mirror of https://github.com/yt-dlp/yt-dlp
Merge branch 'master' into fix/youtube/remove-android-as-default
commit
d2911ea8bb
@ -0,0 +1,444 @@
|
||||
import http.cookies
|
||||
import re
|
||||
import xml.etree.ElementTree
|
||||
|
||||
import pytest
|
||||
|
||||
from yt_dlp.utils import dict_get, int_or_none, str_or_none
|
||||
from yt_dlp.utils.traversal import traverse_obj
|
||||
|
||||
_TEST_DATA = {
|
||||
100: 100,
|
||||
1.2: 1.2,
|
||||
'str': 'str',
|
||||
'None': None,
|
||||
'...': ...,
|
||||
'urls': [
|
||||
{'index': 0, 'url': 'https://www.example.com/0'},
|
||||
{'index': 1, 'url': 'https://www.example.com/1'},
|
||||
],
|
||||
'data': (
|
||||
{'index': 2},
|
||||
{'index': 3},
|
||||
),
|
||||
'dict': {},
|
||||
}
|
||||
|
||||
|
||||
class TestTraversal:
|
||||
def test_traversal_base(self):
|
||||
assert traverse_obj(_TEST_DATA, ('str',)) == 'str', \
|
||||
'allow tuple path'
|
||||
assert traverse_obj(_TEST_DATA, ['str']) == 'str', \
|
||||
'allow list path'
|
||||
assert traverse_obj(_TEST_DATA, (value for value in ("str",))) == 'str', \
|
||||
'allow iterable path'
|
||||
assert traverse_obj(_TEST_DATA, 'str') == 'str', \
|
||||
'single items should be treated as a path'
|
||||
assert traverse_obj(_TEST_DATA, 100) == 100, \
|
||||
'allow int path'
|
||||
assert traverse_obj(_TEST_DATA, 1.2) == 1.2, \
|
||||
'allow float path'
|
||||
assert traverse_obj(_TEST_DATA, None) == _TEST_DATA, \
|
||||
'`None` should not perform any modification'
|
||||
|
||||
def test_traversal_ellipsis(self):
|
||||
assert traverse_obj(_TEST_DATA, ...) == [x for x in _TEST_DATA.values() if x not in (None, {})], \
|
||||
'`...` should give all non discarded values'
|
||||
assert traverse_obj(_TEST_DATA, ('urls', 0, ...)) == list(_TEST_DATA['urls'][0].values()), \
|
||||
'`...` selection for dicts should select all values'
|
||||
assert traverse_obj(_TEST_DATA, (..., ..., 'url')) == ['https://www.example.com/0', 'https://www.example.com/1'], \
|
||||
'nested `...` queries should work'
|
||||
assert traverse_obj(_TEST_DATA, (..., ..., 'index')) == list(range(4)), \
|
||||
'`...` query result should be flattened'
|
||||
assert traverse_obj(iter(range(4)), ...) == list(range(4)), \
|
||||
'`...` should accept iterables'
|
||||
|
||||
def test_traversal_function(self):
|
||||
filter_func = lambda x, y: x == 'urls' and isinstance(y, list)
|
||||
assert traverse_obj(_TEST_DATA, filter_func) == [_TEST_DATA['urls']], \
|
||||
'function as query key should perform a filter based on (key, value)'
|
||||
assert traverse_obj(_TEST_DATA, lambda _, x: isinstance(x[0], str)) == ['str'], \
|
||||
'exceptions in the query function should be catched'
|
||||
assert traverse_obj(iter(range(4)), lambda _, x: x % 2 == 0) == [0, 2], \
|
||||
'function key should accept iterables'
|
||||
# Wrong function signature should raise (debug mode)
|
||||
with pytest.raises(Exception):
|
||||
traverse_obj(_TEST_DATA, lambda a: ...)
|
||||
with pytest.raises(Exception):
|
||||
traverse_obj(_TEST_DATA, lambda a, b, c: ...)
