Tvheadend

I'm talking about the Tvheadend client addon in Kodi having channel numbers that start from 10000, both in the Channels view and the EPG (it looks similar to your TVGUIDE screenshot but the numbers are in the 10000s and there is no show data).

The tvheadend server web interface is showing all the Locast channel numbers as 0 in the Channels tab. The EPG source field is blank.

I'm assuming that's a space between -u and http://your ip address:6077/xmltv.xml.

The TVHeadend log shows:

2021-06-06 14:51:27.813 xmltv: /storage/.kodi/addons/service.tvheadend42/bin/tv_grab_url: grab /storage/.kodi/addons/service.tvheadend42/bin/tv_grab_url
2021-06-06 14:51:27.824 spawn: Executing "/storage/.kodi/addons/service.tvheadend42/bin/tv_grab_url"
2021-06-06 14:51:27.833 spawn: /storage/.kodi/addons/service.tvheadend42/bin/tv_grab_url: line 6: can't open 1: no such file
2021-06-06 14:51:27.841 spawn: wget: unrecognized option '--retry-connrefused'
2021-06-06 14:51:27.844 spawn: BusyBox v1.32.1 (2021-06-06 11:42:06 CEST) multi-call binary.
2021-06-06 14:51:27.844 spawn: Usage: wget [-c|--continue] [--spider] [-q|--quiet] [-O|--output-document FILE]
2021-06-06 14:51:27.844 spawn: [-o|--output-file FILE] [--header 'header: value'] [-Y|--proxy on/off]
2021-06-06 14:51:27.844 spawn: [--no-check-certificate]
2021-06-06 14:51:27.844 spawn: [-P DIR] [-S|--server-response] [-U|--user-agent AGENT] [-T SEC] URL...
2021-06-06 14:51:27.844 spawn: Retrieve files via HTTP or FTP
2021-06-06 14:51:27.844 spawn: --spider Only check URL existence: $? is 0 if exists
2021-06-06 14:51:27.844 spawn: --no-check-certificate Don't validate the server's certificate
2021-06-06 14:51:27.844 spawn: c Continue retrieval of aborted transfer
2021-06-06 14:51:27.844 spawn: -q Quiet
2021-06-06 14:51:27.844 spawn: -P DIR Save to DIR (default .)
2021-06-06 14:51:27.844 spawn: -S Show server response
2021-06-06 14:51:27.844 spawn: -T SEC Network read timeout is SEC seconds
2021-06-06 14:51:27.844 spawn: -O FILE Save to FILE ('' for stdout)
2021-06-06 14:51:27.844 spawn: -o FILE Log messages to FILE
2021-06-06 14:51:27.844 spawn: -U STR Use STR for User-Agent header
2021-06-06 14:51:27.844 spawn: -Y on/off Use proxy
2021-06-06 14:51:27.845 xmltv: /storage/.kodi/addons/service.tvheadend42/bin/tv_grab_url: no output detected
2021-06-06 14:51:27.845 xmltv: /storage/.kodi/addons/service.tvheadend42/bin/tv_grab_url: grab returned no data

When I type the command: ./tv_grab_url -u http://192.168.1.83:6077/xmltv/xml

I get the below response:

CoreELEC:~/.kodi/addons/service.tvheadend42/bin # ./tv_grab_url -u http://192.168.1.83:6077/xmltv/xml
./tv_grab_url: line 6: can't open 1: no such file
wget: unrecognized option '--retry-connrefused'
BusyBox v1.32.1 (2021-06-06 11:42:06 CEST) multi-call binary.

Usage: wget [-c|--continue] [--spider] [-q|--quiet] [-O|--output-document FILE]
[-o|--output-file FILE] [--header 'header: value'] [-Y|--proxy on/off]
[--no-check-certificate]
[-P DIR] [-S|--server-response] [-U|--user-agent AGENT] [-T SEC] URL...

