github/plexpy
1
0
Fork 0
mirror of https://github.com/Tautulli/Tautulli.git synced 2025-08-20 13:23:24 -07:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

Update bs4 to 4.8.1 (with 2to3)

Browse source
This commit is contained in:
JonnyWong16 2019-11-23 18:54:24 -08:00
commit f28e741ad7
19 changed files with 5487 additions and 792 deletions
Download patch file Download diff file Expand all files Collapse all files

1
lib/bs4/tests/__init__.py Normal file
View file

@ -0,0 +1 @@
"The beautifulsoup tests."

147
lib/bs4/tests/test_builder_registry.py Normal file
View file

@ -0,0 +1,147 @@
"""Tests of the builder registry."""
import unittest
import warnings
from bs4 import BeautifulSoup
from bs4.builder import (
builder_registry as registry,
HTMLParserTreeBuilder,
TreeBuilderRegistry,
)
try:
from bs4.builder import HTML5TreeBuilder
HTML5LIB_PRESENT = True
except ImportError:
HTML5LIB_PRESENT = False
try:
from bs4.builder import (
LXMLTreeBuilderForXML,
LXMLTreeBuilder,
)
LXML_PRESENT = True
except ImportError:
LXML_PRESENT = False
class BuiltInRegistryTest(unittest.TestCase):
"""Test the built-in registry with the default builders registered."""
def test_combination(self):
if LXML_PRESENT:
self.assertEqual(registry.lookup('fast', 'html'),
LXMLTreeBuilder)
if LXML_PRESENT:
self.assertEqual(registry.lookup('permissive', 'xml'),
LXMLTreeBuilderForXML)
self.assertEqual(registry.lookup('strict', 'html'),
HTMLParserTreeBuilder)
if HTML5LIB_PRESENT:
self.assertEqual(registry.lookup('html5lib', 'html'),
HTML5TreeBuilder)
def test_lookup_by_markup_type(self):
if LXML_PRESENT:
self.assertEqual(registry.lookup('html'), LXMLTreeBuilder)
self.assertEqual(registry.lookup('xml'), LXMLTreeBuilderForXML)
else:
self.assertEqual(registry.lookup('xml'), None)
if HTML5LIB_PRESENT:
self.assertEqual(registry.lookup('html'), HTML5TreeBuilder)
else:
self.assertEqual(registry.lookup('html'), HTMLParserTreeBuilder)
def test_named_library(self):
if LXML_PRESENT:
self.assertEqual(registry.lookup('lxml', 'xml'),
LXMLTreeBuilderForXML)
self.assertEqual(registry.lookup('lxml', 'html'),
LXMLTreeBuilder)
if HTML5LIB_PRESENT:
self.assertEqual(registry.lookup('html5lib'),
HTML5TreeBuilder)
self.assertEqual(registry.lookup('html.parser'),
HTMLParserTreeBuilder)
def test_beautifulsoup_constructor_does_lookup(self):
with warnings.catch_warnings(record=True) as w:
# This will create a warning about not explicitly
# specifying a parser, but we'll ignore it.
# You can pass in a string.
BeautifulSoup("", features="html")
# Or a list of strings.
BeautifulSoup("", features=["html", "fast"])
# You'll get an exception if BS can't find an appropriate
# builder.
self.assertRaises(ValueError, BeautifulSoup,
"", features="no-such-feature")
class RegistryTest(unittest.TestCase):
"""Test the TreeBuilderRegistry class in general."""
def setUp(self):
self.registry = TreeBuilderRegistry()
def builder_for_features(self, *feature_list):
cls = type('Builder_' + '_'.join(feature_list),
(object,), {'features' : feature_list})
self.registry.register(cls)
return cls
def test_register_with_no_features(self):
builder = self.builder_for_features()
# Since the builder advertises no features, you can't find it
# by looking up features.
self.assertEqual(self.registry.lookup('foo'), None)
# But you can find it by doing a lookup with no features, if
# this happens to be the only registered builder.
self.assertEqual(self.registry.lookup(), builder)
def test_register_with_features_makes_lookup_succeed(self):
builder = self.builder_for_features('foo', 'bar')
self.assertEqual(self.registry.lookup('foo'), builder)
self.assertEqual(self.registry.lookup('bar'), builder)
def test_lookup_fails_when_no_builder_implements_feature(self):
builder = self.builder_for_features('foo', 'bar')
self.assertEqual(self.registry.lookup('baz'), None)
def test_lookup_gets_most_recent_registration_when_no_feature_specified(self):
builder1 = self.builder_for_features('foo')
builder2 = self.builder_for_features('bar')
self.assertEqual(self.registry.lookup(), builder2)
def test_lookup_fails_when_no_tree_builders_registered(self):
self.assertEqual(self.registry.lookup(), None)
def test_lookup_gets_most_recent_builder_supporting_all_features(self):
has_one = self.builder_for_features('foo')
has_the_other = self.builder_for_features('bar')
has_both_early = self.builder_for_features('foo', 'bar', 'baz')
has_both_late = self.builder_for_features('foo', 'bar', 'quux')
lacks_one = self.builder_for_features('bar')
has_the_other = self.builder_for_features('foo')
# There are two builders featuring 'foo' and 'bar', but
# the one that also features 'quux' was registered later.
self.assertEqual(self.registry.lookup('foo', 'bar'),
has_both_late)
# There is only one builder featuring 'foo', 'bar', and 'baz'.
self.assertEqual(self.registry.lookup('foo', 'bar', 'baz'),
has_both_early)
def test_lookup_fails_when_cannot_reconcile_requested_features(self):
builder1 = self.builder_for_features('foo', 'bar')
builder2 = self.builder_for_features('foo', 'baz')
self.assertEqual(self.registry.lookup('bar', 'baz'), None)

