[vlc-commits] ytdl: extract playlist non-recursively ("flat")

[Rémi Denis-Courmont]

vlc | branch: master | Rémi Denis-Courmont <remi at remlab.net> | Sat Sep 26 20:29:44 2020 +0300| [fa78491d58cac289e63d9d9dbe15ac68c298460b] | committer: Rémi Denis-Courmont

ytdl: extract playlist non-recursively ("flat")

This uses the extract-flat mode of YoutubeDL, which skips parsing
individual items within a playlist. We define a dedicated MRL scheme
to track a YoutubeDL playlist item, and parse it only when the item is
actually opened.

This has two benefits:
1) Extracting a playlist is dramatically faster.
2) Expiring media URL can be played even if the playlist is long.

This does *not* solve the remaining problem that expiring URLs cannot
be saved and replayed later.

> http://git.videolan.org/gitweb.cgi/vlc.git/?a=commit;h=fa78491d58cac289e63d9d9dbe15ac68c298460b
---

 modules/demux/ytdl.c  | 30 +++++++++++++++++++++---------
 share/ytdl-extract.py | 33 ++++++++++++++++++++++++++++++++-
 2 files changed, 53 insertions(+), 10 deletions(-)

diff --git a/modules/demux/ytdl.c b/modules/demux/ytdl.c
index d0066eda24..0fcd0a0c3c 100644
--- a/modules/demux/ytdl.c
+++ b/modules/demux/ytdl.c
@@ -207,18 +207,10 @@ static void Close(vlc_object_t *obj)
     json_free(&sys->json);
 }
 
-static int OpenFilter(vlc_object_t *obj)
+static int OpenCommon(vlc_object_t *obj)
 {
     stream_t *s = (stream_t *)obj;
 
-    if (s->psz_url == NULL)
-        return VLC_EGENERIC;
-    if (strncasecmp(s->psz_url, "http:", 5)
-     && strncasecmp(s->psz_url, "https:", 6))
-        return VLC_EGENERIC;
-    if (!var_InheritBool(s, "ytdl"))
-        return VLC_EGENERIC;
-
     struct ytdl_playlist *sys = vlc_obj_malloc(obj, sizeof (*sys));
     if (unlikely(sys == NULL))
         return VLC_EGENERIC;
@@ -257,6 +249,21 @@ static int OpenFilter(vlc_object_t *obj)
     return VLC_SUCCESS;
 }
 
+static int OpenFilter(vlc_object_t *obj)
+{
+    stream_t *s = (stream_t *)obj;
+
+    if (s->psz_url == NULL)
+        return VLC_EGENERIC;
+    if (strncasecmp(s->psz_url, "http:", 5)
+     && strncasecmp(s->psz_url, "https:", 6))
+        return VLC_EGENERIC;
+    if (!var_InheritBool(obj, "ytdl"))
+        return VLC_EGENERIC;
+
+    return OpenCommon(obj);
+}
+
 vlc_module_begin()
     set_shortname("YT-DL")
     set_description("YoutubeDL")
@@ -267,4 +274,9 @@ vlc_module_begin()
     /* TODO: convert to demux and enable by default */
     add_bool("ytdl", false, N_("Enable YT-DL"), N_("Enable YT-DL"), true)
         change_safe()
+
+    add_submodule()
+    set_capability("access", 0)
+    add_shortcut("ytdl")
+    set_callbacks(OpenCommon, Close)
 vlc_module_end()
diff --git a/share/ytdl-extract.py b/share/ytdl-extract.py
index 107a77ec13..bce6a82ee9 100755
--- a/share/ytdl-extract.py
+++ b/share/ytdl-extract.py
@@ -18,6 +18,7 @@
 
 import sys
 import json
+import urllib.parse
 import youtube_dl
 
 class logger(object):
@@ -32,6 +33,7 @@ class logger(object):
 
 def url_extract(url):
     opts = {
+        'extract_flat': True,
         'logger': logger(),
         'youtube_include_dash_manifest': False,
     }
@@ -40,7 +42,36 @@ def url_extract(url):
 
     # Process a given URL
     infos = dl.extract_info(url, download=False)
+
+    if 'entries' in infos:
+        for entry in infos['entries']:
+             if 'ie_key' in entry and entry['ie_key']:
+                 # Flat-extracted playlist entry
+                 url = 'ytdl:///?' + urllib.parse.urlencode(entry)
+                 entry['url'] = url;
+
+    print(json.dumps(infos))
+
+def url_process(ie_url):
+    opts = {
+        'logger': logger(),
+        'youtube_include_dash_manifest': False,
+    }
+
+    dl = youtube_dl.YoutubeDL(opts)
+
+    # Rebuild the original IE entry
+    entry = { }
+
+    for p in urllib.parse.parse_qsl(url[9:]):
+        entry[p[0]] = p[1]
+
+    infos = dl.process_ie_result(entry, download=False)
     print(json.dumps(infos))
 
 url = sys.argv[1]
-url_extract(url)
+
+if url.startswith('ytdl:///?'):
+    url_process(url)
+else:
+    url_extract(url)