Retrieve files via HTTP or FTP

--spider        Only check URL existence: $? is 0 if exists
        --no-check-certificate  Don't validate the server's certificate
        c              Continue retrieval of aborted transfer
        -q              Quiet
        -P DIR          Save to DIR (default .)
        -S              Show server response
        -T SEC          Network read timeout is SEC seconds
        -O FILE         Save to FILE ('' for stdout)
        -o FILE         Log messages to FILE
        -U STR          Use STR for User-Agent header
        -Y on/off       Use proxy

Line 6 of my tv_grab_url file is:

if (( $# < 1 ))

I think you have a 'slash' instead of a period
Try this:

./tv_grab_url -u http://192.168.1.83:6077/xmltv.xml

G Kazaroth wrote:

Thank you for the information and working this. The change will help others using minimal Linux OSes. It looks like the wget executable for CoreELEC does not have the options, so I can make a change and have you pull it down to try it, or you can make the change yourself which is probably the simple answer. If you can change the line in the grabber file where the "wget" command is from

wget -q --retry-connrefused --tries=5 --timeout=20 -O - $url

to

wget -q -T 20 -O - $url

According to the Usage statement above, that should work. Let me know if you get the data dump.

First, thanks for your help! I don't mind doing what little I can do to help. It's nothing compared to what you've done.

I'm now just getting the following error:

CoreELEC:~/.kodi/addons/service.tvheadend42/bin # ./tv_grab_url -u http://192.168.83:6077/xmltv.xml
./tv_grab_url: line 6: can't open 1: no such file
wget: download timed out

One thing. This may sound stupid but I couldn't figure out how to directly download the file from Github. I used Windows to cut and paste the text into WinSCP text editor to create the file. I've had problems before where using Windows to copy has caused problems. Could this be the problem?

C Island wrote:

I think you have a 'slash' instead of a period
Try this:

./tv_grab_url -u http://192.168.1.83:6077/xmltv.xml

Yea, I pasted the wrong section into the message above but I did try with the ".".

G Kazaroth wrote:

It is very possible that the editor you used added a control-M to the end of the line. I ran a test with the ctrl-M and got a WARNING in the cabernet log

WARNING: "GET /xmltv.xml%0D HTTP/1.1" 501 -

Since some windows editor add Ctrl-Ms to the end of the line, you could have this issue. Check the cabernet log to see what happened. It looks like the wget command worked, but got an unknown response from cabernet. If this is true, you probably need to learn a little vi since you are working in linux. Here is a brief list of commands once you "vi file"
:q! quits without saving
:qw writes updates and then quits
ctrl-G goto end of file
shift-A goto end of line
ESC exit editor mode
x deletes current character
i starts insert mode
dd deletes current line
yy copies current line
p pastes line from yy

I opened the cabernet.log file in the data/logs folder but it is empty.

I tried re-running the internal EGP grabbers to see what it would do. It got stuck in a loop and kept repeating the following information in the Tvheadend log:

2021-06-06 17:49:57.869 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:49:59.869 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:50:00.653 http: 127.0.0.1: HTTP/1.1 HEAD /imagecache/15 -- 404
2021-06-06 17:50:01.904 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:50:03.635 http: 127.0.0.1: HTTP/1.1 HEAD /imagecache/19 -- 404
2021-06-06 17:50:03.904 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:50:05.910 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:50:06.645 http: 127.0.0.1: HTTP/1.1 HEAD /imagecache/15 -- 404
2021-06-06 17:50:07.933 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:50:09.741 http: 127.0.0.1: HTTP/1.1 HEAD /imagecache/19 -- 404
2021-06-06 17:50:09.933 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:50:11.933 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:50:12.665 http: 127.0.0.1: HTTP/1.1 HEAD /imagecache/15 -- 404
2021-06-06 17:50:13.933 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:38:10.310 subscription: 0025: No input source available for subscription "127.0.0.1 [ | Kodi Media Center ]" to channel "29.1L ION"
2021-06-06 17:38:12.299 http: 127.0.0.1: HTTP/1.1 HEAD /imagecache/15 -- 404 ....