36
lib/bs4/tests/test_docs.py Normal file
View file

@ -0,0 +1,36 @@
"Test harness for doctests."
# pylint: disable-msg=E0611,W0142
__metaclass__ = type
__all__ = [
'additional_tests',
]
import atexit
import doctest
import os
#from pkg_resources import (
# resource_filename, resource_exists, resource_listdir, cleanup_resources)
import unittest
DOCTEST_FLAGS = (
doctest.ELLIPSIS |
doctest.NORMALIZE_WHITESPACE |
doctest.REPORT_NDIFF)
# def additional_tests():
# "Run the doc tests (README.txt and docs/*, if any exist)"
# doctest_files = [
# os.path.abspath(resource_filename('bs4', 'README.txt'))]
# if resource_exists('bs4', 'docs'):
# for name in resource_listdir('bs4', 'docs'):
# if name.endswith('.txt'):
# doctest_files.append(
# os.path.abspath(
# resource_filename('bs4', 'docs/%s' % name)))
# kwargs = dict(module_relative=False, optionflags=DOCTEST_FLAGS)
# atexit.register(cleanup_resources)
# return unittest.TestSuite((
# doctest.DocFileSuite(*doctest_files, **kwargs)))

184
lib/bs4/tests/test_html5lib.py Normal file
View file

@ -0,0 +1,184 @@
"""Tests to ensure that the html5lib tree builder generates good trees."""
import warnings
try:
from bs4.builder import HTML5TreeBuilder
HTML5LIB_PRESENT = True
except ImportError as e:
HTML5LIB_PRESENT = False
from bs4.element import SoupStrainer
from bs4.testing import (
HTML5TreeBuilderSmokeTest,
SoupTest,
skipIf,
)
@skipIf(
not HTML5LIB_PRESENT,
"html5lib seems not to be present, not testing its tree builder.")
class HTML5LibBuilderSmokeTest(SoupTest, HTML5TreeBuilderSmokeTest):
"""See ``HTML5TreeBuilderSmokeTest``."""
@property
def default_builder(self):
return HTML5TreeBuilder
def test_soupstrainer(self):
# The html5lib tree builder does not support SoupStrainers.
strainer = SoupStrainer("b")
markup = "<p>A <b>bold</b> statement.</p>"
with warnings.catch_warnings(record=True) as w:
soup = self.soup(markup, parse_only=strainer)
self.assertEqual(
soup.decode(), self.document_for(markup))
self.assertTrue(
"the html5lib tree builder doesn't support parse_only" in
str(w[0].message))
def test_correctly_nested_tables(self):
"""html5lib inserts <tbody> tags where other parsers don't."""
markup = ('<table id="1">'
'<tr>'
"<td>Here's another table:"
'<table id="2">'
'<tr><td>foo</td></tr>'
'</table></td>')
self.assertSoupEquals(
markup,
'<table id="1"><tbody><tr><td>Here\'s another table:'
'<table id="2"><tbody><tr><td>foo</td></tr></tbody></table>'
'</td></tr></tbody></table>')
self.assertSoupEquals(
"<table><thead><tr><td>Foo</td></tr></thead>"
"<tbody><tr><td>Bar</td></tr></tbody>"
"<tfoot><tr><td>Baz</td></tr></tfoot></table>")
def test_xml_declaration_followed_by_doctype(self):
markup = '''<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html>
<html>
<head>
</head>
<body>
<p>foo</p>
</body>
</html>'''
soup = self.soup(markup)
# Verify that we can reach the <p> tag; this means the tree is connected.
self.assertEqual(b"<p>foo</p>", soup.p.encode())
def test_reparented_markup(self):
markup = '<p><em>foo</p>\n<p>bar<a></a></em></p>'
soup = self.soup(markup)
self.assertEqual("<body><p><em>foo</em></p><em>\n</em><p><em>bar<a></a></em></p></body>", soup.body.decode())
self.assertEqual(2, len(soup.find_all('p')))
def test_reparented_markup_ends_with_whitespace(self):
markup = '<p><em>foo</p>\n<p>bar<a></a></em></p>\n'
soup = self.soup(markup)
self.assertEqual("<body><p><em>foo</em></p><em>\n</em><p><em>bar<a></a></em></p>\n</body>", soup.body.decode())
self.assertEqual(2, len(soup.find_all('p')))
def test_reparented_markup_containing_identical_whitespace_nodes(self):
"""Verify that we keep the two whitespace nodes in this
document distinct when reparenting the adjacent <tbody> tags.
"""
markup = '<table> <tbody><tbody><ims></tbody> </table>'
soup = self.soup(markup)
space1, space2 = soup.find_all(string=' ')
tbody1, tbody2 = soup.find_all('tbody')
assert space1.next_element is tbody1
assert tbody2.next_element is space2
def test_reparented_markup_containing_children(self):
markup = '<div><a>aftermath<p><noscript>target</noscript>aftermath</a></p></div>'
soup = self.soup(markup)
noscript = soup.noscript
self.assertEqual("target", noscript.next_element)
target = soup.find(string='target')
# The 'aftermath' string was duplicated; we want the second one.
final_aftermath = soup.find_all(string='aftermath')[-1]
# The <noscript> tag was moved beneath a copy of the <a> tag,
# but the 'target' string within is still connected to the
# (second) 'aftermath' string.
self.assertEqual(final_aftermath, target.next_element)
self.assertEqual(target, final_aftermath.previous_element)
def test_processing_instruction(self):
"""Processing instructions become comments."""
markup = b"""<?PITarget PIContent?>"""
soup = self.soup(markup)
assert str(soup).startswith("<!--?PITarget PIContent?-->")
def test_cloned_multivalue_node(self):
markup = b"""<a class="my_class"><p></a>"""
soup = self.soup(markup)
a1, a2 = soup.find_all('a')
self.assertEqual(a1, a2)
assert a1 is not a2
def test_foster_parenting(self):
markup = b"""<table><td></tbody>A"""
soup = self.soup(markup)
self.assertEqual("<body>A<table><tbody><tr><td></td></tr></tbody></table></body>", soup.body.decode())
def test_extraction(self):
"""
Test that extraction does not destroy the tree.
https://bugs.launchpad.net/beautifulsoup/+bug/1782928
"""
markup = """
<html><head></head>
<style>
</style><script></script><body><p>hello</p></body></html>
"""
soup = self.soup(markup)
[s.extract() for s in soup('script')]
[s.extract() for s in soup('style')]
self.assertEqual(len(soup.find_all("p")), 1)
def test_empty_comment(self):
"""
Test that empty comment does not break structure.
https://bugs.launchpad.net/beautifulsoup/+bug/1806598
"""
markup = """
<html>
<body>
<form>
<!----><input type="text">
</form>
</body>
</html>
"""
soup = self.soup(markup)
inputs = []
for form in soup.find_all('form'):
inputs.extend(form.find_all('input'))
self.assertEqual(len(inputs), 1)
def test_tracking_line_numbers(self):
# The html.parser TreeBuilder keeps track of line number and
# position of each element.
markup = "\n <p>\n\n<sourceline>\n<b>text</b></sourceline><sourcepos></p>"
soup = self.soup(markup)
self.assertEqual(2, soup.p.sourceline)
self.assertEqual(5, soup.p.sourcepos)
self.assertEqual("sourceline", soup.p.find('sourceline').name)
# You can deactivate this behavior.
soup = self.soup(markup, store_line_numbers=False)
self.assertEqual("sourceline", soup.p.sourceline.name)
self.assertEqual("sourcepos", soup.p.sourcepos.name)

