self.count)
1875
1876 def test_doctype_ends_in_newline(self):
1877 # Unlike other NavigableString subclasses, a DOCTYPE always ends
1878 # in a newline.
1879 doctype = Doctype("foo")
1880 soup = self.soup("")
1881 soup.insert(1, doctype)
1882 self.assertEqual(soup.encode(), b"<!DOCTYPE foo>\n")
1883
1884 def test_declaration(self):
1885 d = Declaration("foo")
1886 self.assertEqual("<?foo?>", d.output_ready())
1887
1888 def test_default_string_containers(self):
1889 # In some cases, we use different NavigableString subclasses for
1890 # the same text in different tags.
1891 soup = self.soup(
1892 "<div>text</div><script>text</script><style>text</style>"
1893 )
1894 self.assertEqual(
1895 [NavigableString, Script, Stylesheet],
1896 [x.__class__ for x in soup.find_all(text=True)]
1897 )
1898
1899 # The TemplateString is a little unusual because it's generally found
1900 # _inside_ children of a <template> element, not a direct child of the
1901 # <template> element.
1902 soup = self.soup(
1903 "<template>Some text<p>In a tag</p></template>Some text outside"
1904 )
1905 assert all(isinstance(x, TemplateString) for x in soup.template.strings)
1906
1907 # Once the <template> tag closed, we went back to using
1908 # NavigableString.
1909 outside = soup.template.next_sibling
1910 assert isinstance(outside, NavigableString)
1911 assert not isinstance(outside, TemplateString)
1912
1913 class TestSoupSelector(TreeTest):
1914
1915 HTML = """
1916 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
1917 "http://www.w3.org/TR/html4/strict.dtd">
1918 <html>
1919 <head>
1920 <title>The title</title>
1921 <link rel="stylesheet" href="blah.css" type="text/css" id="l1">
1922 </head>
1923 <body>
1924 <custom-dashed-tag class="dashed" id="dash1">Hello there.</custom-dashed-tag>
1925 <div id="main" class="fancy">
1926 <div id="inner">
1927 <h1 id="header1">An H1</h1>
1928 <p>Some text</p>
1929 <p class="onep" id="p1">Some more text</p>
1930 <h2 id="header2">An H2</h2>
1931 <p class="class1 class2 class3" id="pmulti">Another</p>
1932 <a href="http://bob.example.org/" rel="friend met" id="bob">Bob</a>
1933 <h2 id="header3">Another H2</h2>
1934 <a id="me" href="http://simonwillison.net/" rel="me">me</a>
1935 <span class="s1">
1936 <a href="#" id="s1a1">span1a1</a>
1937 <a href="#" id="s1a2">span1a2 <span id="s1a2s1">test</span></a>
1938 <span class="span2">
1939 <a href="#" id="s2a1">span2a1</a>
1940 </span>
1941 <span class="span3"></span>
1942 <custom-dashed-tag class="dashed" id="dash2"/>
1943 <div data-tag="dashedvalue" id="data1"/>
1944 </span>
1945 </div>
1946 <x id="xid">
1947 <z id="zida"/>
1948 <z id="zidab"/>
1949 <z id="zidac"/>
1950 </x>
1951 <y id="yid">
1952 <z id="zidb"/>
1953 </y>
1954 <p lang="en" id="lang-en">English</p>
1955 <p lang="en-gb" id="lang-en-gb">English UK</p>
1956 <p lang="en-us" id="lang-en-us">English US</p>
1957 <p lang="fr" id="lang-fr">French</p>
1958 </div>
1959
1960 <div id="footer">
1961 </div>
1962 """
1963
1964 def setUp(self):
1965 self.soup = BeautifulSoup(self.HTML, 'html.parser')
1966
1967 def assertSelects(self, selector, expected_ids, **kwargs):
1968 el_ids = [el['id'] for el in self.soup.select(selector, **kwargs)]