|
||||
|
||||
def test_traversal_set(self):
|
||||
# transformation/type, like `expected_type`
|
||||
assert traverse_obj(_TEST_DATA, (..., {str.upper}, )) == ['STR'], \
|
||||
'Function in set should be a transformation'
|
||||
assert traverse_obj(_TEST_DATA, (..., {str})) == ['str'], \
|
||||
'Type in set should be a type filter'
|
||||
assert traverse_obj(_TEST_DATA, (..., {str, int})) == [100, 'str'], \
|
||||
'Multiple types in set should be a type filter'
|
||||
assert traverse_obj(_TEST_DATA, {dict}) == _TEST_DATA, \
|
||||
'A single set should be wrapped into a path'
|
||||
assert traverse_obj(_TEST_DATA, (..., {str.upper})) == ['STR'], \
|
||||
'Transformation function should not raise'
|
||||
expected = [x for x in map(str_or_none, _TEST_DATA.values()) if x is not None]
|
||||
assert traverse_obj(_TEST_DATA, (..., {str_or_none})) == expected, \
|
||||
'Function in set should be a transformation'
|
||||
assert traverse_obj(_TEST_DATA, ('fail', {lambda _: 'const'})) == 'const', \
|
||||
'Function in set should always be called'
|
||||
# Sets with length < 1 or > 1 not including only types should raise
|
||||
with pytest.raises(Exception):
|
||||
traverse_obj(_TEST_DATA, set())
|
||||
with pytest.raises(Exception):
|
||||
traverse_obj(_TEST_DATA, {str.upper, str})
|
||||
|
||||
def test_traversal_slice(self):
|
||||
_SLICE_DATA = [0, 1, 2, 3, 4]
|
||||
|
||||
assert traverse_obj(_TEST_DATA, ('dict', slice(1))) is None, \
|
||||
'slice on a dictionary should not throw'
|
||||
assert traverse_obj(_SLICE_DATA, slice(1)) == _SLICE_DATA[:1], \
|
||||
'slice key should apply slice to sequence'
|
||||
assert traverse_obj(_SLICE_DATA, slice(1, 2)) == _SLICE_DATA[1:2], \
|
||||
'slice key should apply slice to sequence'
|
||||
assert traverse_obj(_SLICE_DATA, slice(1, 4, 2)) == _SLICE_DATA[1:4:2], \
|
||||
'slice key should apply slice to sequence'
|
||||
|
||||
def test_traversal_alternatives(self):
|
||||
assert traverse_obj(_TEST_DATA, 'fail', 'str') == 'str', \
|
||||
'multiple `paths` should be treated as alternative paths'
|
||||
assert traverse_obj(_TEST_DATA, 'str', 100) == 'str', \
|
||||
'alternatives should exit early'
|
||||
assert traverse_obj(_TEST_DATA, 'fail', 'fail') is None, \
|
||||
'alternatives should return `default` if exhausted'
|
||||
assert traverse_obj(_TEST_DATA, (..., 'fail'), 100) == 100, \
|
||||
'alternatives should track their own branching return'
|
||||
assert traverse_obj(_TEST_DATA, ('dict', ...), ('data', ...)) == list(_TEST_DATA['data']), \
|
||||
'alternatives on empty objects should search further'
|
||||
|
||||
def test_traversal_branching_nesting(self):
|
||||
assert traverse_obj(_TEST_DATA, ('urls', (3, 0), 'url')) == ['https://www.example.com/0'], \
|
||||
'tuple as key should be treated as branches'
|
||||
assert traverse_obj(_TEST_DATA, ('urls', [3, 0], 'url')) == ['https://www.example.com/0'], \
|
||||
'list as key should be treated as branches'
|
||||
assert traverse_obj(_TEST_DATA, ('urls', ((1, 'fail'), (0, 'url')))) == ['https://www.example.com/0'], \
|
||||
'double nesting in path should be treated as paths'
|
||||
assert traverse_obj(['0', [1, 2]], [(0, 1), 0]) == [1], \
|
||||
'do not fail early on branching'
|
||||
expected = ['https://www.example.com/0', 'https://www.example.com/1']
|
||||
assert traverse_obj(_TEST_DATA, ('urls', ((0, ('fail', 'url')), (1, 'url')))) == expected, \
|
||||
'tripple nesting in path should be treated as branches'
|
||||
assert traverse_obj(_TEST_DATA, ('urls', ('fail', (..., 'url')))) == expected, \
|
||||
'ellipsis as branch path start gets flattened'
|
||||
|
||||
def test_traversal_dict(self):
|
||||
assert traverse_obj(_TEST_DATA, {0: 100, 1: 1.2}) == {0: 100, 1: 1.2}, \
|
||||
'dict key should result in a dict with the same keys'
|
||||
expected = {0: 'https://www.example.com/0'}
|
||||
assert traverse_obj(_TEST_DATA, {0: ('urls', 0, 'url')}) == expected, \
|
||||
'dict key should allow paths'
|
||||
