XMLTV <episode-num> format?

[MCEabused]

Hi Sub
here is an example as requested of a reoccurring recording being skipped due to an deleted (different) program with the same Season & Episode number.
The Tuesday recording is skipped because it's Unique_ID has a match in the [recently deleted] table = 1:11:0 (S2E12)
it would appear the program title info is not used in the unique id Hash Creation

Sorry about the Pictures as I couldn't see the skip in the text logs.

[Graham]

Forgive me but can you clarify what is happening here. My understanding is that NextPVR avoids recording duplicates when "Avoid duplicate recordings where possible" is ticked on in Settings --> Recording

http://www.nextpvr.com/nwiki/pmwiki.php?....Recording

I have just experimented with this and NextPVR does not schedule duplicate recordings i.e. duplicates are not coloured red in the TV Guide and duplicates do not appear in the Pending Recording list.

In the UK we are blessed to have episodes of Big Bang Theory broadcast about 50 times each week (no exaggeration) plus we have "+1" channels that broadcast the same schedule as the original channel but one hour later. I set NextPVR to record "all episodes all channels" for Big Bang Theory and the duplicate episodes on the "+1" channel were not scheduled to record.

My understanding is that the duplicate episodes are identified based on the unique_id column of the EPG_EVENT table ... for me, this contains values like "The Big Bang Theory-S006E014". The duplication prevention works for recordings that have been recently deleted because the RECENTLY_DELETED table has a unique_identifier column that contains values such as "The Big Bang Theory-S005E020". Curiously, NextPVR appears not to prevent duplicates for earlier recordings that have not yet been deleted. It appears that I might be able to record Series 1 Episode 1 of Big Bang in January and then leave the recording undeleted. I can schedule a new series recording for Big Bang in July and NextPVR will record a duplicate of Series 1 Episode 1.

I understand that,for me, NextPVR gets the series and episode numbers used in the unique_id column of the EPG_EVENT table from the XMLTV file. My XMLTV file has values like
"<episode-num system="xmltv_ns">5.7/24.</episode-num>"
NextPVR also populates the season and episode columns of the EPG_EVENT table using the values from the XMLTV file.

My database does not contain any rows that have unique_id(entifier) values in EPG_EVENT or RECENTLY_DELETED that have values that do not contain the show name. The unique_id and series and episode columns are blank for shows where the XMLTV does not contain series and episode values. I have no clue why you have values with no show name in your RECENTLY_DELETED table. Do you have similar values, with no show name, in the unique_id column of EPG_EVENT?

Your screen shot shows a Manual Recording. Do you have similar problems if you do a Recurring Recording by selecting a show in the TV Guide and then selecting Record and then all episodes all channels?

[sub]

Hi Sub
here is an example as requested of a reoccurring recording being skipped due to an deleted (different) program with the same Season & Episode number.

Can you hit the 'update epg' button, let it complete, then post the npvr.log?

[jcjefferies]

I don’t know if this is relevant at all to the various problems but this is my experience “Avoiding duplicate recordings” in the UK. Graham and I use very different methods to avoid duplicate recordings from the same DVB-T transmissions. We both import the EPG from an XMLTV file but I think Graham uses data from the Internet.

I use the data broadcast with the EPG and used by PVRs to record “Series” and “Avoid duplicate recordings”. It is in the format:-

"<episode-num system="xmltv_ns">/XXX./YYY.0</episode-num>"

Where “XXX” is the “Series ID” and “YYY“is the “Unique programme ID” broadcast with the EPG data extracted by EPGcollector. Next PVR does not know about “SeriesID” so I run a script which deletes it and replaces it with a “0” to produce a value below.

"<episode-num system="xmltv_ns">0.YYY.0</episode-num>"

Each broadcaster uses a different data format up to 12-15 characters. At least one broadcaster uses “illegal” characters which were not liked by NextPVR so my script changes the character to something accepted. It has worked well for almost two years and is only spoilt by one channel which gives new Unique IDs to programmes every January.

MCEabused you will never get the series/episode numbers to act as a unique ID in NextPVR. I think if you can use XMLTV GUI that generates a unique alpha numeric string from the series/episode numbers and adds the programme name to produce a “Unique ID”.

Chris

[MCEabused]

Hi Sub
Cheers for looking at this
I found the relevant part in the logs {...already got this episode: 1:11:0 (recently deleted)}
An ID of just the episode + season is going to be insufficient to be unique,
I do realize there is a alternate of the xmltv_ns tag spec that has a sequential EpisodeNumber that might be just in some of the US EPGs.
ref http://xmltv.cvs.sourceforge.net/xmltv/x...iew=markup around line 363

To jcjefferies
For clarity sake, I am not attempting or do I wish to use a program's valid episode # and Season# to "create" a unique ID.
I would rather disable or avoid NPVR Repeat/Deletion Unique_ID logic altogether and have reliable recording.
I was trying to get the series name, episode # and season # from the EPG into the recordings file system name and maybe also the DB.
I was alerting the original poster of the problem I encountered when I enabled episode # and season numbers in the NPVR EPG load.

log snippet
<Rules>
<ChannelOID>0</ChannelOID>
<ChannelName>All Channels</ChannelName>
<StartTime>2014-07-14T03:10:00.0000000Z</StartTime>
<EndTime>2014-07-14T04:10:00.0000000Z</EndTime>
<PrePadding>3</PrePadding>
<PostPadding>4</PostPadding>
<Quality>0</Quality>
<Keep>0</Keep>
<AdvancedRules>
title like &apos;%million dol%&apos;</AdvancedRules>
</Rules>