1969 el_ids.sort()
1970 expected_ids.sort()
1971 self.assertEqual(expected_ids, el_ids,
1972 "Selector %s, expected [%s], got [%s]" % (
1973 selector, ', '.join(expected_ids), ', '.join(el_ids)
1974 )
1975 )
1976
1977 assertSelect = assertSelects
1978
1979 def assertSelectMultiple(self, *tests):
1980 for selector, expected_ids in tests:
1981 self.assertSelect(selector, expected_ids)
1982
1983 def test_one_tag_one(self):
1984 els = self.soup.select('title')
1985 self.assertEqual(len(els), 1)
1986 self.assertEqual(els[0].name, 'title')
1987 self.assertEqual(els[0].contents, ['The title'])
1988
1989 def test_one_tag_many(self):
1990 els = self.soup.select('div')
1991 self.assertEqual(len(els), 4)
1992 for div in els:
1993 self.assertEqual(div.name, 'div')
1994
1995 el = self.soup.select_one('div')
1996 self.assertEqual('main', el['id'])
1997
1998 def test_select_one_returns_none_if_no_match(self):
1999 match = self.soup.select_one('nonexistenttag')
2000 self.assertEqual(None, match)
2001
2002
2003 def test_tag_in_tag_one(self):
2004 els = self.soup.select('div div')
2005 self.assertSelects('div div', ['inner', 'data1'])
2006
2007 def test_tag_in_tag_many(self):
2008 for selector in ('html div', 'html body div', 'body div'):
2009 self.assertSelects(selector, ['data1', 'main', 'inner', 'footer'])
2010
2011
2012 def test_limit(self):
2013 self.assertSelects('html div', ['main'], limit=1)
2014 self.assertSelects('html body div', ['inner', 'main'], limit=2)
2015 self.assertSelects('body div', ['data1', 'main', 'inner', 'footer'],
2016 limit=10)
2017
2018 def test_tag_no_match(self):
2019 self.assertEqual(len(self.soup.select('del')), 0)
2020
2021 def test_invalid_tag(self):
2022 self.assertRaises(SelectorSyntaxError, self.soup.select, 'tag%t')
2023
2024 def test_select_dashed_tag_ids(self):
2025 self.assertSelects('custom-dashed-tag', ['dash1', 'dash2'])
2026
2027 def test_select_dashed_by_id(self):
2028 dashed = self.soup.select('custom-dashed-tag[id=\"dash2\"]')
2029 self.assertEqual(dashed[0].name, 'custom-dashed-tag')
2030 self.assertEqual(dashed[0]['id'], 'dash2')
2031
2032 def test_dashed_tag_text(self):
2033 self.assertEqual(self.soup.select('body > custom-dashed-tag')[0].text, 'Hello there.')
2034
2035 def test_select_dashed_matches_find_all(self):
2036 self.assertEqual(self.soup.select('custom-dashed-tag'), self.soup.find_all('custom-dashed-tag'))
2037
2038 def test_header_tags(self):
2039 self.assertSelectMultiple(
2040 ('h1', ['header1']),
2041 ('h2', ['header2', 'header3']),
2042 )
2043
2044 def test_class_one(self):
2045 for selector in ('.onep', 'p.onep', 'html p.onep'):
2046 els = self.soup.select(selector)
2047 self.assertEqual(len(els), 1)
2048 self.assertEqual(els[0].name, 'p')
2049 self.assertEqual(els[0]['class'], ['onep'])
2050
2051 def test_class_mismatched_tag(self):
2052 els = self.soup.select('div.onep')
2053 self.assertEqual(len(els), 0)
2054
2055 def test_one_id(self):
2056 for selector in ('div#inner', '#inner', 'div div#inner'):
2057 self.assertSelects(selector, ['inner'])
2058
2059 def test_bad_id(self):
2060 els = self.soup.select('#doesnotexist')
2061 self.assertEqual(len(els), 0)
2062
2063 def test_items_in_id(self):
