#!/usr/bin/env python3
+
# Allow direct execution
import os
+import re
import sys
import unittest
sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
-# Various small unit tests
+import contextlib
import io
import itertools
import json
import xml.etree.ElementTree
from yt_dlp.compat import (
- compat_chr,
compat_etree_fromstring,
- compat_getenv,
compat_HTMLParseError,
compat_os_name,
- compat_setenv,
)
from yt_dlp.utils import (
Config,
datetime_from_str,
detect_exe_version,
determine_ext,
+ determine_file_encoding,
dfxp2srt,
dict_get,
encode_base_n,
fix_xml_ampersands,
float_or_none,
format_bytes,
+ get_compatible_ext,
get_element_by_attribute,
get_element_by_class,
get_element_html_by_attribute,
sanitized_Request,
shell_quote,
smuggle_url,
+ str_or_none,
str_to_int,
strip_jsonp,
strip_or_none,
subtitles_filename,
timeconvert,
+ traverse_obj,
unescapeHTML,
unified_strdate,
unified_timestamp,
self.assertEqual(sanitize_filename('123'), '123')
- self.assertEqual('abc_de', sanitize_filename('abc/de'))
+ self.assertEqual('abc⧸de', sanitize_filename('abc/de'))
self.assertFalse('/' in sanitize_filename('abc/de///'))
- self.assertEqual('abc_de', sanitize_filename('abc/<>\\*|de'))
- self.assertEqual('xxx', sanitize_filename('xxx/<>\\*|'))
- self.assertEqual('yes no', sanitize_filename('yes? no'))
- self.assertEqual('this - that', sanitize_filename('this: that'))
+ self.assertEqual('abc_de', sanitize_filename('abc/<>\\*|de', is_id=False))
+ self.assertEqual('xxx', sanitize_filename('xxx/<>\\*|', is_id=False))
+ self.assertEqual('yes no', sanitize_filename('yes? no', is_id=False))
+ self.assertEqual('this - that', sanitize_filename('this: that', is_id=False))
self.assertEqual(sanitize_filename('AT&T'), 'AT&T')
aumlaut = 'ä'
def env(var):
return f'%{var}%' if sys.platform == 'win32' else f'${var}'
- compat_setenv('yt_dlp_EXPATH_PATH', 'expanded')
+ os.environ['yt_dlp_EXPATH_PATH'] = 'expanded'
self.assertEqual(expand_path(env('yt_dlp_EXPATH_PATH')), 'expanded')
- self.assertEqual(expand_path(env('HOME')), compat_getenv('HOME'))
- self.assertEqual(expand_path('~'), compat_getenv('HOME'))
- self.assertEqual(
- expand_path('~/%s' % env('yt_dlp_EXPATH_PATH')),
- '%s/expanded' % compat_getenv('HOME'))
+
+ old_home = os.environ.get('HOME')
+ test_str = R'C:\Documents and Settings\тест\Application Data'
+ try:
+ os.environ['HOME'] = test_str
+ self.assertEqual(expand_path(env('HOME')), os.getenv('HOME'))
+ self.assertEqual(expand_path('~'), os.getenv('HOME'))
+ self.assertEqual(
+ expand_path('~/%s' % env('yt_dlp_EXPATH_PATH')),
+ '%s/expanded' % os.getenv('HOME'))
+ finally:
+ os.environ['HOME'] = old_home or ''
def test_prepend_extension(self):
self.assertEqual(prepend_extension('abc.ext', 'temp'), 'abc.temp.ext')
self.assertEqual(unified_strdate('2012/10/11 01:56:38 +0000'), '20121011')
self.assertEqual(unified_strdate('1968 12 10'), '19681210')
self.assertEqual(unified_strdate('1968-12-10'), '19681210')
+ self.assertEqual(unified_strdate('31-07-2022 20:00'), '20220731')
self.assertEqual(unified_strdate('28/01/2014 21:00:00 +0100'), '20140128')
self.assertEqual(
unified_strdate('11/26/2014 11:30:00 AM PST', day_first=False),
self.assertEqual(unified_timestamp('December 15, 2017 at 7:49 am'), 1513324140)
self.assertEqual(unified_timestamp('2018-03-14T08:32:43.1493874+00:00'), 1521016363)
+ self.assertEqual(unified_timestamp('December 31 1969 20:00:01 EDT'), 1)