I opened the tv_grab_url file in vi but I don't know how to tell if something is wrong.

Just a couple of general thoughts here worth mentioning, although they may not be all that helpful, but on the off chance...

First, don't just assume that CoreELEC includes bash. Some lightweight, single purpose distributions are built around BusyBox or something similar and only include the sh shell. sh is rather limited compared to bash, so a script written for bash may or may not run in sh (or whatever shell CoreELEC includes). I would enter which bash at a command prompt; if bash is installed it should show a path such as /usr/bin/bash on the following line. If nothing is printed, that means bash is not installed, or is not in your path, and that means some scripts will break.

Second, if you have a file and you are not sure whether it contains Windows-style line endings, you can run the dos2unix command on it. I believe the proper syntax is simply dos2unix filename but I would run man dos2unix to be certain. dos2unix is included in many Linux distributions, or can be added from a distribution's repositories, and you can run which dos2unix to see if you already have it. I think it may even be included in Windows now, but I don't use Windows at all, so can't be certain of that. It's a lot easier than manually searching for ^M characters or something similar (also, if you are like me and prefer nano to vi, you don't have to learn a different editor).

Good progress.

I copied tv_grab_url v 0.8.6a over and ran dos2unix on it (CoreELEC has both bash and dos2unix).

I still get the line 6 no such file error when I re-run the internal EPG grabbers. The cabernet log has the following at the same time:

2021-06-07 07:27:24,629-INFO:web_handler [192.168.1.83] "GET /xmltv.xml HTTP/1.1" 200 -

However, the EPG is populating and things seem to work correctly. I can view shows, record shows, set timers, etc.

My channel numbers still start with 10000 in the Kodi Tvheand client and don't corelate with the actual channel numbers, but I can work around that if necessary.

I'm assuming the best way to get cabernet to run on system restart is to create an autostart.sh file with the correct command, correct?

Thanks guys for all your help!

G Kazaroth wrote:

I think I have answers.

1) channels in Kodi EPG have 10000 numbers.
Follow step 3 and add channel numbers to the number column in TVHeadend

2) You may have OTA grabbers enabled.
I noticed in your log that tvheadend is running scans. This is normally associated with OTA grabbers. In the old days, stations would broadcast tvguide data over the channel itself and could be obtained directly from the station instead of going to zap2it or the like. This is no longer happening and it just causes the tuner to use up streams. Recommend disabling all OTA grabbers in the "EPG Grabber Modules" tab.

3) Still getting the line 6 no such file error.
I agree with Sean. Based on some tests it looks like you do not have BASH shell installed on CoreELEC at the location in the script. type "which bash" and see if it is located at /bin/bash. This is where the script expects to find the executable. If you find it elsewhere, copying it to the /bin location will correct the issue.

1) I'm assuming you mean manually add the channel numbers (don't mind - just want to be clear). However, I have a question. What would happen if I use the same channel numbers as the corresponding OTA channels and record the same show both OTA and from Locast? I want to mainly use OTA for recordings but have Locast be a backup incase my reception drops.

2) I currently have an OTA grabber running as low priority but will disable it.

3) The command "which bash" returns the following:
/usr/bin/bash
Can I just change the first line in the script to "#!/usr/bin/bash"?

Sterling

I believe CoreELEC uses busybox. The 'which' command does not show it is all. If you do a 'ls -al' command you will see that it uses busybox. Here is the output from a CoreELEC box I have.

CoreELEC:~/.kodi/addons # which bash
/usr/bin/bash
CoreELEC:~/.kodi/addons # ls -al /usr/bin/bash
lrwxrwxrwx    1 root     root             7 Jun  6 02:42 /usr/bin/bash -> busybox

BTW: I deleted my previous comments about using a python based tv_grab_url because it does not work on CoreELEC (the script would need to be converted to python3)

TVHeadend is implemented as a Kodi addon in Kodi. Writing a Kodi addon can be somewhat complicated for someone who has not written a Kodi service previously.