2064 els = self.soup.select('div#inner p')
2065 self.assertEqual(len(els), 3)
2066 for el in els:
2067 self.assertEqual(el.name, 'p')
2068 self.assertEqual(els[1]['class'], ['onep'])
2069 self.assertFalse(els[0].has_attr('class'))
2070
2071 def test_a_bunch_of_emptys(self):
2072 for selector in ('div#main del', 'div#main div.oops', 'div div#main'):
2073 self.assertEqual(len(self.soup.select(selector)), 0)
2074
2075 def test_multi_class_support(self):
2076 for selector in ('.class1', 'p.class1', '.class2', 'p.class2',
2077 '.class3', 'p.class3', 'html p.class2', 'div#inner .class2'):
2078 self.assertSelects(selector, ['pmulti'])
2079
2080 def test_multi_class_selection(self):
2081 for selector in ('.class1.class3', '.class3.class2',
2082 '.class1.class2.class3'):
2083 self.assertSelects(selector, ['pmulti'])
2084
2085 def test_child_selector(self):
2086 self.assertSelects('.s1 > a', ['s1a1', 's1a2'])
2087 self.assertSelects('.s1 > a span', ['s1a2s1'])
2088
2089 def test_child_selector_id(self):
2090 self.assertSelects('.s1 > a#s1a2 span', ['s1a2s1'])
2091
2092 def test_attribute_equals(self):
2093 self.assertSelectMultiple(
2094 ('p[class="onep"]', ['p1']),
2095 ('p[id="p1"]', ['p1']),
2096 ('[class="onep"]', ['p1']),
2097 ('[id="p1"]', ['p1']),
2098 ('link[rel="stylesheet"]', ['l1']),
2099 ('link[type="text/css"]', ['l1']),
2100 ('link[href="blah.css"]', ['l1']),
2101 ('link[href="no-blah.css"]', []),
2102 ('[rel="stylesheet"]', ['l1']),
2103 ('[type="text/css"]', ['l1']),
2104 ('[href="blah.css"]', ['l1']),
2105 ('[href="no-blah.css"]', []),
2106 ('p[href="no-blah.css"]', []),
2107 ('[href="no-blah.css"]', []),
2108 )
2109
2110 def test_attribute_tilde(self):
2111 self.assertSelectMultiple(
2112 ('p[class~="class1"]', ['pmulti']),
2113 ('p[class~="class2"]', ['pmulti']),
2114 ('p[class~="class3"]', ['pmulti']),
2115 ('[class~="class1"]', ['pmulti']),
2116 ('[class~="class2"]', ['pmulti']),
2117 ('[class~="class3"]', ['pmulti']),
2118 ('a[rel~="friend"]', ['bob']),
2119 ('a[rel~="met"]', ['bob']),
2120 ('[rel~="friend"]', ['bob']),
2121 ('[rel~="met"]', ['bob']),
2122 )
2123
2124 def test_attribute_startswith(self):
2125 self.assertSelectMultiple(
2126 ('[rel^="style"]', ['l1']),
2127 ('link[rel^="style"]', ['l1']),
2128 ('notlink[rel^="notstyle"]', []),
2129 ('[rel^="notstyle"]', []),
2130 ('link[rel^="notstyle"]', []),
2131 ('link[href^="bla"]', ['l1']),
2132 ('a[href^="http://"]', ['bob', 'me']),
2133 ('[href^="http://"]', ['bob', 'me']),
2134 ('[id^="p"]', ['pmulti', 'p1']),
2135 ('[id^="m"]', ['me', 'main']),
2136 ('div[id^="m"]', ['main']),
2137 ('a[id^="m"]', ['me']),
2138 ('div[data-tag^="dashed"]', ['data1'])
2139 )
2140
2141 def test_attribute_endswith(self):
2142 self.assertSelectMultiple(
2143 ('[href$=".css"]', ['l1']),
2144 ('link[href$=".css"]', ['l1']),
2145 ('link[id$="1"]', ['l1']),
2146 ('[id$="1"]', ['data1', 'l1', 'p1', 'header1', 's1a1', 's2a1', 's1a2s1', 'dash1']),
2147 ('div[id$="1"]', ['data1']),
2148 ('[id$="noending"]', []),
2149 )
2150
2151 def test_attribute_contains(self):
2152 self.assertSelectMultiple(
2153 # From test_attribute_startswith
2154 ('[rel*="style"]', ['l1']),
2155 ('link[rel*="style"]', ['l1']),
