mirror of
https://github.com/ytdl-org/youtube-dl
synced 2025-08-18 00:24:14 +09:00
Compare commits
No commits in common. "21924742f79ccbd62d16ef4120518c6a5da8614e" and "a08f2b7e4567cdc50c0614ee0a4ffdff49b8b6e6" have entirely different histories.
21924742f7
...
a08f2b7e45
@ -153,9 +153,6 @@ class TestInfoExtractor(unittest.TestCase):
|
||||
'''
|
||||
search = self.ie._search_nextjs_data(html, 'testID')
|
||||
self.assertEqual(search['props']['pageProps']['video']['id'], 'testid')
|
||||
search = self.ie._search_nextjs_data(
|
||||
'no next.js data here, move along', 'testID', default={'status': 0})
|
||||
self.assertEqual(search['status'], 0)
|
||||
|
||||
def test_search_nuxt_data(self):
|
||||
html = '''
|
||||
|
||||
@ -1,509 +0,0 @@
|
||||
#!/usr/bin/env python
|
||||
# coding: utf-8
|
||||
from __future__ import unicode_literals
|
||||
|
||||
# Allow direct execution
|
||||
import os
|
||||
import sys
|
||||
import unittest
|
||||
sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||
|
||||
|
||||
import re
|
||||
|
||||
from youtube_dl.traversal import (
|
||||
dict_get,
|
||||
get_first,
|
||||
T,
|
||||
traverse_obj,
|
||||
)
|
||||
from youtube_dl.compat import (
|
||||
compat_etree_fromstring,
|
||||
compat_http_cookies,
|
||||
compat_str,
|
||||
)
|
||||
from youtube_dl.utils import (
|
||||
int_or_none,
|
||||
str_or_none,
|
||||
)
|
||||
|
||||
_TEST_DATA = {
|
||||
100: 100,
|
||||
1.2: 1.2,
|
||||
'str': 'str',
|
||||
'None': None,
|
||||
'...': Ellipsis,
|
||||
'urls': [
|
||||
{'index': 0, 'url': 'https://www.example.com/0'},
|
||||
{'index': 1, 'url': 'https://www.example.com/1'},
|
||||
],
|
||||
'data': (
|
||||
{'index': 2},
|
||||
{'index': 3},
|
||||
),
|
||||
'dict': {},
|
||||
}
|
||||
|
||||
|
||||
if sys.version_info < (3, 0):
|
||||
class _TestCase(unittest.TestCase):
|
||||
|
||||
def assertCountEqual(self, *args, **kwargs):
|
||||
return self.assertItemsEqual(*args, **kwargs)
|
||||
else:
|
||||
_TestCase = unittest.TestCase
|
||||
|
||||
|
||||
class TestTraversal(_TestCase):
|
||||
def assertMaybeCountEqual(self, *args, **kwargs):
|
||||
if sys.version_info < (3, 7):
|
||||
# random dict order
|
||||
return self.assertCountEqual(*args, **kwargs)
|
||||
else:
|
||||
return self.assertEqual(*args, **kwargs)
|
||||
|
||||
def test_traverse_obj(self):
|
||||
# instant compat
|
||||
str = compat_str
|
||||
|
||||
# define a pukka Iterable
|
||||
def iter_range(stop):
|
||||
for from_ in range(stop):
|
||||
yield from_
|
||||
|
||||
# Test base functionality
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('str',)), 'str',
|
||||
msg='allow tuple path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ['str']), 'str',
|
||||
msg='allow list path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (value for value in ("str",))), 'str',
|
||||
msg='allow iterable path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 'str'), 'str',
|
||||
msg='single items should be treated as a path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, None), _TEST_DATA)
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 100), 100)
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 1.2), 1.2)
|
||||
|
||||
# Test Ellipsis behavior
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, Ellipsis),
|
||||
(item for item in _TEST_DATA.values() if item not in (None, {})),
|
||||
msg='`...` should give all non-discarded values')
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, ('urls', 0, Ellipsis)), _TEST_DATA['urls'][0].values(),
|
||||
msg='`...` selection for dicts should select all values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, Ellipsis, 'url')),
|
||||
['https://www.example.com/0', 'https://www.example.com/1'],
|
||||
msg='nested `...` queries should work')
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, (Ellipsis, Ellipsis, 'index')), iter_range(4),
|
||||
msg='`...` query result should be flattened')
|
||||
self.assertEqual(traverse_obj(iter(range(4)), Ellipsis), list(range(4)),
|
||||
msg='`...` should accept iterables')
|
||||
|
||||
# Test function as key
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, lambda x, y: x == 'urls' and isinstance(y, list)),
|
||||
[_TEST_DATA['urls']],
|
||||
msg='function as query key should perform a filter based on (key, value)')
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, lambda _, x: isinstance(x[0], str)), set(('str',)),
|
||||
msg='exceptions in the query function should be caught')
|
||||
self.assertEqual(traverse_obj(iter(range(4)), lambda _, x: x % 2 == 0), [0, 2],
|
||||
msg='function key should accept iterables')
|
||||
if __debug__:
|
||||
with self.assertRaises(Exception, msg='Wrong function signature should raise in debug'):
|
||||
traverse_obj(_TEST_DATA, lambda a: Ellipsis)
|
||||
with self.assertRaises(Exception, msg='Wrong function signature should raise in debug'):
|
||||
traverse_obj(_TEST_DATA, lambda a, b, c: Ellipsis)
|
||||
|
||||
# Test set as key (transformation/type, like `expected_type`)
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str.upper), )), ['STR'],
|
||||
msg='Function in set should be a transformation')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('fail', T(lambda _: 'const'))), 'const',
|
||||
msg='Function in set should always be called')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str))), ['str'],
|
||||
msg='Type in set should be a type filter')
|
||||
self.assertMaybeCountEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str, int))), [100, 'str'],
|
||||
msg='Multiple types in set should be a type filter')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, T(dict)), _TEST_DATA,
|
||||
msg='A single set should be wrapped into a path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str.upper))), ['STR'],
|
||||
msg='Transformation function should not raise')
|
||||
self.assertMaybeCountEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str_or_none))),
|
||||
[item for item in map(str_or_none, _TEST_DATA.values()) if item is not None],
|
||||
msg='Function in set should be a transformation')
|
||||
if __debug__:
|
||||
with self.assertRaises(Exception, msg='Sets with length != 1 should raise in debug'):
|
||||
traverse_obj(_TEST_DATA, set())
|
||||
with self.assertRaises(Exception, msg='Sets with length != 1 should raise in debug'):
|
||||
traverse_obj(_TEST_DATA, set((str.upper, str)))
|
||||
|
||||
# Test `slice` as a key
|
||||
_SLICE_DATA = [0, 1, 2, 3, 4]
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('dict', slice(1))), None,
|
||||
msg='slice on a dictionary should not throw')
|
||||
self.assertEqual(traverse_obj(_SLICE_DATA, slice(1)), _SLICE_DATA[:1],
|
||||
msg='slice key should apply slice to sequence')
|
||||
self.assertEqual(traverse_obj(_SLICE_DATA, slice(1, 2)), _SLICE_DATA[1:2],
|
||||
msg='slice key should apply slice to sequence')
|
||||
self.assertEqual(traverse_obj(_SLICE_DATA, slice(1, 4, 2)), _SLICE_DATA[1:4:2],
|
||||
msg='slice key should apply slice to sequence')
|
||||
|
||||
# Test alternative paths
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 'fail', 'str'), 'str',
|
||||
msg='multiple `paths` should be treated as alternative paths')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 'str', 100), 'str',
|
||||
msg='alternatives should exit early')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 'fail', 'fail'), None,
|
||||
msg='alternatives should return `default` if exhausted')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, 'fail'), 100), 100,
|
||||
msg='alternatives should track their own branching return')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('dict', Ellipsis), ('data', Ellipsis)), list(_TEST_DATA['data']),
|
||||
msg='alternatives on empty objects should search further')
|
||||
|
||||
# Test branch and path nesting
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('urls', (3, 0), 'url')), ['https://www.example.com/0'],
|
||||
msg='tuple as key should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('urls', [3, 0], 'url')), ['https://www.example.com/0'],
|
||||
msg='list as key should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('urls', ((1, 'fail'), (0, 'url')))), ['https://www.example.com/0'],
|
||||
msg='double nesting in path should be treated as paths')
|
||||
self.assertEqual(traverse_obj(['0', [1, 2]], [(0, 1), 0]), [1],
|
||||
msg='do not fail early on branching')
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, ('urls', ((1, ('fail', 'url')), (0, 'url')))),
|
||||
['https://www.example.com/0', 'https://www.example.com/1'],
|
||||
msg='triple nesting in path should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('urls', ('fail', (Ellipsis, 'url')))),
|
||||
['https://www.example.com/0', 'https://www.example.com/1'],
|
||||
msg='ellipsis as branch path start gets flattened')
|
||||
|
||||
# Test dictionary as key
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2}), {0: 100, 1: 1.2},
|
||||
msg='dict key should result in a dict with the same keys')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', 0, 'url')}),
|
||||
{0: 'https://www.example.com/0'},
|
||||
msg='dict key should allow paths')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', (3, 0), 'url')}),