61
lib/bs4/tests/test_htmlparser.py Normal file
View file

@ -0,0 +1,61 @@
"""Tests to ensure that the html.parser tree builder generates good
trees."""
from pdb import set_trace
import pickle
from bs4.testing import SoupTest, HTMLTreeBuilderSmokeTest
from bs4.builder import HTMLParserTreeBuilder
from bs4.builder._htmlparser import BeautifulSoupHTMLParser
class HTMLParserTreeBuilderSmokeTest(SoupTest, HTMLTreeBuilderSmokeTest):
default_builder = HTMLParserTreeBuilder
def test_namespaced_system_doctype(self):
# html.parser can't handle namespaced doctypes, so skip this one.
pass
def test_namespaced_public_doctype(self):
# html.parser can't handle namespaced doctypes, so skip this one.
pass
def test_builder_is_pickled(self):
"""Unlike most tree builders, HTMLParserTreeBuilder and will
be restored after pickling.
"""
tree = self.soup("<a><b>foo</a>")
dumped = pickle.dumps(tree, 2)
loaded = pickle.loads(dumped)
self.assertTrue(isinstance(loaded.builder, type(tree.builder)))
def test_redundant_empty_element_closing_tags(self):
self.assertSoupEquals('<br></br><br></br><br></br>', "<br/><br/><br/>")
self.assertSoupEquals('</br></br></br>', "")
def test_empty_element(self):
# This verifies that any buffered data present when the parser
# finishes working is handled.
self.assertSoupEquals("foo &# bar", "foo &amp;# bar")
def test_tracking_line_numbers(self):
# The html.parser TreeBuilder keeps track of line number and
# position of each element.
markup = "\n <p>\n\n<sourceline>\n<b>text</b></sourceline><sourcepos></p>"
soup = self.soup(markup)
self.assertEqual(2, soup.p.sourceline)
self.assertEqual(3, soup.p.sourcepos)
self.assertEqual("sourceline", soup.p.find('sourceline').name)
# You can deactivate this behavior.
soup = self.soup(markup, store_line_numbers=False)
self.assertEqual("sourceline", soup.p.sourceline.name)
self.assertEqual("sourcepos", soup.p.sourcepos.name)
class TestHTMLParserSubclass(SoupTest):
def test_error(self):
"""Verify that our HTMLParser subclass implements error() in a way
that doesn't cause a crash.
"""
parser = BeautifulSoupHTMLParser()
parser.error("don't crash")