+ self.assertEqual(unified_timestamp('Wednesday 31 December 1969 18:01:26 MDT'), 86)
+ self.assertEqual(unified_timestamp('12/31/1969 20:01:18 EDT', False), 78)
+
def test_determine_ext(self):
self.assertEqual(determine_ext('http://example.com/foo/bar.mp4/?download'), 'mp4')
self.assertEqual(determine_ext('http://example.com/foo/bar/?download', None), None)
self.assertEqual(base_url('http://foo.de/bar/'), 'http://foo.de/bar/')
self.assertEqual(base_url('http://foo.de/bar/baz'), 'http://foo.de/bar/')
self.assertEqual(base_url('http://foo.de/bar/baz?x=z/x/c'), 'http://foo.de/bar/')
+ self.assertEqual(base_url('http://foo.de/bar/baz&x=z&w=y/x/c'), 'http://foo.de/bar/baz&x=z&w=y/x/')
def test_urljoin(self):
self.assertEqual(urljoin('http://foo.de/', '/a/b/c.txt'), 'http://foo.de/a/b/c.txt')
'dynamic_range': 'HDR10',
})
self.assertEqual(parse_codecs('av01.0.12M.10.0.110.09.16.09.0'), {
- 'vcodec': 'av01.0.12M.10',
+ 'vcodec': 'av01.0.12M.10.0.110.09.16.09.0',
'acodec': 'none',
'dynamic_range': 'HDR10',
})
)
self.assertEqual(escape_url('http://vimeo.com/56015672#at=0'), 'http://vimeo.com/56015672#at=0')
+ def test_js_to_json_vars_strings(self):
+ self.assertDictEqual(
+ json.loads(js_to_json(
+ '''{
+ 'null': a,
+ 'nullStr': b,
+ 'true': c,
+ 'trueStr': d,
+ 'false': e,
+ 'falseStr': f,
+ 'unresolvedVar': g,
+ }''',
+ {
+ 'a': 'null',
+ 'b': '"null"',
+ 'c': 'true',
+ 'd': '"true"',
+ 'e': 'false',
+ 'f': '"false"',
+ 'g': 'var',
+ }
+ )),
+ {
+ 'null': None,
+ 'nullStr': 'null',
+ 'true': True,
+ 'trueStr': 'true',
+ 'false': False,
+ 'falseStr': 'false',
+ 'unresolvedVar': 'var'
+ }
+ )
+
+ self.assertDictEqual(
+ json.loads(js_to_json(
+ '''{
+ 'int': a,
+ 'intStr': b,
+ 'float': c,
+ 'floatStr': d,
+ }''',
+ {
+ 'a': '123',
+ 'b': '"123"',
+ 'c': '1.23',
+ 'd': '"1.23"',
+ }
+ )),
+ {
+ 'int': 123,
+ 'intStr': '123',
+ 'float': 1.23,
+ 'floatStr': '1.23',
+ }
+ )
+
+ self.assertDictEqual(
+ json.loads(js_to_json(
+ '''{
+ 'object': a,
+ 'objectStr': b,
+ 'array': c,
+ 'arrayStr': d,
+ }''',
+ {
+ 'a': '{}',
+ 'b': '"{}"',
+ 'c': '[]',
+ 'd': '"[]"',
+ }
+ )),
+ {
+ 'object': {},
+ 'objectStr': '{}',
+ 'array': [],
+ 'arrayStr': '[]',
+ }
+ )
+
def test_js_to_json_realworld(self):
inp = '''{
'clip':{'provider':'pseudo'}
on = js_to_json('[1,//{},\n2]')
self.assertEqual(json.loads(on), [1, 2])
+ on = js_to_json(R'"\^\$\#"')
+ self.assertEqual(json.loads(on), R'^$#', msg='Unnecessary escapes should be stripped')
+
+ on = js_to_json('\'"\\""\'')
+ self.assertEqual(json.loads(on), '"""', msg='Unnecessary quote escape should be escaped')
+
def test_js_to_json_malformed(self):
self.assertEqual(js_to_json('42a1'), '42"a1"')
self.assertEqual(js_to_json('42a-1'), '42"a"-1')
+ def test_js_to_json_template_literal(self):
+ self.assertEqual(js_to_json('`Hello ${name}`', {'name': '"world"'}), '"Hello world"')
+ self.assertEqual(js_to_json('`${name}${name}`', {'name': '"X"'}), '"XX"')
+ self.assertEqual(js_to_json('`${name}${name}`', {'name': '5'}), '"55"')
+ self.assertEqual(js_to_json('`${name}"${name}"`', {'name': '5'}), '"5\\"5\\""')
+ self.assertEqual(js_to_json('`${name}`', {}), '"name"')
+
def test_extract_attributes(self):
self.assertEqual(extract_attributes('<e x="y">'), {'x': 'y'})
self.assertEqual(extract_attributes("<e x='y'>"), {'x': 'y'})
self.assertEqual(extract_attributes('<e x="décomposé">'), {'x': 'décompose\u0301'})