|
||||
{0: ['https://www.example.com/0']},
|
||||
msg='tuple in dict path should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', ((1, 'fail'), (0, 'url')))}),
|
||||
{0: ['https://www.example.com/0']},
|
||||
msg='double nesting in dict path should be treated as paths')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', ((1, ('fail', 'url')), (0, 'url')))}),
|
||||
{0: ['https://www.example.com/1', 'https://www.example.com/0']},
|
||||
msg='triple nesting in dict path should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 'fail'}), {},
|
||||
msg='remove `None` values when top level dict key fails')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 'fail'}, default=Ellipsis), {0: Ellipsis},
|
||||
msg='use `default` if key fails and `default`')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 'dict'}), {},
|
||||
msg='remove empty values when dict key')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 'dict'}, default=Ellipsis), {0: Ellipsis},
|
||||
msg='use `default` when dict key and a default')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 'fail'}}), {},
|
||||
msg='remove empty values when nested dict key fails')
|
||||
self.assertEqual(traverse_obj(None, {0: 'fail'}), {},
|
||||
msg='default to dict if pruned')
|
||||
self.assertEqual(traverse_obj(None, {0: 'fail'}, default=Ellipsis), {0: Ellipsis},
|
||||
msg='default to dict if pruned and default is given')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 'fail'}}, default=Ellipsis), {0: {0: Ellipsis}},
|
||||
msg='use nested `default` when nested dict key fails and `default`')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('dict', Ellipsis)}), {},
|
||||
msg='remove key if branch in dict key not successful')
|
||||
|
||||
# Testing default parameter behavior
|
||||
_DEFAULT_DATA = {'None': None, 'int': 0, 'list': []}
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail'), None,
|
||||
msg='default value should be `None`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail', 'fail', default=Ellipsis), Ellipsis,
|
||||
msg='chained fails should result in default')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'None', 'int'), 0,
|
||||
msg='should not short cirquit on `None`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail', default=1), 1,
|
||||
msg='invalid dict key should result in `default`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'None', default=1), 1,
|
||||
msg='`None` is a deliberate sentinel and should become `default`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, ('list', 10)), None,
|
||||
msg='`IndexError` should result in `default`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail'), default=1), 1,
|
||||
msg='if branched but not successful return `default` if defined, not `[]`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail'), default=None), None,
|
||||
msg='if branched but not successful return `default` even if `default` is `None`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail')), [],
|
||||
msg='if branched but not successful return `[]`, not `default`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, ('list', Ellipsis)), [],
|
||||
msg='if branched but object is empty return `[]`, not `default`')
|
||||
self.assertEqual(traverse_obj(None, Ellipsis), [],
|
||||
msg='if branched but object is `None` return `[]`, not `default`')
|
||||
self.assertEqual(traverse_obj({0: None}, (0, Ellipsis)), [],
|
||||
msg='if branched but state is `None` return `[]`, not `default`')
|
||||
|
||||
branching_paths = [
|
||||
('fail', Ellipsis),
|
||||
(Ellipsis, 'fail'),
|
||||
100 * ('fail',) + (Ellipsis,),
|
||||
(Ellipsis,) + 100 * ('fail',),
|
||||
]
|
||||
for branching_path in branching_paths:
|
||||
self.assertEqual(traverse_obj({}, branching_path), [],
|
||||
msg='if branched but state is `None`, return `[]` (not `default`)')
|
||||
self.assertEqual(traverse_obj({}, 'fail', branching_path), [],
|
||||
msg='if branching in last alternative and previous did not match, return `[]` (not `default`)')
|
||||
self.assertEqual(traverse_obj({0: 'x'}, 0, branching_path), 'x',
|
||||
msg='if branching in last alternative and previous did match, return single value')
|
||||
self.assertEqual(traverse_obj({0: 'x'}, branching_path, 0), 'x',
|
||||
msg='if branching in first alternative and non-branching path does match, return single value')
|
||||
self.assertEqual(traverse_obj({}, branching_path, 'fail'), None,
|
||||
msg='if branching in first alternative and non-branching path does not match, return `default`')
|
||||
|
||||
# Testing expected_type behavior
|
||||
_EXPECTED_TYPE_DATA = {'str': 'str', 'int': 0}
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=str),
|
||||
'str', msg='accept matching `expected_type` type')
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=int),
|
||||
None, msg='reject non-matching `expected_type` type')
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'int', expected_type=lambda x: str(x)),
|
||||
'0', msg='transform type using type function')
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=lambda _: 1 / 0),
|
||||
None, msg='wrap expected_type function in try_call')
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, Ellipsis, expected_type=str),
|
||||
['str'], msg='eliminate items that expected_type fails on')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2}, expected_type=int),
|
||||
{0: 100}, msg='type as expected_type should filter dict values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2, 2: 'None'}, expected_type=str_or_none),
|
||||
{0: '100', 1: '1.2'}, msg='function as expected_type should transform dict values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ({0: 1.2}, 0, set((int_or_none,))), expected_type=int),
|
||||
1, msg='expected_type should not filter non-final dict values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 100, 1: 'str'}}, expected_type=int),
|
||||
{0: {0: 100}}, msg='expected_type should transform deep dict values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [({0: '...'}, {0: '...'})], expected_type=type(Ellipsis)),
|
||||
[{0: Ellipsis}, {0: Ellipsis}], msg='expected_type should transform branched dict values')
|
||||
self.assertEqual(traverse_obj({1: {3: 4}}, [(1, 2), 3], expected_type=int),
|
||||
[4], msg='expected_type regression for type matching in tuple branching')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ['data', Ellipsis], expected_type=int),
|
||||
[], msg='expected_type regression for type matching in dict result')
|
||||
|
||||
# Test get_all behavior
|
||||
_GET_ALL_DATA = {'key': [0, 1, 2]}
|
||||
self.assertEqual(traverse_obj(_GET_ALL_DATA, ('key', Ellipsis), get_all=False), 0,
|
||||
msg='if not `get_all`, return only first matching value')
|
||||
self.assertEqual(traverse_obj(_GET_ALL_DATA, Ellipsis, get_all=False), [0, 1, 2],
|
||||
msg='do not overflatten if not `get_all`')
|
||||
|
||||
# Test casesense behavior
|
||||
_CASESENSE_DATA = {
|
||||
'KeY': 'value0',
|
||||
0: {
|
||||
'KeY': 'value1',
|
||||
0: {'KeY': 'value2'},
|
||||
},
|
||||
# FULLWIDTH LATIN CAPITAL LETTER K
|
||||
'\uff2bey': 'value3',
|
||||
}
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, 'key'), None,
|
||||
msg='dict keys should be case sensitive unless `casesense`')
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, 'keY',
|
||||
casesense=False), 'value0',
|
||||
msg='allow non matching key case if `casesense`')
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, '\uff4bey', # FULLWIDTH LATIN SMALL LETTER K
|
||||
casesense=False), 'value3',
|
||||
msg='allow non matching Unicode key case if `casesense`')
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, (0, ('keY',)),
|
||||
casesense=False), ['value1'],
|
||||
msg='allow non matching key case in branch if `casesense`')
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, (0, ((0, 'keY'),)),
|
||||
casesense=False), ['value2'],
|
||||
msg='allow non matching key case in branch path if `casesense`')
|
||||
|
||||
# Test traverse_string behavior
|
||||
_TRAVERSE_STRING_DATA = {'str': 'str', 1.2: 1.2}
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0)), None,
|
||||
msg='do not traverse into string if not `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0),
|
||||
_traverse_string=True), 's',
|
||||
msg='traverse into string if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, (1.2, 1),
|
||||
_traverse_string=True), '.',
|
||||
msg='traverse into converted data if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', Ellipsis),
|
||||
_traverse_string=True), 'str',
|
||||
msg='`...` should result in string (same value) if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', slice(0, None, 2)),
|
||||
_traverse_string=True), 'sr',
|
||||
msg='`slice` should result in string if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', lambda i, v: i or v == 's'),
|
||||
_traverse_string=True), 'str',
|
||||
msg='function should result in string if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', (0, 2)),
|
||||
_traverse_string=True), ['s', 'r'],
|
||||
msg='branching should result in list if `traverse_string`')
|
||||