115
lib/bs4/tests/test_lxml.py Normal file
View file

@ -0,0 +1,115 @@
"""Tests to ensure that the lxml tree builder generates good trees."""
import re
import warnings
try:
import lxml.etree
LXML_PRESENT = True
LXML_VERSION = lxml.etree.LXML_VERSION
except ImportError as e:
LXML_PRESENT = False
LXML_VERSION = (0,)
if LXML_PRESENT:
from bs4.builder import LXMLTreeBuilder, LXMLTreeBuilderForXML
from bs4 import (
BeautifulSoup,
BeautifulStoneSoup,
)
from bs4.element import Comment, Doctype, SoupStrainer
from bs4.testing import skipIf
from bs4.tests import test_htmlparser
from bs4.testing import (
HTMLTreeBuilderSmokeTest,
XMLTreeBuilderSmokeTest,
SoupTest,
skipIf,
)
@skipIf(
not LXML_PRESENT,
"lxml seems not to be present, not testing its tree builder.")
class LXMLTreeBuilderSmokeTest(SoupTest, HTMLTreeBuilderSmokeTest):
"""See ``HTMLTreeBuilderSmokeTest``."""
@property
def default_builder(self):
return LXMLTreeBuilder
def test_out_of_range_entity(self):
self.assertSoupEquals(
"<p>foo&#10000000000000;bar</p>", "<p>foobar</p>")
self.assertSoupEquals(
"<p>foo&#x10000000000000;bar</p>", "<p>foobar</p>")
self.assertSoupEquals(
"<p>foo&#1000000000;bar</p>", "<p>foobar</p>")
def test_entities_in_foreign_document_encoding(self):
# We can't implement this case correctly because by the time we
# hear about markup like "&#147;", it's been (incorrectly) converted into
# a string like u'\x93'
pass
# In lxml < 2.3.5, an empty doctype causes a segfault. Skip this
# test if an old version of lxml is installed.
@skipIf(
not LXML_PRESENT or LXML_VERSION < (2,3,5,0),
"Skipping doctype test for old version of lxml to avoid segfault.")
def test_empty_doctype(self):
soup = self.soup("<!DOCTYPE>")
doctype = soup.contents[0]
self.assertEqual("", doctype.strip())
def test_beautifulstonesoup_is_xml_parser(self):
# Make sure that the deprecated BSS class uses an xml builder
# if one is installed.
with warnings.catch_warnings(record=True) as w:
soup = BeautifulStoneSoup("<b />")
self.assertEqual("<b/>", str(soup.b))
self.assertTrue("BeautifulStoneSoup class is deprecated" in str(w[0].message))
def test_tracking_line_numbers(self):
# The lxml TreeBuilder cannot keep track of line numbers from
# the original markup. Even if you ask for line numbers, we
# don't have 'em.
#
# This means that if you have a tag like <sourceline> or
# <sourcepos>, attribute access will find it rather than
# giving you a numeric answer.
soup = self.soup(
"\n <p>\n\n<sourceline>\n<b>text</b></sourceline><sourcepos></p>",
store_line_numbers=True
)
self.assertEqual("sourceline", soup.p.sourceline.name)
self.assertEqual("sourcepos", soup.p.sourcepos.name)
@skipIf(
not LXML_PRESENT,
"lxml seems not to be present, not testing its XML tree builder.")
class LXMLXMLTreeBuilderSmokeTest(SoupTest, XMLTreeBuilderSmokeTest):
"""See ``HTMLTreeBuilderSmokeTest``."""
@property
def default_builder(self):
return LXMLTreeBuilderForXML
def test_namespace_indexing(self):
# We should not track un-prefixed namespaces as we can only hold one
# and it will be recognized as the default namespace by soupsieve,
# which may be confusing in some situations. When no namespace is provided
# for a selector, the default namespace (if defined) is assumed.
soup = self.soup(
'<?xml version="1.1"?>\n'
'<root>'
'<tag xmlns="http://unprefixed-namespace.com">content</tag>'
'<prefix:tag xmlns:prefix="http://prefixed-namespace.com">content</tag>'
'</root>'
)
self.assertEqual(
soup._namespaces,
{'xml': 'http://www.w3.org/XML/1998/namespace', 'prefix': 'http://prefixed-namespace.com'}
)

