1
0
Fork 0
mirror of https://gitlab.com/dstftw/youtube-dl.git synced 2020-11-16 09:42:26 +00:00

[pornhub] Improve extraction and style (closes #12515)

This commit is contained in:
Sergey M․ 2017-03-22 01:59:27 +07:00
parent 21fbf0f955
commit e1e35d1ac6
No known key found for this signature in database
GPG key ID: 2C393E0F18A9236D

View file

@ -20,6 +20,7 @@ from ..utils import (
js_to_json, js_to_json,
orderedSet, orderedSet,
# sanitized_Request, # sanitized_Request,
remove_quotes,
str_to_int, str_to_int,
) )
# from ..aes import ( # from ..aes import (
@ -131,38 +132,32 @@ class PornHubIE(InfoExtractor):
tv_webpage = dl_webpage('tv') tv_webpage = dl_webpage('tv')
encoded_url = self._search_regex(r'(var.*mediastring.*)</script>', assignments = self._search_regex(
tv_webpage, 'encoded url') r'(var.+?mediastring.+?)</script>', tv_webpage,
assignments = encoded_url.split(";") 'encoded url').split(';')
js_vars = {} js_vars = {}
def parse_js_value(inp): def parse_js_value(inp):
inp = re.sub(r'/\*[^*]*\*/', "", inp) inp = re.sub(r'/\*(?:(?!\*/).)*?\*/', '', inp)
if '+' in inp:
if "+" in inp: inps = inp.split('+')
inps = inp.split("+") return functools.reduce(
return functools.reduce(operator.concat, map(parse_js_value, inps)) operator.concat, map(parse_js_value, inps))
inp = inp.strip() inp = inp.strip()
if inp in js_vars: if inp in js_vars:
return js_vars[inp] return js_vars[inp]
return remove_quotes(inp)
# Hope it's a string!
assert inp.startswith('"') and inp.endswith('"')
return inp[1:-1]
for assn in assignments: for assn in assignments:
assn = assn.strip() assn = assn.strip()
if len(assn) == 0: if not assn:
continue continue
assn = re.sub(r'var\s+', '', assn)
assert assn.startswith("var ") vname, value = assn.split('=', 1)
assn = assn[4:]
vname, value = assn.split("=", 1)
js_vars[vname] = parse_js_value(value) js_vars[vname] = parse_js_value(value)
video_url = js_vars["mediastring"] video_url = js_vars['mediastring']
title = self._search_regex( title = self._search_regex(
r'<h1>([^>]+)</h1>', tv_webpage, 'title', default=None) r'<h1>([^>]+)</h1>', tv_webpage, 'title', default=None)