expected = {0: ['https://www.example.com/0']}
|
||||
assert traverse_obj(_TEST_DATA, {0: ('urls', (3, 0), 'url')}) == expected, \
|
||||
'tuple in dict path should be treated as branches'
|
||||
assert traverse_obj(_TEST_DATA, {0: ('urls', ((1, 'fail'), (0, 'url')))}) == expected, \
|
||||
'double nesting in dict path should be treated as paths'
|
||||
expected = {0: ['https://www.example.com/1', 'https://www.example.com/0']}
|
||||
assert traverse_obj(_TEST_DATA, {0: ('urls', ((1, ('fail', 'url')), (0, 'url')))}) == expected, \
|
||||
'tripple nesting in dict path should be treated as branches'
|
||||
assert traverse_obj(_TEST_DATA, {0: 'fail'}) == {}, \
|
||||
'remove `None` values when top level dict key fails'
|
||||
assert traverse_obj(_TEST_DATA, {0: 'fail'}, default=...) == {0: ...}, \
|
||||
'use `default` if key fails and `default`'
|
||||
assert traverse_obj(_TEST_DATA, {0: 'dict'}) == {}, \
|
||||
'remove empty values when dict key'
|
||||
assert traverse_obj(_TEST_DATA, {0: 'dict'}, default=...) == {0: ...}, \
|
||||
'use `default` when dict key and `default`'
|
||||
assert traverse_obj(_TEST_DATA, {0: {0: 'fail'}}) == {}, \
|
||||
'remove empty values when nested dict key fails'
|
||||
assert traverse_obj(None, {0: 'fail'}) == {}, \
|
||||
'default to dict if pruned'
|
||||
assert traverse_obj(None, {0: 'fail'}, default=...) == {0: ...}, \
|
||||
'default to dict if pruned and default is given'
|
||||
assert traverse_obj(_TEST_DATA, {0: {0: 'fail'}}, default=...) == {0: {0: ...}}, \
|
||||
'use nested `default` when nested dict key fails and `default`'
|
||||
assert traverse_obj(_TEST_DATA, {0: ('dict', ...)}) == {}, \
|
||||
'remove key if branch in dict key not successful'
|
||||
|
||||
def test_traversal_default(self):
|
||||
_DEFAULT_DATA = {'None': None, 'int': 0, 'list': []}
|
||||
|
||||
assert traverse_obj(_DEFAULT_DATA, 'fail') is None, \
|
||||
'default value should be `None`'
|
||||
assert traverse_obj(_DEFAULT_DATA, 'fail', 'fail', default=...) == ..., \
|
||||
'chained fails should result in default'
|
||||
assert traverse_obj(_DEFAULT_DATA, 'None', 'int') == 0, \
|
||||
'should not short cirquit on `None`'
|
||||
assert traverse_obj(_DEFAULT_DATA, 'fail', default=1) == 1, \
|
||||
'invalid dict key should result in `default`'
|
||||
assert traverse_obj(_DEFAULT_DATA, 'None', default=1) == 1, \
|
||||
'`None` is a deliberate sentinel and should become `default`'
|
||||
assert traverse_obj(_DEFAULT_DATA, ('list', 10)) is None, \
|
||||
'`IndexError` should result in `default`'
|
||||
assert traverse_obj(_DEFAULT_DATA, (..., 'fail'), default=1) == 1, \
|
||||
'if branched but not successful return `default` if defined, not `[]`'
|
||||
assert traverse_obj(_DEFAULT_DATA, (..., 'fail'), default=None) is None, \
|
||||
'if branched but not successful return `default` even if `default` is `None`'
|
||||
assert traverse_obj(_DEFAULT_DATA, (..., 'fail')) == [], \
|
||||
'if branched but not successful return `[]`, not `default`'
|
||||
assert traverse_obj(_DEFAULT_DATA, ('list', ...)) == [], \
|
||||
'if branched but object is empty return `[]`, not `default`'
|
||||
assert traverse_obj(None, ...) == [], \
|
||||
'if branched but object is `None` return `[]`, not `default`'
|
||||
assert traverse_obj({0: None}, (0, ...)) == [], \
|
||||
'if branched but state is `None` return `[]`, not `default`'
|
||||
|
||||
@pytest.mark.parametrize('path', [
|
||||
('fail', ...),
|
||||
(..., 'fail'),
|
||||
100 * ('fail',) + (...,),
|
||||
(...,) + 100 * ('fail',),
|
||||
])
|
||||
def test_traversal_branching(self, path):
|
||||
assert traverse_obj({}, path) == [], \
|
||||
'if branched but state is `None`, return `[]` (not `default`)'
|
||||
assert traverse_obj({}, 'fail', path) == [], \
|
||||
'if branching in last alternative and previous did not match, return `[]` (not `default`)'
|
||||
assert traverse_obj({0: 'x'}, 0, path) == 'x', \
|
||||
'if branching in last alternative and previous did match, return single value'