2014-07-14 15:09:20.565 [DEBUG][1] LoadListingByAdvancedRules(channel='all channels',advancedRules=
title like '%million dol%')
2014-07-14 15:09:20.568 [DEBUG][1] SQL: select * from EPG_EVENT where
title like '%million dol%' order by start_time
2014-07-14 15:09:20.569 [DEBUG][1] - Potential episode at: 14/07/2014 5:25:00 p.m.
2014-07-14 15:09:20.572 [DEBUG][1] ...check IsValid()
2014-07-14 15:09:20.573 [DEBUG][1] ...scheduling episode
2014-07-14 15:09:20.585 [DEBUG][1] - Potential episode at: 14/07/2014 6:25:00 p.m.
2014-07-14 15:09:20.585 [DEBUG][1] ...check IsValid()
2014-07-14 15:09:20.585 [DEBUG][1] ...already got this episode: 1:10:0
2014-07-14 15:09:20.585 [DEBUG][1] existing status: STATUS_PENDING
2014-07-14 15:09:20.585 [DEBUG][1] existing time: 14/07/2014 5:25:00 p.m. - 14/07/2014 6:00:00 p.m.
2014-07-14 15:09:20.585 [DEBUG][1] ...already got this episode scheduled
2014-07-14 15:09:20.585 [DEBUG][1] - Potential episode at: 15/07/2014 5:25:00 p.m.
2014-07-14 15:09:20.585 [DEBUG][1] ...check IsValid()
2014-07-14 15:09:20.585 [DEBUG][1] ...already got this episode: 1:11:0 (recently deleted)
2014-07-14 15:09:20.585 [DEBUG][1] ...already got this episode scheduled
2014-07-14 15:09:20.585 [DEBUG][1] - Potential episode at: 15/07/2014 6:25:00 p.m.
2014-07-14 15:09:20.585 [DEBUG][1] ...check IsValid()
2014-07-14 15:09:20.585 [DEBUG][1] ...already got this episode: 1:11:0 (recently deleted)
2014-07-14 15:09:20.585 [DEBUG][1] ...already got this episode scheduled
2014-07-14 15:09:20.585 [DEBUG][1] - Potential episode at: 16/07/2014 5:25:00 p.m.
2014-07-14 15:09:20.585 [DEBUG][1] ...check IsValid()
2014-07-14 15:09:20.585 [DEBUG][1] ...scheduling episode
2014-07-14 15:09:20.597 [DEBUG][1] - Potential episode at: 16/07/2014 6:25:00 p.m.
2014-07-14 15:09:20.597 [DEBUG][1] ...check IsValid()
2014-07-14 15:09:20.597 [DEBUG][1] ...already got this episode: 1:12:0
2014-07-14 15:09:20.597 [DEBUG][1] existing status: STATUS_PENDING
2014-07-14 15:09:20.597 [DEBUG][1] existing time: 16/07/2014 5:25:00 p.m. - 16/07/2014 6:00:00 p.m.
2014-07-14 15:09:20.597 [DEBUG][1] ...already got this episode scheduled
2014-07-14 15:09:21.632 [DEBUG][1] Recordings.RELOAD_RECORDINGS
2014-07-14 15:09:21.632 [DEBUG][1] Recordings.RELOAD_RECORDINGS@1
2014-07-14 15:09:21.632 [DEBUG][1] LoadList@0
2014-07-14 15:09:21.632 [DEBUG][1] LoadList@setobjects
2014-07-14 15:09:21.641 [DEBUG][1] LoadList@exit
2014-07-14 15:09:21.641 [DEBUG][1] Recordings.RELOAD_RECORDINGS@2
2014-07-14 15:09:21.641 [DEBUG][1] Recordings.RELOAD_RECORDINGS@exit

[sub]

See if this patch helps.

[MCEabused]

Thank you Sub
Looks like that patch nailed it, the skipped 1:11:0 programme is back in the pending list and your new Unique_ID format looks really good.
I'll do some test records/deletes to 100% confirm.

Is there a way or other to access the Unique_ID via Nscripthelper.exe from a Postprocessing.bat ?
Cheers

[sub]

There is no commands NScriptHelper relating to the unique id. If you want to suggest any, I can see about adding.

[MCEabused]

Hi Sub
perhaps something like your existing get description option:
NScriptHelper -UniqueID filename
- Output UniqueID of recording from the EPG
Ex:
c:\Program Files (x86)\NPVR>NScriptHelper -Description "d:\Recordings\Nashville\Nashville_20140716_09301025.ts"
Nashville-0:10:0

But only as a low priority, I can always write some code to parse the XML in the DB or the Alternate data stream instead
Cheers

[sub]

Hi Sub
perhaps something like your existing get description option:
NScriptHelper -UniqueID filename
- Output UniqueID of recording from the EPG
Ex:
c:\Program Files (x86)\NPVR>NScriptHelper -Description "d:\Recordings\Nashville\Nashville_20140716_09301025.ts"
Nashville-0:10:0

But only as a low priority, I can always write some code to parse the XML in the DB or the Alternate data stream instead
Cheers

I've attached an updated version that supports -UniqueID