code.delx.au - webdl/blob - sbs.py

   1 import requests_cache
   2 from common import grab_html, grab_json, grab_xml, download_hls, download_mpd, Node, append_to_qs
   3
   4 import json
   5 import logging
   6 import os
   7 import sys
   8
   9 BASE = "https://www.sbs.com.au"
  10 FULL_VIDEO_LIST = BASE + "/api/video_feed/f/Bgtm9B/sbs-section-programs/"
  11 VIDEO_URL = BASE + "/ondemand/video/single/%s"
  12
  13 NS = {
  14     "smil": "http://www.w3.org/2005/SMIL21/Language",
  15 }
  16
  17
  18 class SbsVideoNode(Node):
  19     def __init__(self, title, parent, url):
  20         Node.__init__(self, title, parent)
  21         self.video_id = url.split("/")[-1]
  22         self.can_download = True
  23
  24     def download(self):
  25         with requests_cache.disabled():
  26             doc = grab_html(VIDEO_URL % self.video_id)
  27         player_params = self.get_player_params(doc)
  28
  29         error = player_params.get("error", None)
  30         if error:
  31             print("Cannot download:", error)
  32             return False
  33
  34         release_url = player_params["releaseUrls"]["html"]
  35         filename = self.title + ".ts"
  36
  37         hls_url = self.get_hls_url(release_url)
  38         if hls_url:
  39             return download_hls(filename, hls_url)
  40         else:
  41             return download_mpd(filename, release_url)
  42
  43     def get_player_params(self, doc):
  44         for script in doc.xpath("//script"):
  45             if not script.text:
  46                 continue
  47             for line in script.text.split("\n"):
  48                 s = "var playerParams = {"
  49                 if s in line:
  50                     p1 = line.find(s) + len(s) - 1
  51                     p2 = line.find("};", p1) + 1
  52                     if p1 >= 0 and p2 > 0:
  53                         return json.loads(line[p1:p2])
  54         raise Exception("Unable to find player params for %s: %s" % (self.video_id, self.title))
  55
  56     def get_hls_url(self, release_url):
  57         with requests_cache.disabled():
  58             doc = grab_xml("https:" + release_url.replace("http:", "").replace("https:", ""))
  59             video = doc.xpath("//smil:video", namespaces=NS)
  60             if not video:
  61                 return
  62             video_url = video[0].attrib["src"]
  63             return video_url
  64
  65 class SbsNavNode(Node):
  66     def create_video_node(self, entry_data):
  67         SbsVideoNode(entry_data["title"], self, entry_data["id"])
  68
  69     def find_existing_child(self, path):
  70         for child in self.children:
  71             if child.title == path:
  72                 return child
  73
  74 class SbsRootNode(SbsNavNode):
  75     def __init__(self, parent):
  76         Node.__init__(self, "SBS", parent)
  77
  78     def fill_children(self):
  79         all_video_entries = self.load_all_video_entries()
  80         category_and_entry_data = self.explode_videos_to_unique_categories(all_video_entries)
  81         for category_path, entry_data in category_and_entry_data:
  82             nav_node = self.create_nav_node(self, category_path)
  83             nav_node.create_video_node(entry_data)
  84
  85     def load_all_video_entries(self):
  86         channels = [
  87             "Channel/SBS1",
  88             "Channel/SBS Food",
  89             "Channel/SBS VICELAND",
  90             "Channel/SBS World Movies",
  91             "Channel/Web Exclusive",
  92         ]
  93
  94         all_entries = {}
  95         for channel in channels:
  96             self.load_all_video_entries_for_channel(all_entries, channel)
  97
  98         all_entries = list(all_entries.values())
  99         print(" SBS fetched", len(all_entries))
 100         return all_entries
 101
 102     def load_all_video_entries_for_channel(self, all_entries, channel):
 103         offset = 1
 104         page_size = 500
 105         duplicate_warning = False
 106
 107         while True:
 108             entries = self.fetch_entries_page(channel, offset, page_size)
 109             if len(entries) == 0:
 110                 break
 111
 112             for entry in entries:
 113                 guid = entry["guid"]
 114                 if guid in entries and not duplicate_warning:
 115                     # https://bitbucket.org/delx/webdl/issues/102/recent-sbs-series-missing
 116                     logging.warn("SBS returned a duplicate response, data is probably missing. Try decreasing page_size.")
 117                     duplicate_warning = True
 118
 119                 all_entries[guid] = entry
 120
 121             offset += page_size
 122             if os.isatty(sys.stdout.fileno()):
 123                 sys.stdout.write(".")
 124                 sys.stdout.flush()
 125
 126     def fetch_entries_page(self, channel, offset, page_size):
 127         url = append_to_qs(FULL_VIDEO_LIST, {
 128             "range": "%s-%s" % (offset, offset+page_size-1),
 129             "byCategories": channel,
 130         })
 131         data = grab_json(url)
 132         if "entries" not in data:
 133             raise Exception("Missing data in SBS response", data)
 134         return data["entries"]
 135
 136     def explode_videos_to_unique_categories(self, all_video_entries):
 137         for entry_data in all_video_entries:
 138             for category_data in entry_data["media$categories"]:
 139                 category_path = self.calculate_category_path(
 140                     category_data["media$scheme"],
 141                     category_data["media$name"],
 142                 )
 143                 if category_path:
 144                     yield category_path, entry_data
 145
 146     def calculate_category_path(self, scheme, name):
 147         if not scheme:
 148             return
 149         if scheme == name:
 150             return
 151         name = name.split("/")
 152         if name[0] != scheme:
 153             name.insert(0, scheme)
 154         return name
 155
 156     def create_nav_node(self, parent, category_path):
 157         if not category_path:
 158             return parent
 159
 160         current_path = category_path[0]
 161         current_node = parent.find_existing_child(current_path)
 162         if not current_node:
 163             current_node = SbsNavNode(current_path, parent)
 164         return self.create_nav_node(current_node, category_path[1:])
 165
 166 def fill_nodes(root_node):
 167     SbsRootNode(root_node)