|
||||
assert traverse_obj({0: 'x'}, path, 0) == 'x', \
|
||||
'if branching in first alternative and non-branching path does match, return single value'
|
||||
assert traverse_obj({}, path, 'fail') is None, \
|
||||
'if branching in first alternative and non-branching path does not match, return `default`'
|
||||
|
||||
def test_traversal_expected_type(self):
|
||||
_EXPECTED_TYPE_DATA = {'str': 'str', 'int': 0}
|
||||
|
||||
assert traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=str) == 'str', \
|
||||
'accept matching `expected_type` type'
|
||||
assert traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=int) is None, \
|
||||
'reject non matching `expected_type` type'
|
||||
assert traverse_obj(_EXPECTED_TYPE_DATA, 'int', expected_type=lambda x: str(x)) == '0', \
|
||||
'transform type using type function'
|
||||
assert traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=lambda _: 1 / 0) is None, \
|
||||
'wrap expected_type fuction in try_call'
|
||||
assert traverse_obj(_EXPECTED_TYPE_DATA, ..., expected_type=str) == ['str'], \
|
||||
'eliminate items that expected_type fails on'
|
||||
assert traverse_obj(_TEST_DATA, {0: 100, 1: 1.2}, expected_type=int) == {0: 100}, \
|
||||
'type as expected_type should filter dict values'
|
||||
assert traverse_obj(_TEST_DATA, {0: 100, 1: 1.2, 2: 'None'}, expected_type=str_or_none) == {0: '100', 1: '1.2'}, \
|
||||
'function as expected_type should transform dict values'
|
||||
assert traverse_obj(_TEST_DATA, ({0: 1.2}, 0, {int_or_none}), expected_type=int) == 1, \
|
||||
'expected_type should not filter non final dict values'
|
||||
assert traverse_obj(_TEST_DATA, {0: {0: 100, 1: 'str'}}, expected_type=int) == {0: {0: 100}}, \
|
||||
'expected_type should transform deep dict values'
|
||||
assert traverse_obj(_TEST_DATA, [({0: '...'}, {0: '...'})], expected_type=type(...)) == [{0: ...}, {0: ...}], \
|
||||
'expected_type should transform branched dict values'
|
||||
assert traverse_obj({1: {3: 4}}, [(1, 2), 3], expected_type=int) == [4], \
|
||||
'expected_type regression for type matching in tuple branching'
|
||||
assert traverse_obj(_TEST_DATA, ['data', ...], expected_type=int) == [], \
|
||||
'expected_type regression for type matching in dict result'
|
||||
|
||||
def test_traversal_get_all(self):
|
||||
_GET_ALL_DATA = {'key': [0, 1, 2]}
|
||||
|
||||
assert traverse_obj(_GET_ALL_DATA, ('key', ...), get_all=False) == 0, \
|
||||
'if not `get_all`, return only first matching value'
|
||||
assert traverse_obj(_GET_ALL_DATA, ..., get_all=False) == [0, 1, 2], \
|
||||
'do not overflatten if not `get_all`'
|
||||
|
||||
def test_traversal_casesense(self):
|
||||
_CASESENSE_DATA = {
|
||||
'KeY': 'value0',
|
||||
0: {
|
||||
'KeY': 'value1',
|
||||
0: {'KeY': 'value2'},
|
||||
},
|
||||
}
|
||||
|
||||
assert traverse_obj(_CASESENSE_DATA, 'key') is None, \
|
||||
'dict keys should be case sensitive unless `casesense`'
|
||||
assert traverse_obj(_CASESENSE_DATA, 'keY', casesense=False) == 'value0', \
|
||||
'allow non matching key case if `casesense`'
|
||||
assert traverse_obj(_CASESENSE_DATA, [0, ('keY',)], casesense=False) == ['value1'], \
|
||||
'allow non matching key case in branch if `casesense`'
|
||||
assert traverse_obj(_CASESENSE_DATA, [0, ([0, 'keY'],)], casesense=False) == ['value2'], \
|
||||
'allow non matching key case in branch path if `casesense`'
|
||||
|
||||
def test_traversal_traverse_string(self):
|
||||
_TRAVERSE_STRING_DATA = {'str': 'str', 1.2: 1.2}
|
||||
|
||||
assert traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0)) is None, \
|
||||
'do not traverse into string if not `traverse_string`'
|
||||
assert traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0), traverse_string=True) == 's', \
|
||||
'traverse into string if `traverse_string`'
|
||||
assert traverse_obj(_TRAVERSE_STRING_DATA, (1.2, 1), traverse_string=True) == '.', \
|
||||
'traverse into converted data if `traverse_string`'