self.assertEqual(traverse_obj({}, (0, Ellipsis), _traverse_string=True), [],
|
||||
msg='branching should result in list if `traverse_string`')
|
||||
self.assertEqual(traverse_obj({}, (0, lambda x, y: True), _traverse_string=True), [],
|
||||
msg='branching should result in list if `traverse_string`')
|
||||
self.assertEqual(traverse_obj({}, (0, slice(1)), _traverse_string=True), [],
|
||||
msg='branching should result in list if `traverse_string`')
|
||||
|
||||
# Test re.Match as input obj
|
||||
mobj = re.match(r'^0(12)(?P<group>3)(4)?$', '0123')
|
||||
self.assertEqual(traverse_obj(mobj, Ellipsis), [x for x in mobj.groups() if x is not None],
|
||||
msg='`...` on a `re.Match` should give its `groups()`')
|
||||
self.assertEqual(traverse_obj(mobj, lambda k, _: k in (0, 2)), ['0123', '3'],
|
||||
msg='function on a `re.Match` should give groupno, value starting at 0')
|
||||
self.assertEqual(traverse_obj(mobj, 'group'), '3',
|
||||
msg='str key on a `re.Match` should give group with that name')
|
||||
self.assertEqual(traverse_obj(mobj, 2), '3',
|
||||
msg='int key on a `re.Match` should give group with that name')
|
||||
self.assertEqual(traverse_obj(mobj, 'gRoUp', casesense=False), '3',
|
||||
msg='str key on a `re.Match` should respect casesense')
|
||||
self.assertEqual(traverse_obj(mobj, 'fail'), None,
|
||||
msg='failing str key on a `re.Match` should return `default`')
|
||||
self.assertEqual(traverse_obj(mobj, 'gRoUpS', casesense=False), None,
|
||||
msg='failing str key on a `re.Match` should return `default`')
|
||||
self.assertEqual(traverse_obj(mobj, 8), None,
|
||||
msg='failing int key on a `re.Match` should return `default`')
|
||||
self.assertEqual(traverse_obj(mobj, lambda k, _: k in (0, 'group')), ['0123', '3'],
|
||||
msg='function on a `re.Match` should give group name as well')
|
||||
|
||||
# Test xml.etree.ElementTree.Element as input obj
|
||||
etree = compat_etree_fromstring('''<?xml version="1.0"?>
|
||||
<data>
|
||||
<country name="Liechtenstein">
|
||||
<rank>1</rank>
|
||||
<year>2008</year>
|
||||
<gdppc>141100</gdppc>
|
||||
<neighbor name="Austria" direction="E"/>
|
||||
<neighbor name="Switzerland" direction="W"/>
|
||||
</country>
|
||||
<country name="Singapore">
|
||||
<rank>4</rank>
|
||||
<year>2011</year>
|
||||
<gdppc>59900</gdppc>
|
||||
<neighbor name="Malaysia" direction="N"/>
|
||||
</country>
|
||||
<country name="Panama">
|
||||
<rank>68</rank>
|
||||
<year>2011</year>
|
||||
<gdppc>13600</gdppc>
|
||||
<neighbor name="Costa Rica" direction="W"/>
|
||||
<neighbor name="Colombia" direction="E"/>
|
||||
</country>
|
||||
</data>''')
|
||||
self.assertEqual(traverse_obj(etree, ''), etree,
|
||||
msg='empty str key should return the element itself')
|
||||
self.assertEqual(traverse_obj(etree, 'country'), list(etree),
|
||||
msg='str key should return all children with that tag name')
|
||||
self.assertEqual(traverse_obj(etree, Ellipsis), list(etree),
|
||||
msg='`...` as key should return all children')
|
||||
self.assertEqual(traverse_obj(etree, lambda _, x: x[0].text == '4'), [etree[1]],
|
||||
msg='function as key should get element as value')
|
||||
self.assertEqual(traverse_obj(etree, lambda i, _: i == 1), [etree[1]],
|
||||
msg='function as key should get index as key')
|
||||
self.assertEqual(traverse_obj(etree, 0), etree[0],
|
||||
msg='int key should return the nth child')
|
||||
self.assertEqual(traverse_obj(etree, './/neighbor/@name'),
|
||||
['Austria', 'Switzerland', 'Malaysia', 'Costa Rica', 'Colombia'],
|
||||
msg='`@<attribute>` at end of path should give that attribute')
|
||||
self.assertEqual(traverse_obj(etree, '//neighbor/@fail'), [None, None, None, None, None],
|
||||
msg='`@<nonexistent>` at end of path should give `None`')
|
||||
self.assertEqual(traverse_obj(etree, ('//neighbor/@', 2)), {'name': 'Malaysia', 'direction': 'N'},
|
||||
msg='`@` should give the full attribute dict')
|
||||
self.assertEqual(traverse_obj(etree, '//year/text()'), ['2008', '2011', '2011'],
|
||||
msg='`text()` at end of path should give the inner text')
|
||||
self.assertEqual(traverse_obj(etree, '//*[@direction]/@direction'), ['E', 'W', 'N', 'W', 'E'],
|
||||
msg='full python xpath features should be supported')
|
||||
self.assertEqual(traverse_obj(etree, (0, '@name')), 'Liechtenstein',
|
||||
msg='special transformations should act on current element')
|
||||
self.assertEqual(traverse_obj(etree, ('country', 0, Ellipsis, 'text()', T(int_or_none))), [1, 2008, 141100],
|
||||
msg='special transformations should act on current element')
|
||||
|
||||
def test_traversal_unbranching(self):
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [(100, 1.2), all]), [100, 1.2],
|
||||
msg='`all` should give all results as list')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [(100, 1.2), any]), 100,
|
||||
msg='`any` should give the first result')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [100, all]), [100],
|
||||
msg='`all` should give list if non branching')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [100, any]), 100,
|
||||
msg='`any` should give single item if non branching')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [('dict', 'None', 100), all]), [100],
|
||||
msg='`all` should filter `None` and empty dict')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [('dict', 'None', 100), any]), 100,
|
||||
msg='`any` should filter `None` and empty dict')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [{
|
||||
'all': [('dict', 'None', 100, 1.2), all],
|
||||
'any': [('dict', 'None', 100, 1.2), any],
|
||||
}]), {'all': [100, 1.2], 'any': 100},
|
||||
msg='`all`/`any` should apply to each dict path separately')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [{
|
||||
'all': [('dict', 'None', 100, 1.2), all],
|
||||
'any': [('dict', 'None', 100, 1.2), any],
|
||||
}], get_all=False), {'all': [100, 1.2], 'any': 100},
|
||||
msg='`all`/`any` should apply to dict regardless of `get_all`')
|
||||
self.assertIs(traverse_obj(_TEST_DATA, [('dict', 'None', 100, 1.2), all, T(float)]), None,
|
||||
msg='`all` should reset branching status')
|
||||
self.assertIs(traverse_obj(_TEST_DATA, [('dict', 'None', 100, 1.2), any, T(float)]), None,
|
||||
msg='`any` should reset branching status')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [('dict', 'None', 100, 1.2), all, Ellipsis, T(float)]), [1.2],
|
||||
msg='`all` should allow further branching')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [('dict', 'None', 'urls', 'data'), any, Ellipsis, 'index']), [0, 1],
|
||||
msg='`any` should allow further branching')
|
||||
|
||||
def test_traversal_morsel(self):
|
||||
values = {
|
||||
'expires': 'a',
|
||||
'path': 'b',
|
||||
'comment': 'c',
|
||||
'domain': 'd',
|
||||
'max-age': 'e',
|
||||
'secure': 'f',
|
||||
'httponly': 'g',
|
||||
'version': 'h',
|
||||
'samesite': 'i',
|
||||
}
|
||||
# SameSite added in Py3.8, breaks .update for 3.5-3.7
|
||||
if sys.version_info < (3, 8):
|
||||
del values['samesite']
|
||||
morsel = compat_http_cookies.Morsel()
|
||||
morsel.set(str('item_key'), 'item_value', 'coded_value')
|
||||
morsel.update(values)
|
||||
values['key'] = str('item_key')
|
||||
values['value'] = 'item_value'
|
||||
values = dict((str(k), v) for k, v in values.items())
|
||||
# make test pass even without ordered dict
|
||||
value_set = set(values.values())
|
||||
|
||||
for key, value in values.items():
|
||||
self.assertEqual(traverse_obj(morsel, key), value,
|
||||
msg='Morsel should provide access to all values')
|
||||
self.assertEqual(set(traverse_obj(morsel, Ellipsis)), value_set,
|
||||
msg='`...` should yield all values')
|
||||
self.assertEqual(set(traverse_obj(morsel, lambda k, v: True)), value_set,
|
||||
msg='function key should yield all values')
|
||||
self.assertIs(traverse_obj(morsel, [(None,), any]), morsel,
|
||||
msg='Morsel should not be implicitly changed to dict on usage')
|
||||
|
||||
def test_get_first(self):
|
||||
self.assertEqual(get_first([{'a': None}, {'a': 'spam'}], 'a'), 'spam')
|
||||
|
||||
def test_dict_get(self):
|
||||
FALSE_VALUES = {
|
||||
'none': None,
|
||||
'false': False,
|
||||
'zero': 0,
|
||||
'empty_string': '',
|
||||
'empty_list': [],
|
||||
}
|
||||
d = FALSE_VALUES.copy()
|
||||
d['a'] = 42
|
||||
self.assertEqual(dict_get(d, 'a'), 42)
|
||||
self.assertEqual(dict_get(d, 'b'), None)
|
||||
self.assertEqual(dict_get(d, 'b', 42), 42)
|
||||
self.assertEqual(dict_get(d, ('a', )), 42)
|
||||
self.assertEqual(dict_get(d, ('b', 'a', )), 42)
|
||||
self.assertEqual(dict_get(d, ('b', 'c', 'a', 'd', )), 42)
|
||||
self.assertEqual(dict_get(d, ('b', 'c', )), None)
|
||||
self.assertEqual(dict_get(d, ('b', 'c', ), 42), 42)
|
||||
for key, false_value in FALSE_VALUES.items():
|
||||
self.assertEqual(dict_get(d, ('b', 'c', key, )), None)
|
||||
self.assertEqual(dict_get(d, ('b', 'c', key, ), skip_false_values=False), false_value)
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
unittest.main()
|
||||
@ -14,6 +14,7 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||
import io
|
||||
import itertools
|
||||
import json
|
||||
import re
|
||||
import xml.etree.ElementTree
|
||||
|
||||
from youtube_dl.utils import (