2156 ('notlink[rel*="notstyle"]', []),
2157 ('[rel*="notstyle"]', []),
2158 ('link[rel*="notstyle"]', []),
2159 ('link[href*="bla"]', ['l1']),
2160 ('[href*="http://"]', ['bob', 'me']),
2161 ('[id*="p"]', ['pmulti', 'p1']),
2162 ('div[id*="m"]', ['main']),
2163 ('a[id*="m"]', ['me']),
2164 # From test_attribute_endswith
2165 ('[href*=".css"]', ['l1']),
2166 ('link[href*=".css"]', ['l1']),
2167 ('link[id*="1"]', ['l1']),
2168 ('[id*="1"]', ['data1', 'l1', 'p1', 'header1', 's1a1', 's1a2', 's2a1', 's1a2s1', 'dash1']),
2169 ('div[id*="1"]', ['data1']),
2170 ('[id*="noending"]', []),
2171 # New for this test
2172 ('[href*="."]', ['bob', 'me', 'l1']),
2173 ('a[href*="."]', ['bob', 'me']),
2174 ('link[href*="."]', ['l1']),
2175 ('div[id*="n"]', ['main', 'inner']),
2176 ('div[id*="nn"]', ['inner']),
2177 ('div[data-tag*="edval"]', ['data1'])
2178 )
2179
2180 def test_attribute_exact_or_hypen(self):
2181 self.assertSelectMultiple(
2182 ('p[lang|="en"]', ['lang-en', 'lang-en-gb', 'lang-en-us']),
2183 ('[lang|="en"]', ['lang-en', 'lang-en-gb', 'lang-en-us']),
2184 ('p[lang|="fr"]', ['lang-fr']),
2185 ('p[lang|="gb"]', []),
2186 )
2187
2188 def test_attribute_exists(self):
2189 self.assertSelectMultiple(
2190 ('[rel]', ['l1', 'bob', 'me']),
2191 ('link[rel]', ['l1']),
2192 ('a[rel]', ['bob', 'me']),
2193 ('[lang]', ['lang-en', 'lang-en-gb', 'lang-en-us', 'lang-fr']),
2194 ('p[class]', ['p1', 'pmulti']),
2195 ('[blah]', []),
2196 ('p[blah]', []),
2197 ('div[data-tag]', ['data1'])
2198 )
2199
2200 def test_quoted_space_in_selector_name(self):
2201 html = """<div style="display: wrong">nope</div>
2202 <div style="display: right">yes</div>
2203 """
2204 soup = BeautifulSoup(html, 'html.parser')
2205 [chosen] = soup.select('div[style="display: right"]')
2206 self.assertEqual("yes", chosen.string)
2207
2208 def test_unsupported_pseudoclass(self):
2209 self.assertRaises(
2210 NotImplementedError, self.soup.select, "a:no-such-pseudoclass")
2211
2212 self.assertRaises(
2213 SelectorSyntaxError, self.soup.select, "a:nth-of-type(a)")
2214
2215 def test_nth_of_type(self):
2216 # Try to select first paragraph
2217 els = self.soup.select('div#inner p:nth-of-type(1)')
2218 self.assertEqual(len(els), 1)
2219 self.assertEqual(els[0].string, 'Some text')
2220
2221 # Try to select third paragraph
2222 els = self.soup.select('div#inner p:nth-of-type(3)')
2223 self.assertEqual(len(els), 1)
2224 self.assertEqual(els[0].string, 'Another')
2225
2226 # Try to select (non-existent!) fourth paragraph
2227 els = self.soup.select('div#inner p:nth-of-type(4)')
2228 self.assertEqual(len(els), 0)
2229
2230 # Zero will select no tags.
2231 els = self.soup.select('div p:nth-of-type(0)')
2232 self.assertEqual(len(els), 0)
2233
2234 def test_nth_of_type_direct_descendant(self):
2235 els = self.soup.select('div#inner > p:nth-of-type(1)')
2236 self.assertEqual(len(els), 1)
2237 self.assertEqual(els[0].string, 'Some text')
2238
2239 def test_id_child_selector_nth_of_type(self):
2240 self.assertSelects('#inner > p:nth-of-type(2)', ['p1'])
2241
2242 def test_select_on_element(self):
2243 # Other tests operate on the tree; this operates on an element
2244 # within the tree.
