Some series/episode data formats not recognised within description

[sjabaker]

I'm using the DVB EPG from UK Freeview, and have noticed that while for many/most shows any series+episode numbering contained in the programme's description is identified and extracted into "proper" episode numbering - there are quite a lot of shows where this is not happening even though the relevant information *is* in the description.

Based on which are and aren't, it looks like just a case of slightly differing formats and I'm hoping that the appropriate regex (I'm assuming that's how it's done) can be updated/refined in a future version to handle these additional cases...

Example working formats: "...(S1 Ep1/8)..." & "...(S1 Ep1)..." & "...S1 Ep1"

Example formats not being recognised:
"...(S1 Ep 1)..."           - e.g. used by '5USA' channel
"...(S1, ep 1)..."          - e.g. used by 'Pick' & 'Sky Arts' channels
"...S1, ep 1/8..."          - e.g. used by 'Forces TV' channel (at least 1 series but not sure if it's common - just found it when I searched on "S1," to find examples!)

It may just be the space before the episode number throwing it off (only factor common to them all), though the comma after the series number &/or the lowercase "ep" might also need checking! (If it isn't simply this difference in formats, I'm happy to provide fuller source data or whatever else might be needed for diagnosis.)

Thanks,
Steve


P.S. - there's also lots and lots of programmes (e.g. most BBC series) that only have episode numbers but no series info, where the description starts (e.g.) "1/8. Blah blah..."
Obviously without any series data I'm not sure it makes sense to pull out the numbering data in these cases - but I wonder if using this to create a *subtitle* of "Episode 1" (assuming no subtitle already exists) would be useful in terms of how it would be presented in the NPVR interface and/or in other front-end clients (e.g. Kodi/Plex/whatever) that people use to build libraries from the NPVR recording folder(s)? It might also help with any matching to online sources like TMDB, as this is commonly used where episodes are not individually named. Just something to perhaps consider?

[mvallevand]

Although the single season episode numbering is common in the UK sub noted he didn't want to support it by assuming season 1 https://forums.nextpvr.com/showthread.ph...#pid563841

I suspect that space is not too hard a change with a regex though, but you also have the option of using Scheduled Direct and getting more standardized S/E data and a few other improvements over OTA and it probably would fix all our issues.

Martin

[sub]

NextPVR looks for a pile of different formats commonly used around the world. There is always some data entry person that is going to use some other random format.

"...(S1 Ep 1)..." - e.g. used by '5USA' channel
"...(S1, ep 1)..." - e.g. used by 'Pick' & 'Sky Arts' channels

I just checked these, and they worked for me. They were picked up by this regular expression:

Code:

*\(S ?(?<SeasonNumber>[0-9]+),? [Ee]p? ?(?<EpisodeNumber>[0-9]+)?(?<Part>.*)\)

[sjabaker]

Thanks to you both for such quick responses :-)

@sub:
That regex does look to handle the comma/space/etc. (though not the versions without brackets?) so not surprising it's working for you... but why doesn't it work for me?
I've double checked and can confirm that none of those patterns are working in either of my two NPVR servers (one Windows on v5.2.7.220410 and one on Synology docker, still v5.2.6.220206 as I haven't updated yet). I've re-updated the EPG just now, and earlier today had done so after emptying it completely - neither made a difference.

How can I help diagnose further? I can upload/share any log files, etc., or just a bunch of full xmltv file entries for affected programmes, if you let me know what'll help...
If it makes a difference, in all cases (that I've seen, anyway) there's further text in the description after the S/Ep info? (I'm at the limit of my 'without google' regex understanding here, but that snippet doesn't seem to specify what can/can't follow the S/Ep element)


@Martin:
I agree assuming season 1 in the absence of season data is not a good idea - that's why I said extracting *numbering* didn't make sense ... it was just an idle thought that perhaps reflecting (only) the episode info into an otherwise non-existent *subtitle* field might be useful? (Not particularly fussed over this)

Re: Schedules Direct - I agree it's got overall "better" data, but as it happens in my specific use case I'm explicitly avoiding it!
I run NPVR in parallel with Emby so for "key/important" programmes I have parallel DVRs on entirely separate hardware, OS, platform, recording engine/application, etc. so they don't both get borked by the same error. Another part of that is using separate EPG sources (Emby's comes from Gracenote, same as Schedules Direct), so that I should hopefully avoid missing recordings when the occasional errors in the Gracenote data occur. (They're pretty rare, but I do encounter one or two every now and then.) It should/could also help on those occasions when schedules get changed "at the last minute", like when a royal dies or some big world event suddenly happens - in these cases the OTA broadcast EPG does seem to get updated during the day (even if not quite immediately), whereas the Gracenote data does not and I've previously missed multiple evening / late night recordings due to that.