The good news is that CoreELEC has 'systemd' that can be used to create your own services which is pretty straightforward.

The directory containing systemd unit files is /storage/.config/system.d. There are some example unit files in that directory.

You can find info about how to create a service using systemd all over the web. Here is an example of a 'how to' https://www.linode.com/docs/guides/start-service-at-boot/

G Kazaroth wrote:

Looks like we are moving in the right direction...
1) The number column does require a number only and cannot use letters. If you use the same number, then Kodi will place the channels next to each other in the list (which is a good thing). Locast will still have the channel number in the channel name, so you will know which is the Locast channel. The ZAP2EPG may not add the channel number to the channel name, but I found this to be deficient since there are numerous channels with the same name (such as COMET). To fix this, I added the channel number to the channel name. I do this even with the Zap2It xmltv.xml file.

2) yes, you can update the top line of the file to point to the correct location of bash. If you could try something first. Try changing the line to a more portable text
[...]
and see if that works. If so, I will make the change.

During my testing of a clean tvheadend install, I did also find that I had made a mistake on settings in the Network area. All of this is going into the configuration notes, but I need to let you know the change. If you edit the Network in TVHeadend and change the view level to at least Advanced, you will see "Re-fetch period (mins)" and is set to 60. There is no need to grab channels every hour, so the recommendation is to make it 1-2 days. One day would be 1440, but I would recommend something like 1460 or slightly over 1 day. I will update the default update rate in cabernet to 1 day and that should make it all work.

1) Thanks, I'll go with your suggestion.

2) I tried both "#!/usr/bin/env bash" and "#!/usr/bin/bash" as the first line in the tv_grab_url file. Both still have the line 6 error in TVHeadend. C Island confirmed CorELEC uses busybox but I don't understand the implications of this.

I changed the fetch period to 1460.

C Island wrote:

TVHeadend is implemented as a Kodi addon in Kodi. Writing a Kodi addon can be somewhat complicated for someone who has not written a Kodi service previously.

The good news is that CoreELEC has 'systemd' that can be used to create your own services which is pretty straightforward.

The directory containing systemd unit files is /storage/.config/system.d. There are some example unit files in that directory.

You can find info about how to create a service using systemd all over the web. Here is an example of a 'how to' https://www.linode.com/docs/guides/start-service-at-boot/

Systemd looks like a little bit of a learning curve for me, which I'm willing to do if necessary.
Is there a problem with using the Python3 tvh_main.py in the autostart.sh file? I don't understand the difference between a service and using autostart.sh.

autostart.sh should work as well. I have used systemd directly but not autostart.sh. I think autostart.sh is a simple service run by systemd so it should work.....

G Kazaroth wrote:

Hi Sterling, I will try to convert the script to Bourne shell and see if that will work. Give me some time and I will get back to you.

Sorry, but I had to let everyone know. I checked edit4ever tv_grab* file that you are using and he uses #!/bin/bash in the top line. However, his script has no bash specific lines in it; it is all bourne shell.

Thanks!

G Kazaroth wrote:

0.8.6b has the updated script written in Bourne Shell. Give it a try and let me know.

That appears to have worked! No errors and a list of updated channels when I forced a re-run of the internal grabbers.

Also, for any CoreELEC user that comes across this thread in the future, the following autostart.sh script seems to work when put in the /storage/.config folder (I unzipped cabernet to a folder called tvheadend-locast in storage):

#!/bin/sh
(
sleep 10
python3 /storage/tvheadend-locast/tvh_main.py
) &

Hi, great to see you are working on this again. Could you perhaps explain a bit further what the scheduler does, and/or what is the advantage over the previous version? Or is this just an interim step needed for some future enhancements?