|
||||
assert traverse_obj(_TRAVERSE_STRING_DATA, ('str', ...), traverse_string=True) == 'str', \
|
||||
'`...` should result in string (same value) if `traverse_string`'
|
||||
assert traverse_obj(_TRAVERSE_STRING_DATA, ('str', slice(0, None, 2)), traverse_string=True) == 'sr', \
|
||||
'`slice` should result in string if `traverse_string`'
|
||||
assert traverse_obj(_TRAVERSE_STRING_DATA, ('str', lambda i, v: i or v == "s"), traverse_string=True) == 'str', \
|
||||
'function should result in string if `traverse_string`'
|
||||
assert traverse_obj(_TRAVERSE_STRING_DATA, ('str', (0, 2)), traverse_string=True) == ['s', 'r'], \
|
||||
'branching should result in list if `traverse_string`'
|
||||
assert traverse_obj({}, (0, ...), traverse_string=True) == [], \
|
||||
'branching should result in list if `traverse_string`'
|
||||
assert traverse_obj({}, (0, lambda x, y: True), traverse_string=True) == [], \
|
||||
'branching should result in list if `traverse_string`'
|
||||
assert traverse_obj({}, (0, slice(1)), traverse_string=True) == [], \
|
||||
'branching should result in list if `traverse_string`'
|
||||
|
||||
def test_traversal_re(self):
|
||||
mobj = re.fullmatch(r'0(12)(?P<group>3)(4)?', '0123')
|
||||
assert traverse_obj(mobj, ...) == [x for x in mobj.groups() if x is not None], \
|
||||
'`...` on a `re.Match` should give its `groups()`'
|
||||
assert traverse_obj(mobj, lambda k, _: k in (0, 2)) == ['0123', '3'], \
|
||||
'function on a `re.Match` should give groupno, value starting at 0'
|
||||
assert traverse_obj(mobj, 'group') == '3', \
|
||||
'str key on a `re.Match` should give group with that name'
|
||||
assert traverse_obj(mobj, 2) == '3', \
|
||||
'int key on a `re.Match` should give group with that name'
|
||||
assert traverse_obj(mobj, 'gRoUp', casesense=False) == '3', \
|
||||
'str key on a `re.Match` should respect casesense'
|
||||
assert traverse_obj(mobj, 'fail') is None, \
|
||||
'failing str key on a `re.Match` should return `default`'
|
||||
assert traverse_obj(mobj, 'gRoUpS', casesense=False) is None, \
|
||||
'failing str key on a `re.Match` should return `default`'
|
||||
assert traverse_obj(mobj, 8) is None, \
|
||||
'failing int key on a `re.Match` should return `default`'
|
||||
assert traverse_obj(mobj, lambda k, _: k in (0, 'group')) == ['0123', '3'], \
|
||||
'function on a `re.Match` should give group name as well'
|
||||
|
||||
def test_traversal_xml_etree(self):
|
||||
etree = xml.etree.ElementTree.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>''')
|
||||
assert traverse_obj(etree, '') == etree, \
|
||||
'empty str key should return the element itself'
|
||||
assert traverse_obj(etree, 'country') == list(etree), \
|
||||
'str key should lead all children with that tag name'
|
||||
assert traverse_obj(etree, ...) == list(etree), \
|
||||
'`...` as key should return all children'
|
||||
assert traverse_obj(etree, lambda _, x: x[0].text == '4') == [etree[1]], \
|
||||
'function as key should get element as value'
|
||||
assert traverse_obj(etree, lambda i, _: i == 1) == [etree[1]], \
|
||||
'function as key should get index as key'
|
||||
assert traverse_obj(etree, 0) == etree[0], \
|
||||
'int key should return the nth child'
|
||||
expected = ['Austria', 'Switzerland', 'Malaysia', 'Costa Rica', 'Colombia']
|
||||
assert traverse_obj(etree, './/neighbor/@name') == expected, \
|
||||
'`@<attribute>` at end of path should give that attribute'
|
||||
assert traverse_obj(etree, '//neighbor/@fail') == [None, None, None, None, None], \
|
||||
'`@<nonexistant>` at end of path should give `None`'
|
||||
assert traverse_obj(etree, ('//neighbor/@', 2)) == {'name': 'Malaysia', 'direction': 'N'}, \
|
||||
'`@` should give the full attribute dict'
|
||||
assert traverse_obj(etree, '//year/text()') == ['2008', '2011', '2011'], \
|
||||
'`text()` at end of path should give the inner text'
|
||||
assert traverse_obj(etree, '//*[@direction]/@direction') == ['E', 'W', 'N', 'W', 'E'], \