682
lib/bs4/tests/test_soup.py Normal file
View file

@ -0,0 +1,682 @@
# -*- coding: utf-8 -*-
"""Tests of Beautiful Soup as a whole."""
from pdb import set_trace
import logging
import unittest
import sys
import tempfile
from bs4 import (
BeautifulSoup,
BeautifulStoneSoup,
)
from bs4.builder import (
TreeBuilder,
ParserRejectedMarkup,
)
from bs4.element import (
CharsetMetaAttributeValue,
Comment,
ContentMetaAttributeValue,
SoupStrainer,
NamespacedAttribute,
Tag,
NavigableString,
)
import bs4.dammit
from bs4.dammit import (
EntitySubstitution,
UnicodeDammit,
EncodingDetector,
)
from bs4.testing import (
default_builder,
SoupTest,
skipIf,
)
import warnings
try:
from bs4.builder import LXMLTreeBuilder, LXMLTreeBuilderForXML
LXML_PRESENT = True
except ImportError as e:
LXML_PRESENT = False
PYTHON_3_PRE_3_2 = (sys.version_info[0] == 3 and sys.version_info < (3,2))
class TestConstructor(SoupTest):
def test_short_unicode_input(self):
data = "<h1>éé</h1>"
soup = self.soup(data)
self.assertEqual("éé", soup.h1.string)
def test_embedded_null(self):
data = "<h1>foo\0bar</h1>"
soup = self.soup(data)
self.assertEqual("foo\0bar", soup.h1.string)
def test_exclude_encodings(self):
utf8_data = "Räksmörgås".encode("utf-8")
soup = self.soup(utf8_data, exclude_encodings=["utf-8"])
self.assertEqual("windows-1252", soup.original_encoding)
def test_custom_builder_class(self):
# Verify that you can pass in a custom Builder class and
# it'll be instantiated with the appropriate keyword arguments.
class Mock(object):
def __init__(self, **kwargs):
self.called_with = kwargs
self.is_xml = True
self.store_line_numbers = False
self.cdata_list_attributes = []
self.preserve_whitespace_tags = []
def initialize_soup(self, soup):
pass
def feed(self, markup):
self.fed = markup
def reset(self):
pass
def ignore(self, ignore):
pass
set_up_substitutions = can_be_empty_element = ignore
def prepare_markup(self, *args, **kwargs):
yield "prepared markup", "original encoding", "declared encoding", "contains replacement characters"
kwargs = dict(
var="value",
# This is a deprecated BS3-era keyword argument, which
# will be stripped out.
convertEntities=True,
)
with warnings.catch_warnings(record=True):
soup = BeautifulSoup('', builder=Mock, **kwargs)
assert isinstance(soup.builder, Mock)
self.assertEqual(dict(var="value"), soup.builder.called_with)
self.assertEqual("prepared markup", soup.builder.fed)
# You can also instantiate the TreeBuilder yourself. In this
# case, that specific object is used and any keyword arguments
# to the BeautifulSoup constructor are ignored.
builder = Mock(**kwargs)
with warnings.catch_warnings(record=True) as w:
soup = BeautifulSoup(
'', builder=builder, ignored_value=True,
)
msg = str(w[0].message)
assert msg.startswith("Keyword arguments to the BeautifulSoup constructor will be ignored.")
self.assertEqual(builder, soup.builder)
self.assertEqual(kwargs, builder.called_with)
def test_parser_markup_rejection(self):
# If markup is completely rejected by the parser, an
# explanatory ParserRejectedMarkup exception is raised.
class Mock(TreeBuilder):
def feed(self, *args, **kwargs):
raise ParserRejectedMarkup("Nope.")
def prepare_markup(self, *args, **kwargs):
# We're going to try two different ways of preparing this markup,
# but feed() will reject both of them.
yield markup, None, None, False
yield markup, None, None, False
import re
self.assertRaisesRegex(
ParserRejectedMarkup,
"The markup you provided was rejected by the parser. Trying a different parser or a different encoding may help.",
BeautifulSoup, '', builder=Mock,
)
def test_cdata_list_attributes(self):
# Most attribute values are represented as scalars, but the
# HTML standard says that some attributes, like 'class' have
# space-separated lists as values.
markup = '<a id=" an id " class=" a class "></a>'
soup = self.soup(markup)
# Note that the spaces are stripped for 'class' but not for 'id'.
a = soup.a
self.assertEqual(" an id ", a['id'])
self.assertEqual(["a", "class"], a['class'])
# TreeBuilder takes an argument called 'mutli_valued_attributes' which lets
# you customize or disable this. As always, you can customize the TreeBuilder
# by passing in a keyword argument to the BeautifulSoup constructor.
soup = self.soup(markup, builder=default_builder, multi_valued_attributes=None)
self.assertEqual(" a class ", soup.a['class'])
# Here are two ways of saying that `id` is a multi-valued
# attribute in this context, but 'class' is not.
for switcheroo in ({'*': 'id'}, {'a': 'id'}):
with warnings.catch_warnings(record=True) as w:
# This will create a warning about not explicitly
# specifying a parser, but we'll ignore it.
soup = self.soup(markup, builder=None, multi_valued_attributes=switcheroo)
a = soup.a
self.assertEqual(["an", "id"], a['id'])
self.assertEqual(" a class ", a['class'])
def test_replacement_classes(self):
# Test the ability to pass in replacements for element classes
# which will be used when building the tree.
class TagPlus(Tag):
pass
class StringPlus(NavigableString):
pass
class CommentPlus(Comment):
pass
soup = self.soup(
"<a><b>foo</b>bar</a><!--whee-->",
element_classes = {
Tag: TagPlus,
NavigableString: StringPlus,
Comment: CommentPlus,
}
)
# The tree was built with TagPlus, StringPlus, and CommentPlus objects,
# rather than Tag, String, and Comment objects.
assert all(
isinstance(x, (TagPlus, StringPlus, CommentPlus))
for x in soup.recursiveChildGenerator()
)
class TestWarnings(SoupTest):
def _no_parser_specified(self, s, is_there=True):
v = s.startswith(BeautifulSoup.NO_PARSER_SPECIFIED_WARNING[:80])
self.assertTrue(v)
def test_warning_if_no_parser_specified(self):
with warnings.catch_warnings(record=True) as w:
soup = self.soup("<a><b></b></a>")
