Changes at BBC Resulting in PVR Duplicates

[Vangelis forthnet]

On Tue Jun 30 02:02:11 BST 2015, I wrote:

> get_iplayer --type=podcast "The Early Music Show"
> yields:
> (humongous snip)
> INFO: 177 Matching Programmes
> (snip)
>  These 177 "The Early Music Show" podcasts
> (MP3 files) are the ones that can be individually
> manually downloaded (in the UK) from:
> http://www.bbc.co.uk/programmes/b006tn49/episodes/downloads
> (snip)
> as for the podcast, it might simply boil down to a manual
> download from the above referenced link.
> (snip)
> the latest 2015 podcast is dated Sun 26 Apr 2015
> (pid=b05s3cjc).

 Hi...
 After examining Budgie's original issue, I was faced
with a GiP issue myself:
Trying to fetch via GiP the podcast I mentioned
(The Early Music Show: Medieval Dance, 26/04/2015)
I found that simply issuing:

get_iplayer --type=podcast "Medieval Dance"
or even
get_iplayer --type=podcast "The Early Music Show - Medieval Dance"

finds nothing :-(
As I am really dumb when it comes to regex, can one of you
suggest a GiP command that would fetch this specific podcast?
This is win32 here, so no grep/sed/awk please...
What I did myself was first issue:

get_iplayer --type=podcast "The Early Music Show"

and then used the "Find" feature of the command prompt
window to search for "Medieval Dance" (without quotes) -
this got me to line:
222118: The Early Music Show - Medieval Dance, Sun, 26 Apr 2015 01:00:00 
+0100,
BBC Radio 3, Music
so then it was a case of:

get_iplayer --type=podcast 222118 -g --no-tag

(side note: I use "--no-tag" because BBC podcast files already come
with a perfectly fine tag/thumbnail, and GiP seems to mess with this tag by 
default...)

On Tue Jun 30 11:48:28 BST 2015, George Eycott wrote:

> The BBC changed their podcast system a few weeks ago
> (snip)
> http://www.bbc.co.uk/faqs/podcast_changes_2015
> (snip)
> http://iplayerhelp.external.bbc.co.uk/help/playing_radio_progs/podcast_changes

Many thanks George for those links!
Link 1:
> The RSS feeds you use to subscribe to podcasts have been redirected to new 
> URLs.
> This is to make sure your existing podcast series subscriptions still 
> work.
> (snip)
> some podcast management apps/clients will download
> all the existing podcast episodes for each series subscribed.
> It's due to how the apps behave with the RSS feed
> (used for subscribing to a programme) being redirected.

GiP is relying upon the old podcast (RSS?) feeds, so
it would appear it's acting like the aforementioned
podcast management apps/clients and causes
Budgie the issue he reported...
I am unfamiliar with the podcast section of the GiP code,
so this is just an educated guess on my part...

Link2:
> Podcast file names are created from the unique code
> that we use to identify all BBC Programmes and digital content,
> (snip)
> A current example of a Podcast file name is 'p02swmcq.mp3'

After the recent BBC changes, every podcast file is treated as
just another form of radio (audio) offer from the beeb and is hence
assigned a unique PID (just if it were a radio/TV programme on iplayer).
My sample "Medieval Dance" podcast has pid=b05s3cjc
however, GiP currently has no support for downloading
podcasts via the --pid method...
My podcast's page is:
http://www.bbc.co.uk/programmes/b05s3cjc
and there's a direct DL link there, but it would be nice if

get_iplayer --type=podcast --pid=b05s3cjc --no-tag

worked...
To this effect, I've explored a bit further the related beeb code:
The "PID.xml" page that already GiP 2.93+ parses by default
to harvest vpid and other metadata is indeed the source for
the podcast's vPID:
http://www.bbc.co.uk/programmes/b05s3cjc.xml  =>

<version 
canonical="0"><pid>p02slpz1</pid><duration>3020</duration><types><type>Podcast 
version</type></types></version>

=> vpid=p02slpz1

Then, one should use the mediaselector/5 API URL
with a new mediaset=audio-nondrm-download to
get access to the actual download links:

http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/audio-nondrm-download/vpid/p02slpz1

(2 CDNs, Akamai & Limelight).

Shouldn't be that hard to patch GiP's code,
if this feature were to be implemented...
(DP, your thoughts?
Nigel Taylor, maybe your perl skills?)

And I suspect "audio-nondrm-download" is a prelude to
"audio-drm-download", the encrypted audio downloads
the BBC announced would've been available in spring of 2014!

Regards,
Vangelis.