# "Narrow" Python builds don't support unicode code points outside BMP.
try:
- compat_chr(0x10000)
+ chr(0x10000)
supports_outside_bmp = True
except ValueError:
supports_outside_bmp = False
self.assertEqual(list(get_elements_text_and_html_by_attribute('class', 'foo', html)), [])
self.assertEqual(list(get_elements_text_and_html_by_attribute('class', 'no-such-foo', html)), [])
+ self.assertEqual(list(get_elements_text_and_html_by_attribute(
+ 'class', 'foo', '<a class="foo">nice</a><span class="foo">nice</span>', tag='a')), [('nice', '<a class="foo">nice</a>')])
+
GET_ELEMENT_BY_TAG_TEST_STRING = '''
random text lorem ipsum</p>
<div>
def test(ll, idx, val, cache):
self.assertEqual(ll[idx], val)
- self.assertEqual(getattr(ll, '_LazyList__cache'), list(cache))
+ self.assertEqual(ll._cache, list(cache))
ll = LazyList(range(10))
test(ll, 0, 0, range(1))
else:
self.assertFalse(testing_write, f'{test_mode} is not blocked by {lock_mode}')
finally:
- try:
+ with contextlib.suppress(OSError):
os.remove(FILE)
- except Exception:
- pass
+
+ def test_determine_file_encoding(self):
+ self.assertEqual(determine_file_encoding(b''), (None, 0))
+ self.assertEqual(determine_file_encoding(b'--verbose -x --audio-format mkv\n'), (None, 0))
+
+ self.assertEqual(determine_file_encoding(b'\xef\xbb\xbf'), ('utf-8', 3))
+ self.assertEqual(determine_file_encoding(b'\x00\x00\xfe\xff'), ('utf-32-be', 4))
+ self.assertEqual(determine_file_encoding(b'\xff\xfe'), ('utf-16-le', 2))
+
+ self.assertEqual(determine_file_encoding(b'\xff\xfe# coding: utf-8\n--verbose'), ('utf-16-le', 2))
+
+ self.assertEqual(determine_file_encoding(b'# coding: utf-8\n--verbose'), ('utf-8', 0))
+ self.assertEqual(determine_file_encoding(b'# coding: someencodinghere-12345\n--verbose'), ('someencodinghere-12345', 0))
+
+ self.assertEqual(determine_file_encoding(b'#coding:utf-8\n--verbose'), ('utf-8', 0))
+ self.assertEqual(determine_file_encoding(b'# coding: utf-8 \r\n--verbose'), ('utf-8', 0))
+
+ self.assertEqual(determine_file_encoding('# coding: utf-32-be'.encode('utf-32-be')), ('utf-32-be', 0))
+ self.assertEqual(determine_file_encoding('# coding: utf-16-le'.encode('utf-16-le')), ('utf-16-le', 0))
+
+ def test_get_compatible_ext(self):
+ self.assertEqual(get_compatible_ext(
+ vcodecs=[None], acodecs=[None, None], vexts=['mp4'], aexts=['m4a', 'm4a']), 'mkv')
+ self.assertEqual(get_compatible_ext(
+ vcodecs=[None], acodecs=[None], vexts=['flv'], aexts=['flv']), 'flv')
+
+ self.assertEqual(get_compatible_ext(
+ vcodecs=[None], acodecs=[None], vexts=['mp4'], aexts=['m4a']), 'mp4')
+ self.assertEqual(get_compatible_ext(
+ vcodecs=[None], acodecs=[None], vexts=['mp4'], aexts=['webm']), 'mkv')
+ self.assertEqual(get_compatible_ext(
+ vcodecs=[None], acodecs=[None], vexts=['webm'], aexts=['m4a']), 'mkv')
+ self.assertEqual(get_compatible_ext(
+ vcodecs=[None], acodecs=[None], vexts=['webm'], aexts=['webm']), 'webm')
+ self.assertEqual(get_compatible_ext(
+ vcodecs=[None], acodecs=[None], vexts=['webm'], aexts=['weba']), 'webm')
+
+ self.assertEqual(get_compatible_ext(
+ vcodecs=['h264'], acodecs=['mp4a'], vexts=['mov'], aexts=['m4a']), 'mp4')
+ self.assertEqual(get_compatible_ext(
+ vcodecs=['av01.0.12M.08'], acodecs=['opus'], vexts=['mp4'], aexts=['webm']), 'webm')
+
+ self.assertEqual(get_compatible_ext(
+ vcodecs=['vp9'], acodecs=['opus'], vexts=['webm'], aexts=['webm'], preferences=['flv', 'mp4']), 'mp4')
+ self.assertEqual(get_compatible_ext(
+ vcodecs=['av1'], acodecs=['mp4a'], vexts=['webm'], aexts=['m4a'], preferences=('webm', 'mkv')), 'mkv')
+
+ def test_traverse_obj(self):
+ _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': {},
+ }
+
+ # 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, ...),
+ (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, ...)), _TEST_DATA['urls'][0].values(),
+ msg='`...` selection for dicts should select all values')
+ self.assertEqual(traverse_obj(_TEST_DATA, (..., ..., 'url')),
+ ['https://www.example.com/0', 'https://www.example.com/1'],
+ msg='nested `...` queries should work')
+ self.assertCountEqual(traverse_obj(_TEST_DATA, (..., ..., 'index')), range(4),
+ msg='`...` query result should be flattened')
+
+ # 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)), {'str'},
+ msg='exceptions in the query function should be catched')
+ if __debug__:
+ with self.assertRaises(Exception, msg='Wrong function signature should raise in debug'):
+ traverse_obj(_TEST_DATA, lambda a: ...)
