devscripts/make_lazy_extractors.py

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