-#!/usr/bin/env python
-
-from common import grab_html, grab_json, grab_xml, download_hls, Node
+import requests_cache
+from common import grab_html, grab_json, grab_xml, download_hls, download_mpd, Node, append_to_qs
import json
+import sys
-BASE = "http://www.sbs.com.au"
-FULL_VIDEO_LIST = BASE + "/api/video_feed/f/Bgtm9B/sbs-section-programs/?form=json"
+BASE = "https://www.sbs.com.au"
+FULL_VIDEO_LIST = BASE + "/api/video_feed/f/Bgtm9B/sbs-section-programs/"
VIDEO_URL = BASE + "/ondemand/video/single/%s"
NS = {
self.can_download = True
def download(self):
- doc = grab_html(VIDEO_URL % self.video_id, 0)
+ with requests_cache.disabled():
+ doc = grab_html(VIDEO_URL % self.video_id)
player_params = self.get_player_params(doc)
- release_url = player_params["releaseUrls"]["html"]
- doc = grab_xml(release_url, 0)
- video = doc.xpath("//smil:video", namespaces=NS)[0]
- video_url = video.attrib["src"]
- if not video_url:
- raise Exception("Unsupported video %s: %s" % (self.video_id, self.title))
+ error = player_params.get("error", None)
+ if error:
+ print("Cannot download:", error)
+ return False
+
+ release_url = player_params["releaseUrls"]["html"]
filename = self.title + ".ts"
- return download_hls(filename, video_url)
+
+ hls_url = self.get_hls_url(release_url)
+ if hls_url:
+ return download_hls(filename, hls_url)
+ else:
+ return download_mpd(filename, release_url)
def get_player_params(self, doc):
for script in doc.xpath("//script"):
return json.loads(line[p1:p2])
raise Exception("Unable to find player params for %s: %s" % (self.video_id, self.title))
+ def get_hls_url(self, release_url):
+ with requests_cache.disabled():
+ doc = grab_xml("https:" + release_url.replace("http:", "").replace("https:", ""))
+ video = doc.xpath("//smil:video", namespaces=NS)
+ if not video:
+ return
+ video_url = video[0].attrib["src"]
+ return video_url
class SbsNavNode(Node):
def create_video_node(self, entry_data):
Node.__init__(self, "SBS", parent)
def fill_children(self):
- full_video_list = grab_json(FULL_VIDEO_LIST, 3600)
- category_and_entry_data = self.explode_videos_to_unique_categories(full_video_list)
+ all_video_entries = self.load_all_video_entries()
+ category_and_entry_data = self.explode_videos_to_unique_categories(all_video_entries)
for category_path, entry_data in category_and_entry_data:
nav_node = self.create_nav_node(self, category_path)
nav_node.create_video_node(entry_data)
- def explode_videos_to_unique_categories(self, full_video_list):
- for entry_data in full_video_list["entries"]:
+ def load_all_video_entries(self):
+ offset = 1
+ amount = 1000
+ uniq = set()
+ while True:
+ url = append_to_qs(FULL_VIDEO_LIST, {"range": "%s-%s" % (offset, offset+amount-1)})
+ data = grab_json(url)
+ if "entries" not in data:
+ raise Exception("Missing data in SBS response", data)
+ entries = data["entries"]
+ if len(entries) == 0:
+ break
+ for i, entry in enumerate(entries):
+ if entry["guid"] not in uniq:
+ uniq.add(entry["guid"])
+ yield entry
+ offset += amount
+ sys.stdout.write(".")
+ sys.stdout.flush()
+ print()
+
+ def explode_videos_to_unique_categories(self, all_video_entries):
+ for entry_data in all_video_entries:
for category_data in entry_data["media$categories"]:
category_path = self.calculate_category_path(
category_data["media$scheme"],