|
||||
@ -27,6 +28,7 @@ from youtube_dl.utils import (
|
||||
DateRange,
|
||||
detect_exe_version,
|
||||
determine_ext,
|
||||
dict_get,
|
||||
encode_base_n,
|
||||
encode_compat_str,
|
||||
encodeFilename,
|
||||
@ -42,6 +44,7 @@ from youtube_dl.utils import (
|
||||
get_element_by_attribute,
|
||||
get_elements_by_class,
|
||||
get_elements_by_attribute,
|
||||
get_first,
|
||||
InAdvancePagedList,
|
||||
int_or_none,
|
||||
intlist_to_bytes,
|
||||
@ -81,11 +84,14 @@ from youtube_dl.utils import (
|
||||
sanitized_Request,
|
||||
shell_quote,
|
||||
smuggle_url,
|
||||
str_or_none,
|
||||
str_to_int,
|
||||
strip_jsonp,
|
||||
strip_or_none,
|
||||
subtitles_filename,
|
||||
T,
|
||||
timeconvert,
|
||||
traverse_obj,
|
||||
try_call,
|
||||
unescapeHTML,
|
||||
unified_strdate,
|
||||
@ -126,6 +132,10 @@ from youtube_dl.compat import (
|
||||
|
||||
class TestUtil(unittest.TestCase):
|
||||
|
||||
# yt-dlp shim
|
||||
def assertCountEqual(self, expected, got, msg='count should be the same'):
|
||||
return self.assertEqual(len(tuple(expected)), len(tuple(got)), msg=msg)
|
||||
|
||||
def test_timeconvert(self):
|
||||
self.assertTrue(timeconvert('') is None)
|
||||
self.assertTrue(timeconvert('bougrg') is None)
|
||||
@ -730,6 +740,28 @@ class TestUtil(unittest.TestCase):
|
||||
self.assertRaises(
|
||||
ValueError, multipart_encode, {b'field': b'value'}, boundary='value')
|
||||
|
||||
def test_dict_get(self):
|
||||
FALSE_VALUES = {
|
||||
'none': None,
|
||||
'false': False,
|
||||
'zero': 0,
|
||||
'empty_string': '',
|
||||
'empty_list': [],
|
||||
}
|
||||
d = FALSE_VALUES.copy()
|
||||
d['a'] = 42
|
||||
self.assertEqual(dict_get(d, 'a'), 42)
|
||||
self.assertEqual(dict_get(d, 'b'), None)
|
||||
self.assertEqual(dict_get(d, 'b', 42), 42)
|
||||
self.assertEqual(dict_get(d, ('a', )), 42)
|
||||
self.assertEqual(dict_get(d, ('b', 'a', )), 42)
|
||||
self.assertEqual(dict_get(d, ('b', 'c', 'a', 'd', )), 42)
|
||||
self.assertEqual(dict_get(d, ('b', 'c', )), None)
|
||||
self.assertEqual(dict_get(d, ('b', 'c', ), 42), 42)
|
||||
for key, false_value in FALSE_VALUES.items():
|
||||
self.assertEqual(dict_get(d, ('b', 'c', key, )), None)
|
||||
self.assertEqual(dict_get(d, ('b', 'c', key, ), skip_false_values=False), false_value)
|
||||
|
||||
def test_merge_dicts(self):
|
||||
self.assertEqual(merge_dicts({'a': 1}, {'b': 2}), {'a': 1, 'b': 2})
|
||||
self.assertEqual(merge_dicts({'a': 1}, {'a': 2}), {'a': 1})
|
||||
@ -1671,6 +1703,336 @@ Line 1
|
||||
self.assertEqual(variadic('spam', allowed_types=dict), 'spam')
|
||||
self.assertEqual(variadic('spam', allowed_types=[dict]), 'spam')
|
||||
|
||||
def test_traverse_obj(self):
|
||||
str = compat_str
|
||||
_TEST_DATA = {
|
||||
100: 100,
|
||||
1.2: 1.2,
|
||||
'str': 'str',
|
||||
'None': None,
|
||||
'...': Ellipsis,
|
||||
'urls': [
|
||||
{'index': 0, 'url': 'https://www.example.com/0'},
|
||||
{'index': 1, 'url': 'https://www.example.com/1'},
|
||||
],
|
||||
'data': (
|
||||
{'index': 2},
|
||||
{'index': 3},
|
||||
),
|
||||
'dict': {},
|
||||
}
|
||||
|
||||
# define a pukka Iterable
|
||||
def iter_range(stop):
|
||||
for from_ in range(stop):
|
||||
yield from_
|
||||
|
||||
# Test base functionality
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('str',)), 'str',
|
||||
msg='allow tuple path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ['str']), 'str',
|
||||
msg='allow list path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (value for value in ("str",))), 'str',
|
||||
msg='allow iterable path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 'str'), 'str',
|
||||
msg='single items should be treated as a path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, None), _TEST_DATA)
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 100), 100)
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 1.2), 1.2)
|
||||
|
||||
# Test Ellipsis behavior
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, Ellipsis),
|
||||
(item for item in _TEST_DATA.values() if item not in (None, {})),
|
||||
msg='`...` should give all non-discarded values')
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, ('urls', 0, Ellipsis)), _TEST_DATA['urls'][0].values(),
|
||||
msg='`...` selection for dicts should select all values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, Ellipsis, 'url')),
|
||||
['https://www.example.com/0', 'https://www.example.com/1'],
|
||||
msg='nested `...` queries should work')
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, (Ellipsis, Ellipsis, 'index')), iter_range(4),
|
||||
msg='`...` query result should be flattened')
|
||||
self.assertEqual(traverse_obj(iter(range(4)), Ellipsis), list(range(4)),
|
||||
msg='`...` should accept iterables')
|
||||
|
||||
# Test function as key
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, lambda x, y: x == 'urls' and isinstance(y, list)),
|
||||
[_TEST_DATA['urls']],
|
||||
msg='function as query key should perform a filter based on (key, value)')
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, lambda _, x: isinstance(x[0], str)), set(('str',)),
|
||||
msg='exceptions in the query function should be caught')
|
||||
self.assertEqual(traverse_obj(iter(range(4)), lambda _, x: x % 2 == 0), [0, 2],
|
||||
msg='function key should accept iterables')
|
||||
if __debug__:
|
||||
with self.assertRaises(Exception, msg='Wrong function signature should raise in debug'):
|
||||
traverse_obj(_TEST_DATA, lambda a: Ellipsis)
|
||||
with self.assertRaises(Exception, msg='Wrong function signature should raise in debug'):
|
||||
traverse_obj(_TEST_DATA, lambda a, b, c: Ellipsis)
|
||||
|
||||
# Test set as key (transformation/type, like `expected_type`)
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str.upper), )), ['STR'],
|
||||
msg='Function in set should be a transformation')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str))), ['str'],
|
||||
msg='Type in set should be a type filter')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, T(dict)), _TEST_DATA,
|
||||
msg='A single set should be wrapped into a path')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str.upper))), ['STR'],
|
||||
msg='Transformation function should not raise')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str_or_none))),
|
||||
[item for item in map(str_or_none, _TEST_DATA.values()) if item is not None],
|
||||
msg='Function in set should be a transformation')
|
||||
if __debug__:
|
||||
with self.assertRaises(Exception, msg='Sets with length != 1 should raise in debug'):
|
||||
traverse_obj(_TEST_DATA, set())
|
||||
with self.assertRaises(Exception, msg='Sets with length != 1 should raise in debug'):
|
||||
traverse_obj(_TEST_DATA, set((str.upper, str)))
|
||||
|
||||
# Test `slice` as a key
|
||||
_SLICE_DATA = [0, 1, 2, 3, 4]
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('dict', slice(1))), None,
|
||||
msg='slice on a dictionary should not throw')
|
||||
self.assertEqual(traverse_obj(_SLICE_DATA, slice(1)), _SLICE_DATA[:1],
|
||||
msg='slice key should apply slice to sequence')
|
||||
self.assertEqual(traverse_obj(_SLICE_DATA, slice(1, 2)), _SLICE_DATA[1:2],
|
||||
msg='slice key should apply slice to sequence')
|
||||
self.assertEqual(traverse_obj(_SLICE_DATA, slice(1, 4, 2)), _SLICE_DATA[1:4:2],
|
||||
msg='slice key should apply slice to sequence')
|
||||
|
||||
# Test alternative paths
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 'fail', 'str'), 'str',
|
||||
msg='multiple `paths` should be treated as alternative paths')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 'str', 100), 'str',
|
||||
msg='alternatives should exit early')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, 'fail', 'fail'), None,
|
||||
msg='alternatives should return `default` if exhausted')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, 'fail'), 100), 100,
|
||||
msg='alternatives should track their own branching return')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('dict', Ellipsis), ('data', Ellipsis)), list(_TEST_DATA['data']),
|
||||
msg='alternatives on empty objects should search further')
|
||||
|
||||
# Test branch and path nesting
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('urls', (3, 0), 'url')), ['https://www.example.com/0'],
|
||||
msg='tuple as key should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('urls', [3, 0], 'url')), ['https://www.example.com/0'],
|
||||
msg='list as key should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('urls', ((1, 'fail'), (0, 'url')))), ['https://www.example.com/0'],
|
||||
msg='double nesting in path should be treated as paths')
|
||||
self.assertEqual(traverse_obj(['0', [1, 2]], [(0, 1), 0]), [1],
|
||||
msg='do not fail early on branching')
|
||||
self.assertCountEqual(traverse_obj(_TEST_DATA, ('urls', ((1, ('fail', 'url')), (0, 'url')))),
|
||||
['https://www.example.com/0', 'https://www.example.com/1'],
|
||||
msg='triple nesting in path should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ('urls', ('fail', (Ellipsis, 'url')))),
|
||||
['https://www.example.com/0', 'https://www.example.com/1'],
|
||||
msg='ellipsis as branch path start gets flattened')
|
||||
|
||||