msg = str(w[0].message)
self._assert_no_parser_specified(msg)
def test_warning_if_parser_specified_too_vague(self):
with warnings.catch_warnings(record=True) as w:
soup = self.soup("<a><b></b></a>", "html")
msg = str(w[0].message)
self._assert_no_parser_specified(msg)
def test_no_warning_if_explicit_parser_specified(self):
with warnings.catch_warnings(record=True) as w:
soup = self.soup("<a><b></b></a>", "html.parser")
self.assertEqual([], w)
def test_parseOnlyThese_renamed_to_parse_only(self):
with warnings.catch_warnings(record=True) as w:
soup = self.soup("<a><b></b></a>", parseOnlyThese=SoupStrainer("b"))
msg = str(w[0].message)
self.assertTrue("parseOnlyThese" in msg)
self.assertTrue("parse_only" in msg)
self.assertEqual(b"<b></b>", soup.encode())
def test_fromEncoding_renamed_to_from_encoding(self):
with warnings.catch_warnings(record=True) as w:
utf8 = b"\xc3\xa9"
soup = self.soup(utf8, fromEncoding="utf8")
msg = str(w[0].message)
self.assertTrue("fromEncoding" in msg)
self.assertTrue("from_encoding" in msg)
self.assertEqual("utf8", soup.original_encoding)
def test_unrecognized_keyword_argument(self):
self.assertRaises(
TypeError, self.soup, "<a>", no_such_argument=True)
class TestWarnings(SoupTest):
def test_disk_file_warning(self):
filehandle = tempfile.NamedTemporaryFile()
filename = filehandle.name
try:
with warnings.catch_warnings(record=True) as w:
soup = self.soup(filename)
msg = str(w[0].message)
self.assertTrue("looks like a filename" in msg)
finally:
filehandle.close()
# The file no longer exists, so Beautiful Soup will no longer issue the warning.
with warnings.catch_warnings(record=True) as w:
soup = self.soup(filename)
self.assertEqual(0, len(w))
def test_url_warning_with_bytes_url(self):
with warnings.catch_warnings(record=True) as warning_list:
soup = self.soup(b"http://www.crummybytes.com/")
# Be aware this isn't the only warning that can be raised during
# execution..
self.assertTrue(any("looks like a URL" in str(w.message)
for w in warning_list))
def test_url_warning_with_unicode_url(self):
with warnings.catch_warnings(record=True) as warning_list:
# note - this url must differ from the bytes one otherwise
# python's warnings system swallows the second warning
soup = self.soup("http://www.crummyunicode.com/")
self.assertTrue(any("looks like a URL" in str(w.message)
for w in warning_list))
def test_url_warning_with_bytes_and_space(self):
with warnings.catch_warnings(record=True) as warning_list:
soup = self.soup(b"http://www.crummybytes.com/ is great")
self.assertFalse(any("looks like a URL" in str(w.message)
for w in warning_list))
def test_url_warning_with_unicode_and_space(self):
with warnings.catch_warnings(record=True) as warning_list:
soup = self.soup("http://www.crummyuncode.com/ is great")
self.assertFalse(any("looks like a URL" in str(w.message)
for w in warning_list))
class TestSelectiveParsing(SoupTest):
def test_parse_with_soupstrainer(self):
markup = "No<b>Yes</b><a>No<b>Yes <c>Yes</c></b>"
strainer = SoupStrainer("b")
soup = self.soup(markup, parse_only=strainer)
self.assertEqual(soup.encode(), b"<b>Yes</b><b>Yes <c>Yes</c></b>")
class TestEntitySubstitution(unittest.TestCase):
"""Standalone tests of the EntitySubstitution class."""
def setUp(self):
self.sub = EntitySubstitution
def test_simple_html_substitution(self):
# Unicode characters corresponding to named HTML entites
# are substituted, and no others.
s = "foo\u2200\N{SNOWMAN}\u00f5bar"
self.assertEqual(self.sub.substitute_html(s),
"foo&forall;\N{SNOWMAN}&otilde;bar")
def test_smart_quote_substitution(self):
# MS smart quotes are a common source of frustration, so we
# give them a special test.
quotes = b"\x91\x92foo\x93\x94"
dammit = UnicodeDammit(quotes)
self.assertEqual(self.sub.substitute_html(dammit.markup),
"&lsquo;&rsquo;foo&ldquo;&rdquo;")
def test_xml_converstion_includes_no_quotes_if_make_quoted_attribute_is_false(self):
s = 'Welcome to "my bar"'
self.assertEqual(self.sub.substitute_xml(s, False), s)
def test_xml_attribute_quoting_normally_uses_double_quotes(self):
self.assertEqual(self.sub.substitute_xml("Welcome", True),
'"Welcome"')
self.assertEqual(self.sub.substitute_xml("Bob's Bar", True),
'"Bob\'s Bar"')
def test_xml_attribute_quoting_uses_single_quotes_when_value_contains_double_quotes(self):
s = 'Welcome to "my bar"'
self.assertEqual(self.sub.substitute_xml(s, True),
"'Welcome to \"my bar\"'")
def test_xml_attribute_quoting_escapes_single_quotes_when_value_contains_both_single_and_double_quotes(self):
s = 'Welcome to "Bob\'s Bar"'
self.assertEqual(
self.sub.substitute_xml(s, True),
'"Welcome to &quot;Bob\'s Bar&quot;"')
def test_xml_quotes_arent_escaped_when_value_is_not_being_quoted(self):
quoted = 'Welcome to "Bob\'s Bar"'
self.assertEqual(self.sub.substitute_xml(quoted), quoted)
def test_xml_quoting_handles_angle_brackets(self):
self.assertEqual(
self.sub.substitute_xml("foo<bar>"),
"foo&lt;bar&gt;")
def test_xml_quoting_handles_ampersands(self):
self.assertEqual(self.sub.substitute_xml("AT&T"), "AT&amp;T")
def test_xml_quoting_including_ampersands_when_they_are_part_of_an_entity(self):
self.assertEqual(
self.sub.substitute_xml("&Aacute;T&T"),
"&amp;Aacute;T&amp;T")
def test_xml_quoting_ignoring_ampersands_when_they_are_part_of_an_entity(self):
self.assertEqual(
self.sub.substitute_xml_containing_entities("&Aacute;T&T"),
"&Aacute;T&amp;T")
def test_quotes_not_html_substituted(self):
"""There's no need to do this except inside attribute values."""
text = 'Bob\'s "bar"'
self.assertEqual(self.sub.substitute_html(text), text)
class TestEncodingConversion(SoupTest):
# Test Beautiful Soup's ability to decode and encode from various
# encodings.
def setUp(self):
super(TestEncodingConversion, self).setUp()