Again, thanks for your attention.
-Steve

[sub]

If you can use the evaluation version of tscapture (https://tscapture.com) to capture a 1 minute full transport stream from the frequency containing '5USA' or 'Pick' & 'Sky Arts', then put the file somewhere like onedrive or googledrive, and let me know the download link, I'll load it up here and see why those strings might not be parsing.

[sjabaker]

Capture files for the following frequencies should be available at:
Freq 482 (inc. Pick & Sky Arts) - https://1drv.ms/u/s!AtsVBvATpsLhgYmKUsL1...0?e=aL1mE2
Freq 506 (inc. 5USA) - https://1drv.ms/u/s!AtsVBvATpsLhgYmKULTe...w?e=pM6T0m
Freq 746 (inc. Forces TV) - https://1drv.ms/u/s!AtsVBvATpsLhgYmKUY-6...o?e=LoAuK5
(Note: I believe in the UK every mux broadcasts the EPG for all channels, not only those on that mux/freq, so you may even find everything you need in just one of them)

If you need to find particular programmes to check the descriptions/results, I saw the different 'problem' formats mentioned in episodes of: "NCIS" (on 5USA, episodes on other channels use 'working' formats), "Hawaii Five-0" (on Pick) and "Tales Of The Unexpected" (on Sky Arts), and "Oh No It's Selwyn Froggitt" (on Forces TV). Each should have multiple episodes all with the same format. These are just the first examples of programmes I found with non-working formats, there are plenty of others the same, at least for the first two formats.

[mvallevand]

Are you sure those are the requested one minute? Typically a full ts mux capture would be more the 10x longer.

Martin

[sub]

Yeah, that seems to missing the bulk of the data.

[sjabaker]

That's what tscapture produced, and I definitely had the "whole mux, not single channel" option selected... I'll try again and see if I get the same.
(Maybe it's due to whatever causes tscapture to give a .NET framework "unhandled exception" error every time it starts ("only one usage of each socket address is normally permitted"). When I choose to 'continue' anyway it appears to run normally, but...)

I'll report back once it's finished re-scanning for channels and I've tried the captures again.

[mvallevand]

I get that error too (maybe it is trying to open the discovery port?) but did you let it run the evaluation timeout?

Code:

System.Net.Sockets.SocketException (0x80004005): Only one usage of each socket address (protocol/network address/port) is normally permitted
   at System.Net.Sockets.Socket.DoBind(EndPoint endPointSnapshot, SocketAddress socketAddress)
   at System.Net.Sockets.Socket.Bind(EndPoint localEP)
   at System.Net.Sockets.UdpClient..ctor(IPEndPoint localEP)
   at NCapture.SSDP.Start()
   at NCapture.Form1.Form1_Load(Object sender, EventArgs e)
   at System.Windows.Forms.Form.OnLoad(EventArgs e)
   at System.Windows.Forms.Form.OnCreateControl()
   at System.Windows.Forms.Control.CreateControl(Boolean fIgnoreVisible)
   at System.Windows.Forms.Control.CreateControl()
   at System.Windows.Forms.Control.WmShowWindow(Message& m)
   at System.Windows.Forms.Control.WndProc(Message& m)
   at System.Windows.Forms.ScrollableControl.WndProc(Message& m)
   at System.Windows.Forms.Form.WmShowWindow(Message& m)
   at System.Windows.Forms.Form.WndProc(Message& m)
   at System.Windows.Forms.Control.ControlNativeWindow.OnMessage(Message& m)
   at System.Windows.Forms.Control.ControlNativeWindow.WndProc(Message& m)
   at System.Windows.Forms.NativeWindow.Callback(IntPtr hWnd, Int32 msg, IntPtr wparam, IntPtr lparam)

Martin