|
||||
'full Python xpath features should be supported'
|
||||
assert traverse_obj(etree, (0, '@name')) == 'Liechtenstein', \
|
||||
'special transformations should act on current element'
|
||||
assert traverse_obj(etree, ('country', 0, ..., 'text()', {int_or_none})) == [1, 2008, 141100], \
|
||||
'special transformations should act on current element'
|
||||
|
||||
def test_traversal_unbranching(self):
|
||||
assert traverse_obj(_TEST_DATA, [(100, 1.2), all]) == [100, 1.2], \
|
||||
'`all` should give all results as list'
|
||||
assert traverse_obj(_TEST_DATA, [(100, 1.2), any]) == 100, \
|
||||
'`any` should give the first result'
|
||||
assert traverse_obj(_TEST_DATA, [100, all]) == [100], \
|
||||
'`all` should give list if non branching'
|
||||
assert traverse_obj(_TEST_DATA, [100, any]) == 100, \
|
||||
'`any` should give single item if non branching'
|
||||
assert traverse_obj(_TEST_DATA, [('dict', 'None', 100), all]) == [100], \
|
||||
'`all` should filter `None` and empty dict'
|
||||
assert traverse_obj(_TEST_DATA, [('dict', 'None', 100), any]) == 100, \
|
||||
'`any` should filter `None` and empty dict'
|
||||
assert traverse_obj(_TEST_DATA, [{
|
||||
'all': [('dict', 'None', 100, 1.2), all],
|
||||
'any': [('dict', 'None', 100, 1.2), any],
|
||||
}]) == {'all': [100, 1.2], 'any': 100}, \
|
||||
'`all`/`any` should apply to each dict path separately'
|
||||
assert 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}, \
|
||||
'`all`/`any` should apply to dict regardless of `get_all`'
|
||||
assert traverse_obj(_TEST_DATA, [('dict', 'None', 100, 1.2), all, {float}]) is None, \
|
||||
'`all` should reset branching status'
|
||||
assert traverse_obj(_TEST_DATA, [('dict', 'None', 100, 1.2), any, {float}]) is None, \
|
||||
'`any` should reset branching status'
|
||||
assert traverse_obj(_TEST_DATA, [('dict', 'None', 100, 1.2), all, ..., {float}]) == [1.2], \
|
||||
'`all` should allow further branching'
|
||||
assert traverse_obj(_TEST_DATA, [('dict', 'None', 'urls', 'data'), any, ..., 'index']) == [0, 1], \
|
||||
'`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',
|
||||
}
|
||||
morsel = http.cookies.Morsel()
|
||||
morsel.set('item_key', 'item_value', 'coded_value')
|
||||
morsel.update(values)
|
||||
values['key'] = 'item_key'
|
||||
values['value'] = 'item_value'
|
||||
|
||||
for key, value in values.items():
|
||||
assert traverse_obj(morsel, key) == value, \
|
||||
'Morsel should provide access to all values'
|
||||
assert traverse_obj(morsel, ...) == list(values.values()), \
|
||||
'`...` should yield all values'
|
||||
assert traverse_obj(morsel, lambda k, v: True) == list(values.values()), \
|
||||
'function key should yield all values'
|
||||
assert traverse_obj(morsel, [(None,), any]) == morsel, \
|
||||
'Morsel should not be implicitly changed to dict on usage'
|
||||
|
||||
|
||||
class TestDictGet:
|
||||
def test_dict_get(self):
|
||||
FALSE_VALUES = {
|
||||
'none': None,
|
||||
'false': False,
|
||||
'zero': 0,
|
||||
'empty_string': '',
|
||||
'empty_list': [],
|
||||
}
|
||||
d = {**FALSE_VALUES, 'a': 42}
|
||||
assert dict_get(d, 'a') == 42
|
||||
assert dict_get(d, 'b') is None
|
||||
assert dict_get(d, 'b', 42) == 42
|
||||
assert dict_get(d, ('a',)) == 42
|
||||
assert dict_get(d, ('b', 'a')) == 42
|
||||
assert dict_get(d, ('b', 'c', 'a', 'd')) == 42
|
||||
assert dict_get(d, ('b', 'c')) is None
|
||||
assert dict_get(d, ('b', 'c'), 42) == 42
|
||||
for key, false_value in FALSE_VALUES.items():
|
||||
assert dict_get(d, ('b', 'c', key)) is None
|
||||
assert dict_get(d, ('b', 'c', key), skip_false_values=False) == false_value
|
@ -1,67 +1,153 @@
|
||||
import urllib.parse
|
||||
|
||||
from .common import InfoExtractor
|
||||
from ..utils import (
|
||||
unified_strdate,
|
||||
update_url_query,
|
||||
urlencode_postdata,
|
||||
filter_dict,
|
||||
parse_iso8601,
|
||||
traverse_obj,
|
||||
try_call,
|
||||
url_or_none,