+ with self.assertRaises(Exception, msg='Wrong function signature should raise in debug'):
+ traverse_obj(_TEST_DATA, lambda a, b, c: ...)
+
+ # Test set as key (transformation/type, like `expected_type`)
+ self.assertEqual(traverse_obj(_TEST_DATA, (..., {str.upper}, )), ['STR'],
+ msg='Function in set should be a transformation')
+ self.assertEqual(traverse_obj(_TEST_DATA, (..., {str})), ['str'],
+ msg='Type in set should be a type filter')
+ self.assertEqual(traverse_obj(_TEST_DATA, {dict}), _TEST_DATA,
+ msg='A single set should be wrapped into a path')
+ self.assertEqual(traverse_obj(_TEST_DATA, (..., {str.upper})), ['STR'],
+ msg='Transformation function should not raise')
+ self.assertEqual(traverse_obj(_TEST_DATA, (..., {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, {str.upper, str})
+
+ # 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, (..., 'fail'), 100), 100,
+ msg='alternatives should track their own branching return')
+ self.assertEqual(traverse_obj(_TEST_DATA, ('dict', ...), ('data', ...)), 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='tripple nesting in path should be treated as branches')
+ self.assertEqual(traverse_obj(_TEST_DATA, ('urls', ('fail', (..., '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='tripple 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=...), {0: ...},
+ 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=...), {0: ...},
+ msg='use `default` when dict key and `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=...), {0: ...},
+ msg='default to dict if pruned and default is given')
+ self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 'fail'}}, default=...), {0: {0: ...}},
+ msg='use nested `default` when nested dict key fails and `default`')
+ self.assertEqual(traverse_obj(_TEST_DATA, {0: ('dict', ...)}), {},
+ 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=...), ...,
+ 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, (..., 'fail'), default=1), 1,
+ msg='if branched but not successful return `default` if defined, not `[]`')
+ self.assertEqual(traverse_obj(_DEFAULT_DATA, (..., 'fail'), default=None), None,
+ msg='if branched but not successful return `default` even if `default` is `None`')
+ self.assertEqual(traverse_obj(_DEFAULT_DATA, (..., 'fail')), [],
+ msg='if branched but not successful return `[]`, not `default`')
+ self.assertEqual(traverse_obj(_DEFAULT_DATA, ('list', ...)), [],
+ msg='if branched but object is empty return `[]`, not `default`')
+ self.assertEqual(traverse_obj(None, ...), [],
+ msg='if branched but object is `None` return `[]`, not `default`')
+ self.assertEqual(traverse_obj({0: None}, (0, ...)), [],
+ msg='if branched but state is `None` return `[]`, not `default`')
+
+ branching_paths = [
+ ('fail', ...),
+ (..., 'fail'),
+ 100 * ('fail',) + (...,),
+ (...,) + 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 fuction in try_call')
+ self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, ..., 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, {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(...)),
+ [{0: ...}, {0: ...}], 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', ...], 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', ...), get_all=False), 0,
+ msg='if not `get_all`, return only first matching value')
+ self.assertEqual(traverse_obj(_GET_ALL_DATA, ..., 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'},
+ },
+ }
+ 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, (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', ...),
+ 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`')
+
+ # 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.fullmatch(r'0(12)(?P<group>3)(4)?', '0123')
+ self.assertEqual(traverse_obj(mobj, ...), [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')
if __name__ == '__main__':