self.unicode_data = '<html><head><meta charset="utf-8"/></head><body><foo>Sacr\N{LATIN SMALL LETTER E WITH ACUTE} bleu!</foo></body></html>'
self.utf8_data = self.unicode_data.encode("utf-8")
# Just so you know what it looks like.
self.assertEqual(
self.utf8_data,
b'<html><head><meta charset="utf-8"/></head><body><foo>Sacr\xc3\xa9 bleu!</foo></body></html>')
def test_ascii_in_unicode_out(self):
# ASCII input is converted to Unicode. The original_encoding
# attribute is set to 'utf-8', a superset of ASCII.
chardet = bs4.dammit.chardet_dammit
logging.disable(logging.WARNING)
try:
def noop(str):
return None
# Disable chardet, which will realize that the ASCII is ASCII.
bs4.dammit.chardet_dammit = noop
ascii = b"<foo>a</foo>"
soup_from_ascii = self.soup(ascii)
unicode_output = soup_from_ascii.decode()
self.assertTrue(isinstance(unicode_output, str))
self.assertEqual(unicode_output, self.document_for(ascii.decode()))
self.assertEqual(soup_from_ascii.original_encoding.lower(), "utf-8")
finally:
logging.disable(logging.NOTSET)
bs4.dammit.chardet_dammit = chardet
def test_unicode_in_unicode_out(self):
# Unicode input is left alone. The original_encoding attribute
# is not set.
soup_from_unicode = self.soup(self.unicode_data)
self.assertEqual(soup_from_unicode.decode(), self.unicode_data)
self.assertEqual(soup_from_unicode.foo.string, 'Sacr\xe9 bleu!')
self.assertEqual(soup_from_unicode.original_encoding, None)
def test_utf8_in_unicode_out(self):
# UTF-8 input is converted to Unicode. The original_encoding
# attribute is set.
soup_from_utf8 = self.soup(self.utf8_data)
self.assertEqual(soup_from_utf8.decode(), self.unicode_data)
self.assertEqual(soup_from_utf8.foo.string, 'Sacr\xe9 bleu!')
def test_utf8_out(self):
# The internal data structures can be encoded as UTF-8.
soup_from_unicode = self.soup(self.unicode_data)
self.assertEqual(soup_from_unicode.encode('utf-8'), self.utf8_data)
@skipIf(
PYTHON_3_PRE_3_2,
"Bad HTMLParser detected; skipping test of non-ASCII characters in attribute name.")
def test_attribute_name_containing_unicode_characters(self):
markup = '<div><a \N{SNOWMAN}="snowman"></a></div>'
self.assertEqual(self.soup(markup).div.encode("utf8"), markup.encode("utf8"))
class TestUnicodeDammit(unittest.TestCase):
"""Standalone tests of UnicodeDammit."""
def test_unicode_input(self):
markup = "I'm already Unicode! \N{SNOWMAN}"
dammit = UnicodeDammit(markup)
self.assertEqual(dammit.unicode_markup, markup)
def test_smart_quotes_to_unicode(self):
markup = b"<foo>\x91\x92\x93\x94</foo>"
dammit = UnicodeDammit(markup)
self.assertEqual(
dammit.unicode_markup, "<foo>\u2018\u2019\u201c\u201d</foo>")
def test_smart_quotes_to_xml_entities(self):
markup = b"<foo>\x91\x92\x93\x94</foo>"
dammit = UnicodeDammit(markup, smart_quotes_to="xml")
self.assertEqual(
dammit.unicode_markup, "<foo>&#x2018;&#x2019;&#x201C;&#x201D;</foo>")
def test_smart_quotes_to_html_entities(self):
markup = b"<foo>\x91\x92\x93\x94</foo>"
dammit = UnicodeDammit(markup, smart_quotes_to="html")
self.assertEqual(
dammit.unicode_markup, "<foo>&lsquo;&rsquo;&ldquo;&rdquo;</foo>")
def test_smart_quotes_to_ascii(self):
markup = b"<foo>\x91\x92\x93\x94</foo>"
dammit = UnicodeDammit(markup, smart_quotes_to="ascii")
self.assertEqual(
dammit.unicode_markup, """<foo>''""</foo>""")
def test_detect_utf8(self):
utf8 = b"Sacr\xc3\xa9 bleu! \xe2\x98\x83"
dammit = UnicodeDammit(utf8)
self.assertEqual(dammit.original_encoding.lower(), 'utf-8')
self.assertEqual(dammit.unicode_markup, 'Sacr\xe9 bleu! \N{SNOWMAN}')
def test_convert_hebrew(self):
hebrew = b"\xed\xe5\xec\xf9"
dammit = UnicodeDammit(hebrew, ["iso-8859-8"])
self.assertEqual(dammit.original_encoding.lower(), 'iso-8859-8')
self.assertEqual(dammit.unicode_markup, '\u05dd\u05d5\u05dc\u05e9')
def test_dont_see_smart_quotes_where_there_are_none(self):
utf_8 = b"\343\202\261\343\203\274\343\202\277\343\202\244 Watch"
dammit = UnicodeDammit(utf_8)
self.assertEqual(dammit.original_encoding.lower(), 'utf-8')
self.assertEqual(dammit.unicode_markup.encode("utf-8"), utf_8)
def test_ignore_inappropriate_codecs(self):
utf8_data = "Räksmörgås".encode("utf-8")
dammit = UnicodeDammit(utf8_data, ["iso-8859-8"])
self.assertEqual(dammit.original_encoding.lower(), 'utf-8')
def test_ignore_invalid_codecs(self):
utf8_data = "Räksmörgås".encode("utf-8")
for bad_encoding in ['.utf8', '...', 'utF---16.!']:
dammit = UnicodeDammit(utf8_data, [bad_encoding])
self.assertEqual(dammit.original_encoding.lower(), 'utf-8')
def test_exclude_encodings(self):
# This is UTF-8.
utf8_data = "Räksmörgås".encode("utf-8")
# But if we exclude UTF-8 from consideration, the guess is
# Windows-1252.
dammit = UnicodeDammit(utf8_data, exclude_encodings=["utf-8"])
self.assertEqual(dammit.original_encoding.lower(), 'windows-1252')
# And if we exclude that, there is no valid guess at all.
dammit = UnicodeDammit(
utf8_data, exclude_encodings=["utf-8", "windows-1252"])
self.assertEqual(dammit.original_encoding, None)
def test_encoding_detector_replaces_junk_in_encoding_name_with_replacement_character(self):
detected = EncodingDetector(
b'<?xml version="1.0" encoding="UTF-\xdb" ?>')
encodings = list(detected.encodings)
assert 'utf-\N{REPLACEMENT CHARACTER}' in encodings
def test_detect_html5_style_meta_tag(self):
for data in (
b'<html><meta charset="euc-jp" /></html>',
b"<html><meta charset='euc-jp' /></html>",
b"<html><meta charset=euc-jp /></html>",
b"<html><meta charset=euc-jp/></html>"):
dammit = UnicodeDammit(data, is_html=True)
self.assertEqual(
"euc-jp", dammit.original_encoding)
def test_last_ditch_entity_replacement(self):