|
||||
)
|
||||
|
||||
|
||||
class MediciIE(InfoExtractor):
|
||||
_VALID_URL = r'https?://(?:www\.)?medici\.tv/#!/(?P<id>[^?#&]+)'
|
||||
_TEST = {
|
||||
'url': 'http://www.medici.tv/#!/daniel-harding-frans-helmerson-verbier-festival-music-camp',
|
||||
'md5': '004c21bb0a57248085b6ff3fec72719d',
|
||||
_VALID_URL = r'https?://(?:(?P<sub>www|edu)\.)?medici\.tv/[a-z]{2}/[\w.-]+/(?P<id>[^/?#&]+)'
|
||||
_TESTS = [{
|
||||
'url': 'https://www.medici.tv/en/operas/thomas-ades-the-exterminating-angel-calixto-bieito-opera-bastille-paris',
|
||||
'md5': 'd483f74e7a7a9eac0dbe152ab189050d',
|
||||
'info_dict': {
|
||||
'id': '8032',
|
||||
'ext': 'mp4',
|
||||
'title': 'Thomas Adès\'s The Exterminating Angel',
|
||||
'description': 'md5:708ae6350dadc604225b4a6e32482bab',
|
||||
'thumbnail': r're:https://.+/.+\.jpg',
|
||||
'upload_date': '20240304',
|
||||
'timestamp': 1709561766,
|
||||
'display_id': 'thomas-ades-the-exterminating-angel-calixto-bieito-opera-bastille-paris',
|
||||
},
|
||||
'expected_warnings': [r'preview'],
|
||||
}, {
|
||||
'url': 'https://edu.medici.tv/en/operas/wagner-lohengrin-paris-opera-kirill-serebrennikov-piotr-beczala-kwangchul-youn-johanni-van-oostrum',
|
||||
'md5': '4ef3f4079a6e1c617584463a9eb84f99',
|
||||
'info_dict': {
|
||||
'id': '7900',
|
||||
'ext': 'mp4',
|
||||
'title': 'Wagner\'s Lohengrin',
|
||||
'description': 'md5:a384a62937866101f86902f21752cd89',
|
||||
'thumbnail': r're:https://.+/.+\.jpg',
|
||||
'upload_date': '20231017',
|
||||
'timestamp': 1697554771,
|
||||
'display_id': 'wagner-lohengrin-paris-opera-kirill-serebrennikov-piotr-beczala-kwangchul-youn-johanni-van-oostrum',
|
||||
},
|
||||
'expected_warnings': [r'preview'],
|
||||
}, {
|
||||
'url': 'https://www.medici.tv/en/concerts/sergey-smbatyan-conducts-mansurian-chouchane-siranossian-mario-brunello',
|
||||
'md5': '9dd757e53b22b2511e85ea9ea60e4815',
|
||||
'info_dict': {
|
||||
'id': '5712',
|
||||
'ext': 'mp4',
|
||||
'title': 'Sergey Smbatyan conducts Tigran Mansurian — With Chouchane Siranossian and Mario Brunello',
|
||||
'thumbnail': r're:https://.+/.+\.jpg',
|
||||
'description': 'md5:9411fe44c874bb10e9af288c65816e41',
|
||||
'upload_date': '20200323',
|
||||
'timestamp': 1584975600,
|
||||
'display_id': 'sergey-smbatyan-conducts-mansurian-chouchane-siranossian-mario-brunello',
|
||||
},
|
||||
'expected_warnings': [r'preview'],
|
||||
}, {
|
||||
'url': 'https://www.medici.tv/en/ballets/carmen-ballet-choregraphie-de-jiri-bubenicek-teatro-dellopera-di-roma',
|
||||
'md5': '40f5e76cb701a97a6d7ba23b62c49990',
|
||||
'info_dict': {
|
||||
'id': '7857',
|
||||
'ext': 'mp4',
|
||||
'title': 'Carmen by Jiří Bubeníček after Roland Petit, music by Bizet, de Falla, Castelnuovo-Tedesco, and Bonolis',
|
||||
'thumbnail': r're:https://.+/.+\.jpg',
|
||||
'description': 'md5:0f15a15611ed748020c769873e10a8bb',
|
||||
'upload_date': '20240223',
|
||||
'timestamp': 1708707600,
|
||||
'display_id': 'carmen-ballet-choregraphie-de-jiri-bubenicek-teatro-dellopera-di-roma',
|
||||
},
|
||||
'expected_warnings': [r'preview'],
|
||||
}, {
|
||||
'url': 'https://www.medici.tv/en/documentaries/la-sonnambula-liege-2023-documentaire',
|
||||
'md5': '87ff198018ce79a34757ab0dd6f21080',
|
||||
'info_dict': {
|
||||
'id': '7513',
|
||||
'ext': 'mp4',
|
||||
'title': 'La Sonnambula',
|
||||
'thumbnail': r're:https://.+/.+\.jpg',
|
||||
'description': 'md5:0caf9109a860fd50cd018df062a67f34',
|
||||
'upload_date': '20231103',
|
||||
'timestamp': 1699010830,
|
||||
'display_id': 'la-sonnambula-liege-2023-documentaire',
|
||||
},
|
||||
'expected_warnings': [r'preview'],
|
||||
}, {
|
||||
'url': 'https://edu.medici.tv/en/masterclasses/yvonne-loriod-olivier-messiaen',
|
||||
'md5': 'fb5dcec46d76ad20fbdbaabb01da191d',
|
||||
'info_dict': {
|
||||
'id': '3024',
|
||||
'ext': 'mp4',
|
||||
'title': 'Olivier Messiaen and Yvonne Loriod, pianists and teachers',
|
||||
'thumbnail': r're:https://.+/.+\.jpg',
|
||||