# Test dictionary as key
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2}), {0: 100, 1: 1.2},
|
||||
msg='dict key should result in a dict with the same keys')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', 0, 'url')}),
|
||||
{0: 'https://www.example.com/0'},
|
||||
msg='dict key should allow paths')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', (3, 0), 'url')}),
|
||||
{0: ['https://www.example.com/0']},
|
||||
msg='tuple in dict path should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', ((1, 'fail'), (0, 'url')))}),
|
||||
{0: ['https://www.example.com/0']},
|
||||
msg='double nesting in dict path should be treated as paths')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', ((1, ('fail', 'url')), (0, 'url')))}),
|
||||
{0: ['https://www.example.com/1', 'https://www.example.com/0']},
|
||||
msg='triple nesting in dict path should be treated as branches')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 'fail'}), {},
|
||||
msg='remove `None` values when top level dict key fails')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 'fail'}, default=Ellipsis), {0: Ellipsis},
|
||||
msg='use `default` if key fails and `default`')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 'dict'}), {},
|
||||
msg='remove empty values when dict key')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 'dict'}, default=Ellipsis), {0: Ellipsis},
|
||||
msg='use `default` when dict key and a default')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 'fail'}}), {},
|
||||
msg='remove empty values when nested dict key fails')
|
||||
self.assertEqual(traverse_obj(None, {0: 'fail'}), {},
|
||||
msg='default to dict if pruned')
|
||||
self.assertEqual(traverse_obj(None, {0: 'fail'}, default=Ellipsis), {0: Ellipsis},
|
||||
msg='default to dict if pruned and default is given')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 'fail'}}, default=Ellipsis), {0: {0: Ellipsis}},
|
||||
msg='use nested `default` when nested dict key fails and `default`')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: ('dict', Ellipsis)}), {},
|
||||
msg='remove key if branch in dict key not successful')
|
||||
|
||||
# Testing default parameter behavior
|
||||
_DEFAULT_DATA = {'None': None, 'int': 0, 'list': []}
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail'), None,
|
||||
msg='default value should be `None`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail', 'fail', default=Ellipsis), Ellipsis,
|
||||
msg='chained fails should result in default')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'None', 'int'), 0,
|
||||
msg='should not short cirquit on `None`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail', default=1), 1,
|
||||
msg='invalid dict key should result in `default`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, 'None', default=1), 1,
|
||||
msg='`None` is a deliberate sentinel and should become `default`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, ('list', 10)), None,
|
||||
msg='`IndexError` should result in `default`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail'), default=1), 1,
|
||||
msg='if branched but not successful return `default` if defined, not `[]`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail'), default=None), None,
|
||||
msg='if branched but not successful return `default` even if `default` is `None`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail')), [],
|
||||
msg='if branched but not successful return `[]`, not `default`')
|
||||
self.assertEqual(traverse_obj(_DEFAULT_DATA, ('list', Ellipsis)), [],
|
||||
msg='if branched but object is empty return `[]`, not `default`')
|
||||
self.assertEqual(traverse_obj(None, Ellipsis), [],
|
||||
msg='if branched but object is `None` return `[]`, not `default`')
|
||||
self.assertEqual(traverse_obj({0: None}, (0, Ellipsis)), [],
|
||||
msg='if branched but state is `None` return `[]`, not `default`')
|
||||
|
||||
branching_paths = [
|
||||
('fail', Ellipsis),
|
||||
(Ellipsis, 'fail'),
|
||||
100 * ('fail',) + (Ellipsis,),
|
||||
(Ellipsis,) + 100 * ('fail',),
|
||||
]
|
||||
for branching_path in branching_paths:
|
||||
self.assertEqual(traverse_obj({}, branching_path), [],
|
||||
msg='if branched but state is `None`, return `[]` (not `default`)')
|
||||
self.assertEqual(traverse_obj({}, 'fail', branching_path), [],
|
||||
msg='if branching in last alternative and previous did not match, return `[]` (not `default`)')
|
||||
self.assertEqual(traverse_obj({0: 'x'}, 0, branching_path), 'x',
|
||||
msg='if branching in last alternative and previous did match, return single value')
|
||||
self.assertEqual(traverse_obj({0: 'x'}, branching_path, 0), 'x',
|
||||
msg='if branching in first alternative and non-branching path does match, return single value')
|
||||
self.assertEqual(traverse_obj({}, branching_path, 'fail'), None,
|
||||
msg='if branching in first alternative and non-branching path does not match, return `default`')
|
||||
|
||||
# Testing expected_type behavior
|
||||
_EXPECTED_TYPE_DATA = {'str': 'str', 'int': 0}
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=str),
|
||||
'str', msg='accept matching `expected_type` type')
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=int),
|
||||
None, msg='reject non-matching `expected_type` type')
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'int', expected_type=lambda x: str(x)),
|
||||
'0', msg='transform type using type function')
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=lambda _: 1 / 0),
|
||||
None, msg='wrap expected_type function in try_call')
|
||||
self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, Ellipsis, expected_type=str),
|
||||
['str'], msg='eliminate items that expected_type fails on')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2}, expected_type=int),
|
||||
{0: 100}, msg='type as expected_type should filter dict values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2, 2: 'None'}, expected_type=str_or_none),
|
||||
{0: '100', 1: '1.2'}, msg='function as expected_type should transform dict values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ({0: 1.2}, 0, set((int_or_none,))), expected_type=int),
|
||||
1, msg='expected_type should not filter non-final dict values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 100, 1: 'str'}}, expected_type=int),
|
||||
{0: {0: 100}}, msg='expected_type should transform deep dict values')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, [({0: '...'}, {0: '...'})], expected_type=type(Ellipsis)),
|
||||
[{0: Ellipsis}, {0: Ellipsis}], msg='expected_type should transform branched dict values')
|
||||
self.assertEqual(traverse_obj({1: {3: 4}}, [(1, 2), 3], expected_type=int),
|
||||
[4], msg='expected_type regression for type matching in tuple branching')
|
||||
self.assertEqual(traverse_obj(_TEST_DATA, ['data', Ellipsis], expected_type=int),
|
||||
[], msg='expected_type regression for type matching in dict result')
|
||||
|
||||
# Test get_all behavior
|
||||
_GET_ALL_DATA = {'key': [0, 1, 2]}
|
||||
self.assertEqual(traverse_obj(_GET_ALL_DATA, ('key', Ellipsis), get_all=False), 0,
|
||||
msg='if not `get_all`, return only first matching value')
|
||||
self.assertEqual(traverse_obj(_GET_ALL_DATA, Ellipsis, get_all=False), [0, 1, 2],
|
||||
msg='do not overflatten if not `get_all`')
|
||||
|
||||
# Test casesense behavior
|
||||
_CASESENSE_DATA = {
|
||||
'KeY': 'value0',
|
||||
0: {
|
||||
'KeY': 'value1',
|
||||
0: {'KeY': 'value2'},
|
||||
},
|
||||
# FULLWIDTH LATIN CAPITAL LETTER K
|
||||
'\uff2bey': 'value3',
|
||||
}
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, 'key'), None,
|
||||
msg='dict keys should be case sensitive unless `casesense`')
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, 'keY',
|
||||
casesense=False), 'value0',
|
||||
msg='allow non matching key case if `casesense`')
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, '\uff4bey', # FULLWIDTH LATIN SMALL LETTER K
|
||||
casesense=False), 'value3',
|
||||
msg='allow non matching Unicode key case if `casesense`')
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, (0, ('keY',)),
|
||||
casesense=False), ['value1'],
|
||||
msg='allow non matching key case in branch if `casesense`')
|
||||
self.assertEqual(traverse_obj(_CASESENSE_DATA, (0, ((0, 'keY'),)),
|
||||
casesense=False), ['value2'],
|
||||
msg='allow non matching key case in branch path if `casesense`')
|
||||
|
||||
# Test traverse_string behavior
|
||||
_TRAVERSE_STRING_DATA = {'str': 'str', 1.2: 1.2}
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0)), None,
|
||||
msg='do not traverse into string if not `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0),
|
||||
_traverse_string=True), 's',
|
||||
msg='traverse into string if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, (1.2, 1),
|
||||
_traverse_string=True), '.',
|
||||
msg='traverse into converted data if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', Ellipsis),
|
||||
_traverse_string=True), 'str',
|
||||
msg='`...` should result in string (same value) if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', slice(0, None, 2)),
|
||||
_traverse_string=True), 'sr',
|
||||
