autograbber: implement .excludes.txt (fixes #75)

[webdl] / sbs.py

diff --git a/sbs.py b/sbs.py
index 568ef60ee1d9e5ce8546e30e56c7247edd4c7050..1b6af3a084da80fbe7876d79961f27300c4d4709 100644 (file)
--- a/sbs.py
+++ b/sbs.py
@@ -1,11 +1,11 @@
-#!/usr/bin/env python
-
-from common import grab_html, grab_json, grab_xml, download_hls, Node, append_to_qs
+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_search/v2/?m=1&filters={section}{Programs}"
+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 = {
@@ -20,17 +20,23 @@ class SbsVideoNode(Node):
         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"):
@@ -45,6 +51,14 @@ class SbsVideoNode(Node):
                         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):
@@ -68,16 +82,24 @@ class SbsRootNode(SbsNavNode):
 
     def load_all_video_entries(self):
         offset = 1
-        amount = 500
+        amount = 1000
+        uniq = set()
         while True:
-            url = append_to_qs(FULL_VIDEO_LIST, {"range": "%s-%s" % (offset, offset+amount)})
-            data = grab_json(url, 3600)
+            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 entry in entries:
-                yield entry
+            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: