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 sys.path.insert(0, dirn(dirn((os.path.abspath(__file__)))))
  11
  12 lazy_extractors_filename = sys.argv[1] if len(sys.argv) > 1 else 'yt_dlp/extractor/lazy_extractors.py'
  13 if os.path.exists(lazy_extractors_filename):
  14     os.remove(lazy_extractors_filename)
  15
  16 # Block plugins from loading
  17 plugins_dirname = 'ytdlp_plugins'
  18 plugins_blocked_dirname = 'ytdlp_plugins_blocked'
  19 if os.path.exists(plugins_dirname):
  20     os.rename(plugins_dirname, plugins_blocked_dirname)
  21
  22 from yt_dlp.extractor import _ALL_CLASSES
  23 from yt_dlp.extractor.common import InfoExtractor, SearchInfoExtractor
  24
  25 if os.path.exists(plugins_blocked_dirname):
  26     os.rename(plugins_blocked_dirname, plugins_dirname)
  27
  28 with open('devscripts/lazy_load_template.py', 'rt') as f:
  29     module_template = f.read()
  30
  31 CLASS_PROPERTIES = ['ie_key', 'working', '_match_valid_url', 'suitable', '_match_id', 'get_temp_id']
  32 module_contents = [
  33     module_template,
  34     *[getsource(getattr(InfoExtractor, k)) for k in CLASS_PROPERTIES],
  35     '\nclass LazyLoadSearchExtractor(LazyLoadExtractor):\n    pass\n']
  36
  37 ie_template = '''
  38 class {name}({bases}):
  39     _module = '{module}'
  40 '''
  41
  42
  43 def get_base_name(base):
  44     if base is InfoExtractor:
  45         return 'LazyLoadExtractor'
  46     elif base is SearchInfoExtractor:
  47         return 'LazyLoadSearchExtractor'
  48     else:
  49         return base.__name__
  50
  51
  52 def build_lazy_ie(ie, name):
  53     s = ie_template.format(
  54         name=name,
  55         bases=', '.join(map(get_base_name, ie.__bases__)),
  56         module=ie.__module__)
  57     valid_url = getattr(ie, '_VALID_URL', None)
  58     if not valid_url and hasattr(ie, '_make_valid_url'):
  59         valid_url = ie._make_valid_url()
  60     if valid_url:
  61         s += f'    _VALID_URL = {valid_url!r}\n'
  62     if not ie._WORKING:
  63         s += '    _WORKING = False\n'
  64     if ie.suitable.__func__ is not InfoExtractor.suitable.__func__:
  65         s += f'\n{getsource(ie.suitable)}'
  66     return s
  67
  68
  69 # find the correct sorting and add the required base classes so that subclasses
  70 # can be correctly created
  71 classes = _ALL_CLASSES[:-1]
  72 ordered_cls = []
  73 while classes:
  74     for c in classes[:]:
  75         bases = set(c.__bases__) - set((object, InfoExtractor, SearchInfoExtractor))
  76         stop = False
  77         for b in bases:
  78             if b not in classes and b not in ordered_cls:
  79                 if b.__name__ == 'GenericIE':
  80                     exit()
  81                 classes.insert(0, b)
  82                 stop = True
  83         if stop:
  84             break
  85         if all(b in ordered_cls for b in bases):
  86             ordered_cls.append(c)
  87             classes.remove(c)
  88             break
  89 ordered_cls.append(_ALL_CLASSES[-1])
  90
  91 names = []
  92 for ie in ordered_cls:
  93     name = ie.__name__
  94     src = build_lazy_ie(ie, name)
  95     module_contents.append(src)
  96     if ie in _ALL_CLASSES:
  97         names.append(name)
  98
  99 module_contents.append(
 100     '\n_ALL_CLASSES = [{0}]'.format(', '.join(names)))
 101
 102 module_src = '\n'.join(module_contents) + '\n'
 103
 104 with io.open(lazy_extractors_filename, 'wt', encoding='utf-8') as f:
 105     f.write(module_src)