2245 inner = self.soup.find("div", id="main")
2246 selected = inner.select("div")
2247 # The <div id="inner"> tag was selected. The <div id="footer">
2248 # tag was not.
2249 self.assertSelectsIDs(selected, ['inner', 'data1'])
2250
2251 def test_overspecified_child_id(self):
2252 self.assertSelects(".fancy #inner", ['inner'])
2253 self.assertSelects(".normal #inner", [])
2254
2255 def test_adjacent_sibling_selector(self):
2256 self.assertSelects('#p1 + h2', ['header2'])
2257 self.assertSelects('#p1 + h2 + p', ['pmulti'])
2258 self.assertSelects('#p1 + #header2 + .class1', ['pmulti'])
2259 self.assertEqual([], self.soup.select('#p1 + p'))
2260
2261 def test_general_sibling_selector(self):
2262 self.assertSelects('#p1 ~ h2', ['header2', 'header3'])
2263 self.assertSelects('#p1 ~ #header2', ['header2'])
2264 self.assertSelects('#p1 ~ h2 + a', ['me'])
2265 self.assertSelects('#p1 ~ h2 + [rel="me"]', ['me'])
2266 self.assertEqual([], self.soup.select('#inner ~ h2'))
2267
2268 def test_dangling_combinator(self):
2269 self.assertRaises(SelectorSyntaxError, self.soup.select, 'h1 >')
2270
2271 def test_sibling_combinator_wont_select_same_tag_twice(self):
2272 self.assertSelects('p[lang] ~ p', ['lang-en-gb', 'lang-en-us', 'lang-fr'])
2273
2274 # Test the selector grouping operator (the comma)
2275 def test_multiple_select(self):
2276 self.assertSelects('x, y', ['xid', 'yid'])
2277
2278 def test_multiple_select_with_no_space(self):
2279 self.assertSelects('x,y', ['xid', 'yid'])
2280
2281 def test_multiple_select_with_more_space(self):
2282 self.assertSelects('x, y', ['xid', 'yid'])
2283
2284 def test_multiple_select_duplicated(self):
2285 self.assertSelects('x, x', ['xid'])
2286
2287 def test_multiple_select_sibling(self):
2288 self.assertSelects('x, y ~ p[lang=fr]', ['xid', 'lang-fr'])
2289
2290 def test_multiple_select_tag_and_direct_descendant(self):
2291 self.assertSelects('x, y > z', ['xid', 'zidb'])
2292
2293 def test_multiple_select_direct_descendant_and_tags(self):
2294 self.assertSelects('div > x, y, z', ['xid', 'yid', 'zida', 'zidb', 'zidab', 'zidac'])
2295
2296 def test_multiple_select_indirect_descendant(self):
2297 self.assertSelects('div x,y, z', ['xid', 'yid', 'zida', 'zidb', 'zidab', 'zidac'])
2298
2299 def test_invalid_multiple_select(self):
2300 self.assertRaises(SelectorSyntaxError, self.soup.select, ',x, y')
2301 self.assertRaises(SelectorSyntaxError, self.soup.select, 'x,,y')
2302
2303 def test_multiple_select_attrs(self):
2304 self.assertSelects('p[lang=en], p[lang=en-gb]', ['lang-en', 'lang-en-gb'])
2305
2306 def test_multiple_select_ids(self):
2307 self.assertSelects('x, y > z[id=zida], z[id=zidab], z[id=zidb]', ['xid', 'zidb', 'zidab'])
2308
2309 def test_multiple_select_nested(self):
2310 self.assertSelects('body > div > x, y > z', ['xid', 'zidb'])
2311
2312 def test_select_duplicate_elements(self):
2313 # When markup contains duplicate elements, a multiple select
2314 # will find all of them.
2315 markup = '<div class="c1"/><div class="c2"/><div class="c1"/>'
2316 soup = BeautifulSoup(markup, 'html.parser')
2317 selected = soup.select(".c1, .c2")
2318 self.assertEqual(3, len(selected))
2319
2320 # Verify that find_all finds the same elements, though because
2321 # of an implementation detail it finds them in a different
2322 # order.
2323 for element in soup.find_all(class_=['c1', 'c2']):
2324 assert element in selected