msg='`slice` should result in string if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', lambda i, v: i or v == 's'),
|
||||
_traverse_string=True), 'str',
|
||||
msg='function should result in string if `traverse_string`')
|
||||
self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', (0, 2)),
|
||||
_traverse_string=True), ['s', 'r'],
|
||||
msg='branching should result in list if `traverse_string`')
|
||||
self.assertEqual(traverse_obj({}, (0, Ellipsis), _traverse_string=True), [],
|
||||
msg='branching should result in list if `traverse_string`')
|
||||
self.assertEqual(traverse_obj({}, (0, lambda x, y: True), _traverse_string=True), [],
|
||||
msg='branching should result in list if `traverse_string`')
|
||||
self.assertEqual(traverse_obj({}, (0, slice(1)), _traverse_string=True), [],
|
||||
msg='branching should result in list if `traverse_string`')
|
||||
|
||||
# Test is_user_input behavior
|
||||
_IS_USER_INPUT_DATA = {'range8': list(range(8))}
|
||||
self.assertEqual(traverse_obj(_IS_USER_INPUT_DATA, ('range8', '3'),
|
||||
_is_user_input=True), 3,
|
||||
msg='allow for string indexing if `is_user_input`')
|
||||
self.assertCountEqual(traverse_obj(_IS_USER_INPUT_DATA, ('range8', '3:'),
|
||||
_is_user_input=True), tuple(range(8))[3:],
|
||||
msg='allow for string slice if `is_user_input`')
|
||||
self.assertCountEqual(traverse_obj(_IS_USER_INPUT_DATA, ('range8', ':4:2'),
|
||||
_is_user_input=True), tuple(range(8))[:4:2],
|
||||
msg='allow step in string slice if `is_user_input`')
|
||||
self.assertCountEqual(traverse_obj(_IS_USER_INPUT_DATA, ('range8', ':'),
|
||||
_is_user_input=True), range(8),
|
||||
msg='`:` should be treated as `...` if `is_user_input`')
|
||||
with self.assertRaises(TypeError, msg='too many params should result in error'):
|
||||
traverse_obj(_IS_USER_INPUT_DATA, ('range8', ':::'), _is_user_input=True)
|
||||
|
||||
# Test re.Match as input obj
|
||||
mobj = re.match(r'^0(12)(?P<group>3)(4)?$', '0123')
|
||||
self.assertEqual(traverse_obj(mobj, Ellipsis), [x for x in mobj.groups() if x is not None],
|
||||
msg='`...` on a `re.Match` should give its `groups()`')
|
||||
self.assertEqual(traverse_obj(mobj, lambda k, _: k in (0, 2)), ['0123', '3'],
|
||||
msg='function on a `re.Match` should give groupno, value starting at 0')
|
||||
self.assertEqual(traverse_obj(mobj, 'group'), '3',
|
||||
msg='str key on a `re.Match` should give group with that name')
|
||||
self.assertEqual(traverse_obj(mobj, 2), '3',
|
||||
msg='int key on a `re.Match` should give group with that name')
|
||||
self.assertEqual(traverse_obj(mobj, 'gRoUp', casesense=False), '3',
|
||||
msg='str key on a `re.Match` should respect casesense')
|
||||
self.assertEqual(traverse_obj(mobj, 'fail'), None,
|
||||
msg='failing str key on a `re.Match` should return `default`')
|
||||
self.assertEqual(traverse_obj(mobj, 'gRoUpS', casesense=False), None,
|
||||
msg='failing str key on a `re.Match` should return `default`')
|
||||
self.assertEqual(traverse_obj(mobj, 8), None,
|
||||
msg='failing int key on a `re.Match` should return `default`')
|
||||
self.assertEqual(traverse_obj(mobj, lambda k, _: k in (0, 'group')), ['0123', '3'],
|
||||
msg='function on a `re.Match` should give group name as well')
|
||||
|
||||
def test_get_first(self):
|
||||
self.assertEqual(get_first([{'a': None}, {'a': 'spam'}], 'a'), 'spam')
|
||||
|
||||
def test_join_nonempty(self):
|
||||
self.assertEqual(join_nonempty('a', 'b'), 'a-b')
|
||||
self.assertEqual(join_nonempty(
|
||||
|
||||
@ -2719,222 +2719,8 @@ if sys.version_info < (2, 7):
|
||||
if isinstance(xpath, compat_str):
|
||||
xpath = xpath.encode('ascii')
|
||||
return xpath
|
||||
|
||||
# further code below based on CPython 2.7 source
|
||||
import functools
|
||||
|
||||
_xpath_tokenizer_re = re.compile(r'''(?x)
|
||||
( # (1)
|
||||
'[^']*'|"[^"]*"| # quoted strings, or
|
||||
::|//?|\.\.|\(\)|[/.*:[\]()@=] # navigation specials
|
||||
)| # or (2)
|
||||
((?:\{[^}]+\})?[^/[\]()@=\s]+)| # token: optional {ns}, no specials
|
||||
\s+ # or white space
|
||||
''')
|
||||
|
||||
def _xpath_tokenizer(pattern, namespaces=None):
|
||||
for token in _xpath_tokenizer_re.findall(pattern):
|
||||
tag = token[1]
|
||||
if tag and tag[0] != "{" and ":" in tag:
|
||||
try:
|
||||
if not namespaces:
|
||||
raise KeyError
|
||||
prefix, uri = tag.split(":", 1)
|
||||
yield token[0], "{%s}%s" % (namespaces[prefix], uri)
|
||||
except KeyError:
|
||||
raise SyntaxError("prefix %r not found in prefix map" % prefix)
|
||||
else:
|
||||
yield token
|
||||
|
||||
def _get_parent_map(context):
|
||||
parent_map = context.parent_map
|
||||
if parent_map is None:
|
||||
context.parent_map = parent_map = {}
|
||||
for p in context.root.getiterator():
|
||||
for e in p:
|
||||
parent_map[e] = p
|
||||
return parent_map
|
||||
|
||||
def _select(context, result, filter_fn=lambda *_: True):
|
||||
for elem in result:
|
||||
for e in elem:
|
||||
if filter_fn(e, elem):
|
||||
yield e
|
||||
|
||||
def _prepare_child(next_, token):
|
||||
tag = token[1]
|
||||
return functools.partial(_select, filter_fn=lambda e, _: e.tag == tag)
|
||||
|
||||
def _prepare_star(next_, token):
|
||||
return _select
|
||||
|
||||
def _prepare_self(next_, token):
|
||||
return lambda _, result: (e for e in result)
|
||||
|
||||
def _prepare_descendant(next_, token):
|
||||
token = next(next_)
|
||||
if token[0] == "*":
|
||||
tag = "*"
|
||||
elif not token[0]:
|
||||
tag = token[1]
|
||||
else:
|
||||
raise SyntaxError("invalid descendant")
|
||||
|
||||
def select(context, result):
|
||||
for elem in result:
|
||||
for e in elem.getiterator(tag):
|
||||
if e is not elem:
|
||||
yield e
|
||||
return select
|
||||
|
||||
def _prepare_parent(next_, token):
|
||||
def select(context, result):
|
||||
# FIXME: raise error if .. is applied at toplevel?
|
||||
parent_map = _get_parent_map(context)
|
||||
result_map = {}
|
||||
for elem in result:
|
||||
if elem in parent_map:
|
||||
parent = parent_map[elem]
|
||||
if parent not in result_map:
|
||||
result_map[parent] = None
|
||||
yield parent
|
||||
return select
|
||||
|
||||
def _prepare_predicate(next_, token):
|
||||
signature = []
|
||||
predicate = []
|
||||
for token in next_:
|
||||
if token[0] == "]":
|
||||
break
|
||||
if token[0] and token[0][:1] in "'\"":
|
||||
token = "'", token[0][1:-1]
|
||||
signature.append(token[0] or "-")
|
||||
predicate.append(token[1])
|
||||
|
||||
def select(context, result, filter_fn=lambda _: True):
|
||||
for elem in result:
|
||||
if filter_fn(elem):
|
||||
yield elem
|
||||
|
||||
signature = "".join(signature)
|
||||
# use signature to determine predicate type
|
||||
if signature == "@-":
|
||||
# [@attribute] predicate
|
||||
key = predicate[1]
|
||||
return functools.partial(
|
||||
select, filter_fn=lambda el: el.get(key) is not None)
|
||||
if signature == "@-='":
|
||||
# [@attribute='value']
|
||||
key = predicate[1]
|
||||
value = predicate[-1]
|
||||
return functools.partial(
|
||||
select, filter_fn=lambda el: el.get(key) == value)
|
||||
if signature == "-" and not re.match(r"\d+$", predicate[0]):
|
||||
# [tag]
|
||||
tag = predicate[0]
|
||||
return functools.partial(
|
||||
select, filter_fn=lambda el: el.find(tag) is not None)
|
||||
if signature == "-='" and not re.match(r"\d+$", predicate[0]):
|
||||
# [tag='value']
|
||||
tag = predicate[0]
|
||||
value = predicate[-1]
|
||||
|
||||
def itertext(el):
|
||||
for e in el.getiterator():
|
||||
e = e.text
|
||||
if e:
|
||||
yield e
|
||||
|
||||
def select(context, result):
|
||||
for elem in result:
|
||||
for e in elem.findall(tag):
|
||||
if "".join(itertext(e)) == value:
|
||||
yield elem
|
||||
break
|
||||
return select
|
||||
if signature == "-" or signature == "-()" or signature == "-()-":
|
||||
# [index] or [last()] or [last()-index]
|
||||
if signature == "-":
|
||||
index = int(predicate[0]) - 1
|
||||
else:
|
||||
if predicate[0] != "last":
|
||||
raise SyntaxError("unsupported function")
|
||||
if signature == "-()-":
|
||||
try:
|
||||
index = int(predicate[2]) - 1
|
||||
except ValueError:
|
||||
raise SyntaxError("unsupported expression")
|
||||
else:
|
||||
index = -1
|
||||
|
||||
def select(context, result):
|
||||
parent_map = _get_parent_map(context)
|
||||
for elem in result:
|
||||
try:
|
||||
parent = parent_map[elem]
|
||||
# FIXME: what if the selector is "*" ?
|
||||
elems = list(parent.findall(elem.tag))
|
||||
if elems[index] is elem:
|
||||
yield elem
|
||||
except (IndexError, KeyError):
|
||||
pass
|
||||
return select
|
||||
raise SyntaxError("invalid predicate")
|
||||
|
||||
ops = {
|
||||
"": _prepare_child,
|
||||
"*": _prepare_star,
|
||||
".": _prepare_self,
|
||||
"..": _prepare_parent,
|
||||
"//": _prepare_descendant,
|
||||
"[": _prepare_predicate,
|
||||
}
|
||||
|
||||
_cache = {}
|
||||
|
||||
class _SelectorContext:
|
||||
parent_map = None
|
||||
|
||||
def __init__(self, root):
|
||||
self.root = root
|
||||
|
||||
##
|
||||
# Generate all matching objects.
|
||||
|
||||
def compat_etree_iterfind(elem, path, namespaces=None):
|
||||
# compile selector pattern
|
||||
if path[-1:] == "/":
|
||||
path = path + "*" # implicit all (FIXME: keep this?)