# This is a UTF-8 document that contains bytestrings
# completely incompatible with UTF-8 (ie. encoded with some other
# encoding).
#
# Since there is no consistent encoding for the document,
# Unicode, Dammit will eventually encode the document as UTF-8
# and encode the incompatible characters as REPLACEMENT
# CHARACTER.
#
# If chardet is installed, it will detect that the document
# can be converted into ISO-8859-1 without errors. This happens
# to be the wrong encoding, but it is a consistent encoding, so the
# code we're testing here won't run.
#
# So we temporarily disable chardet if it's present.
doc = b"""\357\273\277<?xml version="1.0" encoding="UTF-8"?>
<html><b>\330\250\330\252\330\261</b>
<i>\310\322\321\220\312\321\355\344</i></html>"""
chardet = bs4.dammit.chardet_dammit
logging.disable(logging.WARNING)
try:
def noop(str):
return None
bs4.dammit.chardet_dammit = noop
dammit = UnicodeDammit(doc)
self.assertEqual(True, dammit.contains_replacement_characters)
self.assertTrue("\ufffd" in dammit.unicode_markup)
soup = BeautifulSoup(doc, "html.parser")
self.assertTrue(soup.contains_replacement_characters)
finally:
logging.disable(logging.NOTSET)
bs4.dammit.chardet_dammit = chardet
def test_byte_order_mark_removed(self):
# A document written in UTF-16LE will have its byte order marker stripped.
data = b'\xff\xfe<\x00a\x00>\x00\xe1\x00\xe9\x00<\x00/\x00a\x00>\x00'
dammit = UnicodeDammit(data)
self.assertEqual("<a>áé</a>", dammit.unicode_markup)
self.assertEqual("utf-16le", dammit.original_encoding)
def test_detwingle(self):
# Here's a UTF8 document.
utf8 = ("\N{SNOWMAN}" * 3).encode("utf8")
# Here's a Windows-1252 document.
windows_1252 = (
"\N{LEFT DOUBLE QUOTATION MARK}Hi, I like Windows!"
"\N{RIGHT DOUBLE QUOTATION MARK}").encode("windows_1252")
# Through some unholy alchemy, they've been stuck together.
doc = utf8 + windows_1252 + utf8
# The document can't be turned into UTF-8:
self.assertRaises(UnicodeDecodeError, doc.decode, "utf8")
# Unicode, Dammit thinks the whole document is Windows-1252,
# and decodes it into "☃☃☃“Hi, I like Windows!”☃☃☃"
# But if we run it through fix_embedded_windows_1252, it's fixed:
fixed = UnicodeDammit.detwingle(doc)
self.assertEqual(
"☃☃☃“Hi, I like Windows!”☃☃☃", fixed.decode("utf8"))
def test_detwingle_ignores_multibyte_characters(self):
# Each of these characters has a UTF-8 representation ending
# in \x93. \x93 is a smart quote if interpreted as
# Windows-1252. But our code knows to skip over multibyte
# UTF-8 characters, so they'll survive the process unscathed.
for tricky_unicode_char in (
"\N{LATIN SMALL LIGATURE OE}", # 2-byte char '\xc5\x93'
"\N{LATIN SUBSCRIPT SMALL LETTER X}", # 3-byte char '\xe2\x82\x93'
"\xf0\x90\x90\x93", # This is a CJK character, not sure which one.
):
input = tricky_unicode_char.encode("utf8")
self.assertTrue(input.endswith(b'\x93'))
output = UnicodeDammit.detwingle(input)
self.assertEqual(output, input)
def test_find_declared_encoding(self):
# Test our ability to find a declared encoding inside an
# XML or HTML document.
#
# Even if the document comes in as Unicode, it may be
# interesting to know what encoding was claimed
# originally.
html_unicode = '<html><head><meta charset="utf-8"></head></html>'
html_bytes = html_unicode.encode("ascii")
xml_unicode= '<?xml version="1.0" encoding="ISO-8859-1" ?>'
xml_bytes = xml_unicode.encode("ascii")
m = EncodingDetector.find_declared_encoding
self.assertEqual(None, m(html_unicode, is_html=False))
self.assertEqual("utf-8", m(html_unicode, is_html=True))
self.assertEqual("utf-8", m(html_bytes, is_html=True))
self.assertEqual("iso-8859-1", m(xml_unicode))
self.assertEqual("iso-8859-1", m(xml_bytes))
# Normally, only the first few kilobytes of a document are checked for
# an encoding.
spacer = b' ' * 5000
self.assertEqual(None, m(spacer + html_bytes))
self.assertEqual(None, m(spacer + xml_bytes))
# But you can tell find_declared_encoding to search an entire
# HTML document.
self.assertEqual(
"utf-8",
m(spacer + html_bytes, is_html=True, search_entire_document=True)
)
# The XML encoding declaration has to be the very first thing
# in the document. We'll allow whitespace before the document
# starts, but nothing else.
self.assertEqual(
"iso-8859-1",
m(xml_bytes, search_entire_document=True)
)
self.assertEqual(
None, m(b'a' + xml_bytes, search_entire_document=True)
)
class TestNamedspacedAttribute(SoupTest):
def test_name_may_be_none_or_missing(self):
a = NamespacedAttribute("xmlns", None)
self.assertEqual(a, "xmlns")
a = NamespacedAttribute("xmlns")
self.assertEqual(a, "xmlns")
def test_attribute_is_equivalent_to_colon_separated_string(self):
a = NamespacedAttribute("a", "b")
self.assertEqual("a:b", a)
def test_attributes_are_equivalent_if_prefix_and_name_identical(self):
a = NamespacedAttribute("a", "b", "c")
b = NamespacedAttribute("a", "b", "c")
self.assertEqual(a, b)
# The actual namespace is not considered.
c = NamespacedAttribute("a", "b", None)
self.assertEqual(a, c)
# But name and prefix are important.
d = NamespacedAttribute("a", "z", "c")
self.assertNotEqual(a, d)
e = NamespacedAttribute("z", "b", "c")
self.assertNotEqual(a, e)
class TestAttributeValueWithCharsetSubstitution(unittest.TestCase):
def test_content_meta_attribute_value(self):
value = CharsetMetaAttributeValue("euc-jp")
self.assertEqual("euc-jp", value)
self.assertEqual("euc-jp", value.original_value)
self.assertEqual("utf8", value.encode("utf8"))
def test_content_meta_attribute_value(self):
value = ContentMetaAttributeValue("text/html; charset=euc-jp")
self.assertEqual("text/html; charset=euc-jp", value)
self.assertEqual("text/html; charset=euc-jp", value.original_value)
self.assertEqual("text/html; charset=utf8", value.encode("utf8"))

2254
lib/bs4/tests/test_tree.py Normal file
View file

File diff suppressed because it is too large Load diff