]> jfr.im git - yt-dlp.git/blob - devscripts/make_lazy_extractors.py
8189c77dce5a588ac5796c8453e6080a9ab6b920
[yt-dlp.git] / devscripts / make_lazy_extractors.py
1 #!/usr/bin/env python3
2 from __future__ import unicode_literals, print_function
3
4 from inspect import getsource
5 import io
6 import os
7 from os.path import dirname as dirn
8 import sys
9
10 print('WARNING: Lazy loading extractors is an experimental feature that may not always work', file=sys.stderr)
11
12 sys.path.insert(0, dirn(dirn((os.path.abspath(__file__)))))
13
14 lazy_extractors_filename = sys.argv[1]
15 if os.path.exists(lazy_extractors_filename):
16 os.remove(lazy_extractors_filename)
17
18 # Block plugins from loading
19 plugins_dirname = 'ytdlp_plugins'
20 plugins_blocked_dirname = 'ytdlp_plugins_blocked'
21 if os.path.exists(plugins_dirname):
22 os.rename(plugins_dirname, plugins_blocked_dirname)
23
24 from yt_dlp.extractor import _ALL_CLASSES
25 from yt_dlp.extractor.common import InfoExtractor, SearchInfoExtractor
26
27 if os.path.exists(plugins_blocked_dirname):
28 os.rename(plugins_blocked_dirname, plugins_dirname)
29
30 with open('devscripts/lazy_load_template.py', 'rt') as f:
31 module_template = f.read()
32
33 module_contents = [
34 module_template,
35 getsource(InfoExtractor.ie_key),
36 getsource(InfoExtractor._match_valid_url),
37 getsource(InfoExtractor.suitable),
38 '\nclass LazyLoadSearchExtractor(LazyLoadExtractor):\n pass\n']
39
40 ie_template = '''
41 class {name}({bases}):
42 _VALID_URL = {valid_url!r}
43 _module = '{module}'
44 '''
45
46 make_valid_template = '''
47 @classmethod
48 def _make_valid_url(cls):
49 return {valid_url!r}
50 '''
51
52
53 def get_base_name(base):
54 if base is InfoExtractor:
55 return 'LazyLoadExtractor'
56 elif base is SearchInfoExtractor:
57 return 'LazyLoadSearchExtractor'
58 else:
59 return base.__name__
60
61
62 def build_lazy_ie(ie, name):
63 valid_url = getattr(ie, '_VALID_URL', None)
64 s = ie_template.format(
65 name=name,
66 bases=', '.join(map(get_base_name, ie.__bases__)),
67 valid_url=valid_url,
68 module=ie.__module__)
69 if ie.suitable.__func__ is not InfoExtractor.suitable.__func__:
70 s += '\n' + getsource(ie.suitable)
71 if hasattr(ie, '_make_valid_url'):
72 # search extractors
73 s += make_valid_template.format(valid_url=ie._make_valid_url())
74 return s
75
76
77 # find the correct sorting and add the required base classes so that subclasses
78 # can be correctly created
79 classes = _ALL_CLASSES[:-1]
80 ordered_cls = []
81 while classes:
82 for c in classes[:]:
83 bases = set(c.__bases__) - set((object, InfoExtractor, SearchInfoExtractor))
84 stop = False
85 for b in bases:
86 if b not in classes and b not in ordered_cls:
87 if b.__name__ == 'GenericIE':
88 exit()
89 classes.insert(0, b)
90 stop = True
91 if stop:
92 break
93 if all(b in ordered_cls for b in bases):
94 ordered_cls.append(c)
95 classes.remove(c)
96 break
97 ordered_cls.append(_ALL_CLASSES[-1])
98
99 names = []
100 for ie in ordered_cls:
101 name = ie.__name__
102 src = build_lazy_ie(ie, name)
103 module_contents.append(src)
104 if ie in _ALL_CLASSES:
105 names.append(name)
106
107 module_contents.append(
108 '\n_ALL_CLASSES = [{0}]'.format(', '.join(names)))
109
110 module_src = '\n'.join(module_contents) + '\n'
111
112 with io.open(lazy_extractors_filename, 'wt', encoding='utf-8') as f:
113 f.write(module_src)