'description': 'md5:aab948e2f7690214b5c28896c83f1fc1',
|
||||
'upload_date': '20150223',
|
||||
'timestamp': 1424706608,
|
||||
'display_id': 'yvonne-loriod-olivier-messiaen',
|
||||
},
|
||||
'skip': 'Requires authentication; preview starts in the middle',
|
||||
}, {
|
||||
'url': 'https://www.medici.tv/en/jazz/makaya-mccraven-la-rochelle',
|
||||
'md5': '4cc279a8b06609782747c8f50beea2b3',
|
||||
'info_dict': {
|
||||
'id': '3059',
|
||||
'ext': 'flv',
|
||||
'title': 'Daniel Harding conducts the Verbier Festival Music Camp \u2013 With Frans Helmerson',
|
||||
'description': 'md5:322a1e952bafb725174fd8c1a8212f58',
|
||||
'thumbnail': r're:^https?://.*\.jpg$',
|
||||
'upload_date': '20170408',
|
||||
'id': '7922',
|
||||
'ext': 'mp4',
|
||||
'title': 'NEW: Makaya McCraven in La Rochelle',
|
||||
'thumbnail': r're:https://.+/.+\.jpg',
|
||||
'description': 'md5:b5a8aaeb6993d8ccb18bde8abb8aa8d2',
|
||||
'upload_date': '20231228',
|
||||
'timestamp': 1703754863,
|
||||
'display_id': 'makaya-mccraven-la-rochelle',
|
||||
},
|
||||
}
|
||||
'expected_warnings': [r'preview'],
|
||||
}]
|
||||
|
||||
def _real_extract(self, url):
|
||||
video_id = self._match_id(url)
|
||||
|
||||
# Sets csrftoken cookie
|
||||
self._download_webpage(url, video_id)
|
||||
display_id, subdomain = self._match_valid_url(url).group('id', 'sub')
|
||||
self._request_webpage(url, display_id, 'Requesting CSRF token cookie')
|
||||
|
||||
MEDICI_URL = 'http://www.medici.tv/'
|
||||
subdomain = 'edu-' if subdomain == 'edu' else ''
|
||||
origin = f'https://{urllib.parse.urlparse(url).hostname}'
|
||||
|
||||
data = self._download_json(
|
||||
MEDICI_URL, video_id,
|
||||
data=urlencode_postdata({
|
||||
'json': 'true',
|
||||
'page': '/%s' % video_id,
|
||||
'timezone_offset': -420,
|
||||
}), headers={
|
||||
'X-CSRFToken': self._get_cookies(url)['csrftoken'].value,
|
||||
'X-Requested-With': 'XMLHttpRequest',
|
||||
'Referer': MEDICI_URL,
|
||||
'Content-Type': 'application/x-www-form-urlencoded',
|
||||
})
|
||||
|
||||
video = data['video']['videos']['video1']
|
||||
|
||||
title = video.get('nom') or data['title']
|
||||
f'https://api.medici.tv/{subdomain}satie/edito/movie-file/{display_id}/', display_id,
|
||||
headers=filter_dict({
|
||||
'Authorization': try_call(
|
||||
lambda: urllib.parse.unquote(self._get_cookies(url)['auth._token.mAuth'].value)),
|
||||
'Device-Type': 'web',
|
||||
'Origin': origin,
|
||||
'Referer': f'{origin}/',
|
||||
'Accept': 'application/json, text/plain, */*',
|
||||
}))
|
||||
|
||||
video_id = video.get('id') or video_id
|
||||
formats = self._extract_f4m_formats(
|
||||
update_url_query(video['url_akamai'], {
|
||||
'hdcore': '3.1.0',
|
||||
'plugin=aasp': '3.1.0.43.124',
|
||||
}), video_id, f4m_id='hds')
|
||||
if not traverse_obj(data, ('video', 'is_full_video')) and traverse_obj(
|
||||
data, ('video', 'is_limited_by_user_access')):
|
||||
self.report_warning(
|
||||
'The full video is for subscribers only. Only previews will be downloaded. If you '
|
||||
'have used the --cookies-from-browser option, try using the --cookies option instead')
|
||||
|
||||
description = data.get('meta_description')
|
||||
thumbnail = video.get('url_thumbnail') or data.get('main_image')
|
||||
upload_date = unified_strdate(data['video'].get('date'))
|
||||
formats, subtitles = self._extract_m3u8_formats_and_subtitles(
|
||||
data['video']['video_url'], display_id, 'mp4')
|
||||
|
||||
return {
|
||||
'id': video_id,
|
||||
'title': title,
|
||||
'description': description,
|
||||
'thumbnail': thumbnail,
|
||||
'upload_date': upload_date,
|
||||
'id': str(data['id']),
|
||||
'display_id': display_id,
|
||||
'formats': formats,
|
||||
'subtitles': subtitles,
|
||||
**traverse_obj(data, {
|
||||
'title': ('title', {str}),
|
||||
'description': ('subtitle', {str}),
|
||||
'thumbnail': ('picture', {url_or_none}),
|
||||
'timestamp': ('date_publish', {parse_iso8601}),
|
||||
}),
|
||||
}
|
||||
|
Loading…
Reference in New Issue