|
||||
try:
|
||||
selector = _cache[path]
|
||||
except KeyError:
|
||||
if len(_cache) > 100:
|
||||
_cache.clear()
|
||||
if path[:1] == "/":
|
||||
raise SyntaxError("cannot use absolute path on element")
|
||||
tokens = _xpath_tokenizer(path, namespaces)
|
||||
selector = []
|
||||
for token in tokens:
|
||||
if token[0] == "/":
|
||||
continue
|
||||
try:
|
||||
selector.append(ops[token[0]](tokens, token))
|
||||
except StopIteration:
|
||||
raise SyntaxError("invalid path")
|
||||
_cache[path] = selector
|
||||
# execute selector pattern
|
||||
result = [elem]
|
||||
context = _SelectorContext(elem)
|
||||
for select in selector:
|
||||
result = select(context, result)
|
||||
return result
|
||||
|
||||
# end of code based on CPython 2.7 source
|
||||
|
||||
|
||||
else:
|
||||
compat_xpath = lambda xpath: xpath
|
||||
compat_etree_iterfind = lambda element, match: element.iterfind(match)
|
||||
|
||||
|
||||
compat_os_name = os._name if os.name == 'java' else os.name
|
||||
@ -2970,7 +2756,7 @@ except (AssertionError, UnicodeEncodeError):
|
||||
|
||||
|
||||
def compat_ord(c):
|
||||
if isinstance(c, int):
|
||||
if type(c) is int:
|
||||
return c
|
||||
else:
|
||||
return ord(c)
|
||||
@ -3169,7 +2955,7 @@ except ImportError:
|
||||
return self
|
||||
|
||||
def __exit__(self, exc_type, exc_val, exc_tb):
|
||||
return exc_type is not None and issubclass(exc_type, self._exceptions or tuple())
|
||||
return exc_val is not None and isinstance(exc_val, self._exceptions or tuple())
|
||||
|
||||
|
||||
# subprocess.Popen context manager
|
||||
@ -3522,7 +3308,6 @@ __all__ = [
|
||||
'compat_contextlib_suppress',
|
||||
'compat_ctypes_WINFUNCTYPE',
|
||||
'compat_etree_fromstring',
|
||||
'compat_etree_iterfind',
|
||||
'compat_filter',
|
||||
'compat_get_terminal_size',
|
||||
'compat_getenv',
|
||||
|
||||
@ -1169,10 +1169,10 @@ class InfoExtractor(object):
|
||||
def _get_netrc_login_info(self, netrc_machine=None):
|
||||
username = None
|
||||
password = None
|
||||
netrc_machine = netrc_machine or self._NETRC_MACHINE
|
||||
|
||||
if self._downloader.params.get('usenetrc', False):
|
||||
try:
|
||||
netrc_machine = netrc_machine or self._NETRC_MACHINE
|
||||
info = netrc.netrc().authenticators(netrc_machine)
|
||||
if info is not None:
|
||||
username = info[0]
|
||||
@ -1180,7 +1180,7 @@ class InfoExtractor(object):
|
||||
else:
|
||||
raise netrc.NetrcParseError(
|
||||
'No authenticators for %s' % netrc_machine)
|
||||
except (AttributeError, IOError, netrc.NetrcParseError) as err:
|
||||
except (IOError, netrc.NetrcParseError) as err:
|
||||
self._downloader.report_warning(
|
||||
'parsing .netrc: %s' % error_to_compat_str(err))
|
||||
|
||||
@ -1490,18 +1490,14 @@ class InfoExtractor(object):
|
||||
return dict((k, v) for k, v in info.items() if v is not None)
|
||||
|
||||
def _search_nextjs_data(self, webpage, video_id, **kw):
|
||||
# ..., *, transform_source=None, fatal=True, default=NO_DEFAULT
|
||||
|
||||
# TODO: remove this backward compat
|
||||
default = kw.get('default', NO_DEFAULT)
|
||||
if default == '{}':
|
||||
kw['default'] = {}
|
||||
kw = compat_kwargs(kw)
|
||||
|
||||
return self._search_json(
|
||||
r'''<script\s[^>]*?\bid\s*=\s*('|")__NEXT_DATA__\1[^>]*>''',
|
||||
webpage, 'next.js data', video_id, end_pattern='</script>',
|
||||
**kw)
|
||||
nkw = dict((k, v) for k, v in kw.items() if k in ('transform_source', 'fatal'))
|
||||
kw.pop('transform_source', None)
|
||||
next_data = self._search_regex(
|
||||
r'''<script[^>]+\bid\s*=\s*('|")__NEXT_DATA__\1[^>]*>(?P<nd>[^<]+)</script>''',
|
||||
webpage, 'next.js data', group='nd', **kw)
|
||||
if not next_data:
|
||||
return {}
|
||||
return self._parse_json(next_data, video_id, **nkw)
|
||||
|
||||
def _search_nuxt_data(self, webpage, video_id, *args, **kwargs):
|
||||
"""Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""
|
||||
@ -3300,16 +3296,12 @@ class InfoExtractor(object):
|
||||
return ret
|
||||
|
||||
@classmethod
|
||||
def _merge_subtitles(cls, subtitle_dict1, *subtitle_dicts, **kwargs):
|
||||
""" Merge subtitle dictionaries, language by language. """
|
||||
|
||||
# ..., * , target=None
|
||||
target = kwargs.get('target') or dict(subtitle_dict1)
|
||||
|
||||
for subtitle_dict in subtitle_dicts:
|
||||
for lang in subtitle_dict:
|
||||
target[lang] = cls._merge_subtitle_items(target.get(lang, []), subtitle_dict[lang])
|
||||
return target
|
||||
def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2):
|
||||
""" Merge two subtitle dictionaries, language by language. """
|
||||
ret = dict(subtitle_dict1)
|
||||
for lang in subtitle_dict2:
|
||||
ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang])
|
||||
return ret
|
||||
|
||||
def extract_automatic_captions(self, *args, **kwargs):
|
||||
if (self._downloader.params.get('writeautomaticsub', False)
|
||||
@ -3342,29 +3334,6 @@ class InfoExtractor(object):
|
||||
def _generic_title(self, url):
|
||||
return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
|
||||
|
||||
def _yes_playlist(self, playlist_id, video_id, *args, **kwargs):
|
||||
# smuggled_data=None, *, playlist_label='playlist', video_label='video'
|
||||
smuggled_data = args[0] if len(args) == 1 else kwargs.get('smuggled_data')
|
||||
playlist_label = kwargs.get('playlist_label', 'playlist')
|
||||
video_label = kwargs.get('video_label', 'video')
|
||||
|
||||
if not playlist_id or not video_id:
|
||||
return not video_id
|
||||
|
||||
no_playlist = (smuggled_data or {}).get('force_noplaylist')
|
||||
if no_playlist is not None:
|
||||
return not no_playlist
|
||||
|
||||
video_id = '' if video_id is True else ' ' + video_id
|
||||
noplaylist = self.get_param('noplaylist')
|
||||
self.to_screen(
|
||||
'Downloading just the {0}{1} because of --no-playlist'.format(video_label, video_id)
|
||||
if noplaylist else
|
||||
'Downloading {0}{1} - add --no-playlist to download just the {2}{3}'.format(
|
||||
playlist_label, '' if playlist_id is True else ' ' + playlist_id,
|
||||
video_label, video_id))
|
||||
return not noplaylist
|
||||
|
||||
|
||||
class SearchInfoExtractor(InfoExtractor):
|
||||
"""
|
||||
|
||||
@ -1,10 +0,0 @@
|
||||
# coding: utf-8
|
||||
|
||||
# TODO: move these utils.fns here and move import to utils
|
||||
# flake8: noqa
|
||||
from .utils import (
|
||||
dict_get,
|
||||
get_first,
|
||||
T,
|
||||
traverse_obj,
|
||||
)
|
||||
@ -49,14 +49,11 @@ from .compat import (
|
||||
compat_cookiejar,
|
||||
compat_ctypes_WINFUNCTYPE,
|
||||
compat_datetime_timedelta_total_seconds,
|
||||
compat_etree_Element,
|
||||
compat_etree_fromstring,
|
||||
compat_etree_iterfind,
|
||||
compat_expanduser,
|
||||
compat_html_entities,
|
||||
compat_html_entities_html5,
|
||||
compat_http_client,
|
||||
compat_http_cookies,
|
||||
compat_integer_types,
|
||||
compat_kwargs,
|
||||
compat_ncompress as ncompress,
|
||||
@ -6256,16 +6253,15 @@ if __debug__:
|
||||
|
||||
def traverse_obj(obj, *paths, **kwargs):
|
||||
"""
|
||||
Safely traverse nested `dict`s and `Iterable`s, etc
|
||||
Safely traverse nested `dict`s and `Iterable`s
|
||||
|
||||
>>> obj = [{}, {"key": "value"}]
|
||||
>>> traverse_obj(obj, (1, "key"))
|
||||
'value'
|
||||
"value"
|
||||
|
||||
Each of the provided `paths` is tested and the first producing a valid result will be returned.
|
||||
The next path will also be tested if the path branched but no results could be found.
|
||||
Supported values for traversal are `Mapping`, `Iterable`, `re.Match`, `xml.etree.ElementTree`
|
||||
(xpath) and `http.cookies.Morsel`.
|
||||
Supported values for traversal are `Mapping`, `Iterable` and `re.Match`.
|
||||
Unhelpful values (`{}`, `None`) are treated as the absence of a value and discarded.
|
||||
|
||||
The paths will be wrapped in `variadic`, so that `'key'` is conveniently the same as `('key', )`.
|
||||
@ -6273,9 +6269,8 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
The keys in the path can be one of:
|
||||
- `None`: Return the current object.
|
||||
- `set`: Requires the only item in the set to be a type or function,
|
||||
like `{type}`/`{type, type, ...}`/`{func}`. If one or more `type`s,
|
||||
return only values that have one of the types. If a function,
|
||||
return `func(obj)`.
|
||||
like `{type}`/`{func}`. If a `type`, returns only values
|
||||
of this type. If a function, returns `func(obj)`.
|
||||
- `str`/`int`: Return `obj[key]`. For `re.Match`, return `obj.group(key)`.
|
||||
- `slice`: Branch out and return all values in `obj[key]`.
|
||||
- `Ellipsis`: Branch out and return a list of all values.
|
||||
@ -6287,10 +6282,8 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
For `Iterable`s, `key` is the enumeration count of the value.
|
||||
For `re.Match`es, `key` is the group number (0 = full match)
|
||||
as well as additionally any group names, if given.
|
||||
- `dict`: Transform the current object and return a matching dict.
|
||||
- `dict` Transform the current object and return a matching dict.
|
||||
Read as: `{key: traverse_obj(obj, path) for key, path in dct.items()}`.
|
||||
- `any`-builtin: Take the first matching object and return it, resetting branching.
|
||||
- `all`-builtin: Take all matching objects and return them as a list, resetting branching.
|
||||
|
||||
`tuple`, `list`, and `dict` all support nested paths and branches.
|
||||
|
||||
@ -6306,8 +6299,10 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
@param get_all If `False`, return the first matching result, otherwise all matching ones.
|
||||
@param casesense If `False`, consider string dictionary keys as case insensitive.
|
||||
|
||||
The following is only meant to be used by YoutubeDL.prepare_outtmpl and is not part of the API
|
||||
The following are only meant to be used by YoutubeDL.prepare_outtmpl and are not part of the API
|
||||
|
||||
@param _is_user_input Whether the keys are generated from user input.
|
||||
If `True` strings get converted to `int`/`slice` if needed.
|
||||
@param _traverse_string Whether to traverse into objects as strings.
|
||||
If `True`, any non-compatible object will first be
|
||||
converted into a string and then traversed into.
|
||||
@ -6327,6 +6322,7 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
expected_type = kwargs.get('expected_type')
|
||||
get_all = kwargs.get('get_all', True)
|
||||
casesense = kwargs.get('casesense', True)
|
||||
_is_user_input = kwargs.get('_is_user_input', False)
|
||||
_traverse_string = kwargs.get('_traverse_string', False)
|
||||
|
||||
# instant compat
|
||||
@ -6340,8 +6336,10 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
type_test = lambda val: try_call(expected_type or IDENTITY, args=(val,))
|
||||
|
||||
def lookup_or_none(v, k, getter=None):
|
||||
with compat_contextlib_suppress(LookupError):
|
||||
try:
|
||||
return getter(v, k) if getter else v[k]
|
||||
except IndexError:
|
||||
return None
|
||||
|
||||
def from_iterable(iterables):
|
||||
# chain.from_iterable(['ABC', 'DEF']) --> A B C D E F
|
||||
@ -6363,13 +6361,12 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
result = obj
|
||||
|
||||
elif isinstance(key, set):
|
||||
assert len(key) >= 1, 'At least one item is required in a `set` key'
|
||||
if all(isinstance(item, type) for item in key):
|
||||
result = obj if isinstance(obj, tuple(key)) else None
|
||||
else:
|
||||
assert len(key) == 1, 'Set should only be used to wrap a single item'
|
||||
item = next(iter(key))
|
||||
assert len(key) == 1, 'Multiple items in a `set` key must all be types'
|
||||
result = try_call(item, args=(obj,)) if not isinstance(item, type) else None
|
||||
if isinstance(item, type):
|
||||
result = obj if isinstance(obj, item) else None
|
||||
else:
|
||||
result = try_call(item, args=(obj,))
|
||||
|
||||
elif isinstance(key, (list, tuple)):
|
||||
branching = True
|
||||
@ -6378,11 +6375,9 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
|
||||
elif key is Ellipsis:
|
||||
branching = True
|
||||
if isinstance(obj, compat_http_cookies.Morsel):
|
||||
obj = dict(obj, key=obj.key, value=obj.value)
|
||||
if isinstance(obj, compat_collections_abc.Mapping):
|
||||
result = obj.values()
|
||||
elif is_iterable_like(obj, (compat_collections_abc.Iterable, compat_etree_Element)):
|
||||
elif is_iterable_like(obj):
|
||||
result = obj
|
||||
elif isinstance(obj, compat_re_Match):
|
||||
result = obj.groups()
|
||||
@ -6394,11 +6389,9 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
|
||||
elif callable(key):
|
||||
branching = True
|
||||
if isinstance(obj, compat_http_cookies.Morsel):
|
||||
obj = dict(obj, key=obj.key, value=obj.value)
|
||||
if isinstance(obj, compat_collections_abc.Mapping):
|
||||
iter_obj = obj.items()
|
||||
elif is_iterable_like(obj, (compat_collections_abc.Iterable, compat_etree_Element)):
|
||||
elif is_iterable_like(obj):
|
||||
iter_obj = enumerate(obj)
|
||||
elif isinstance(obj, compat_re_Match):
|
||||
iter_obj = itertools.chain(
|
||||
@ -6420,8 +6413,6 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
if v is not None or default is not NO_DEFAULT) or None
|
||||
|
||||
elif isinstance(obj, compat_collections_abc.Mapping):
|
||||
if isinstance(obj, compat_http_cookies.Morsel):
|
||||
obj = dict(obj, key=obj.key, value=obj.value)
|
||||
result = (try_call(obj.get, args=(key,))
|
||||
if casesense or try_call(obj.__contains__, args=(key,))
|
||||
else next((v for k, v in obj.items() if casefold(k) == key), None))
|
||||
@ -6439,40 +6430,12 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
else:
|
||||
result = None
|
||||
if isinstance(key, (int, slice)):
|
||||
if is_iterable_like(obj, (compat_collections_abc.Sequence, compat_etree_Element)):
|
||||
if is_iterable_like(obj, compat_collections_abc.Sequence):
|
||||
branching = isinstance(key, slice)
|
||||
result = lookup_or_none(obj, key)
|
||||
elif _traverse_string:
|
||||
result = lookup_or_none(str(obj), key)
|
||||
|
||||
elif isinstance(obj, compat_etree_Element) and isinstance(key, str):
|
||||
xpath, _, special = key.rpartition('/')
|
||||
if not special.startswith('@') and not special.endswith('()'):
|
||||
xpath = key
|
||||
special = None
|
||||
|
||||
# Allow abbreviations of relative paths, absolute paths error
|
||||
if xpath.startswith('/'):
|
||||
xpath = '.' + xpath
|
||||
elif xpath and not xpath.startswith('./'):
|
||||
xpath = './' + xpath
|
||||
|
||||
def apply_specials(element):
|
||||
if special is None:
|
||||
return element
|
||||
if special == '@':
|
||||
return element.attrib
|
||||
if special.startswith('@'):
|
||||
return try_call(element.attrib.get, args=(special[1:],))
|
||||
if special == 'text()':
|
||||
return element.text
|
||||
raise SyntaxError('apply_specials is missing case for {0!r}'.format(special))
|
||||
|
||||
if xpath:
|
||||
result = list(map(apply_specials, compat_etree_iterfind(obj, xpath)))
|
||||
else:
|
||||
result = apply_specials(obj)
|
||||
|
||||
return branching, result if branching else (result,)
|
||||
|
||||
def lazy_last(iterable):
|
||||
@ -6493,18 +6456,17 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
|
||||
key = None
|
||||
for last, key in lazy_last(variadic(path, (str, bytes, dict, set))):
|
||||
if _is_user_input and isinstance(key, str):
|
||||
if key == ':':
|
||||
key = Ellipsis
|
||||
elif ':' in key:
|
||||
key = slice(*map(int_or_none, key.split(':')))
|
||||
elif int_or_none(key) is not None:
|
||||
key = int(key)
|
||||
|
||||
if not casesense and isinstance(key, str):
|
||||
key = compat_casefold(key)
|
||||
|
||||
if key in (any, all):
|
||||
has_branched = False
|
||||
filtered_objs = (obj for obj in objs if obj not in (None, {}))
|
||||
if key is any:
|
||||
objs = (next(filtered_objs, None),)
|
||||
else:
|
||||
objs = (list(filtered_objs),)
|
||||
continue
|
||||
|
||||
if __debug__ and callable(key):
|
||||
# Verify function signature
|
||||
_try_bind_args(key, None, None)
|
||||
@ -6543,9 +6505,9 @@ def traverse_obj(obj, *paths, **kwargs):
|
||||
return None if default is NO_DEFAULT else default
|
||||
|
||||
|
||||
def T(*x):
|
||||
""" For use in yt-dl instead of {type, ...} or set((type, ...)) """
|
||||
return set(x)
|
||||
def T(x):
|
||||
""" For use in yt-dl instead of {type} or set((type,)) """
|
||||
return set((x,))
|
||||
|
||||
|
||||
def get_first(obj, keys, **kwargs):
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user