If I install 0.8.7 over version 0.8.6d will there be any issues, or will I need to adjust my configuration in any way?

Thanks!

I had a power outage while I was on vacation. When I came back my Tvheadend box had rebooted but was having problems, didn't do any recordings after the outage, etc. I also lost all my custom channel groups on my client box.

I've got everything working again except for Locast. I get a no signal error when I try to watch a Locast channel, both from the guide and if I try to watch the channel directly. I get an unknown error message if I try to watch a Locast channel using the Tvheadend browser. I can watch Locast from the app on my Android box.

I can enter the Cabernet webpage and everything looks OK except that the Locast password field is blank (I don't know if this is normal). I re-entered the password and rebooted but it didn't help, and the password field is blank again.

Any recommendations on what I should do? Should I try upgrading to version 0.8.7?

G Kazaroth wrote:

You should not have to upgrade. Lets fix the issue first.
First, the password is always blank since we do not send your password to the browser as a security reason.

Much of below requires an expert level display. You can permanently change it to expert for now by going to the website > Internal > Display and changing the Display Level to 2-Expert.

Also, lets turn on debug local logging by going to the website > Logging > and change the display level in the far right from 0-Basic to 2-Expert. Then go to File Log Handler and enable it and change the level to DEBUG. The log file will be in the <install>/data/logs/ folder. Restart the app and see what the log indicates.

You might have seen a log entry indicating the locast account is locked. Check if locast locked out your account during the issue by going to the cabernet website > Providers > Locast and checking the login_invalid value (again change the display level to 2-Expert). It should be blank. If it is not, then the app got a message from locast that your account was locked. Try clearing the login_invalid field and restarting the app.

The logs should indicate the issue.

The problem was I had a number in the invalid login field and I guess my account was locked (although I could use the same Locast account on other devices). I didn't try reviewing the log once that fixed it.

Thanks

G Kazaroth wrote:

Well, ok. You asked for the data dump. I will try to answer the questions...

Thanks. These kinds of under-the-hood improvements are appreciated even if they don't seem to make much of a visible difference to the end user.

I was wondering, at one time I thought you had said something about the possibility of adding other sources besides Locast - was that something I imagined or were you still thinking about doing that? The reason I ask is because I saw an announcement that Plex added some new free channels, so I went to their site and saw that they have a guide that goes out for a week (https://app.plex.tv/desktop/#!/media/tv.plex.provider.epg) and they do actually have some interesting channels. It would be useful if those channels could appear as channels in Tvheadend but I have no idea how difficult that might be, or if that's something you'd even have any interest in doing.

EDIT: It appears that each channel has its own m3u8 link that can be found using the web browser's Developer Tools to monitor the network connections, but I would have no idea how you'd find those links other than by playing a little from each channel individually. And I also don't know if those links change periodically.

That's great, looking forward to it! Thanks!

By the way, just so there is no confusion, I do not now and do not want to install or run the Plex software, my desire was to see the free Plex channels that you can watch on the Plex web site appear as channels in Tvheadend. It sounds like that's where you are headed, but I've stumbled across several programs that do the opposite (make Tvheadend channels available in Plex) and that's not what I'm wanting to be able to do.

Just wanted to thank you for your efforts. I'm in a "If it's not broke, don't fix it" mode right now but I'll probably try the updates in the future.

Hey guys, I am in need of some help. I have followed the directions within this script and I cannot get it going. When I enter cabernet port 6077, my m3u list is blank. This is the message that I get within terminal when I eecute the script:

me@tvheadend:~/locast/tvheadend-locast-0.8.8a$ python3 tvh_main.py
2021-08-01 17:50:28,936-WARNING:main #########################################
2021-08-01 17:50:28,936-WARNING:main MIT License, Copyright (C) 2021 ROCKY4546
2021-08-01 17:50:29,951-WARNING:user_config INVALID VALUE (0) FOR CONFIG ITEM [locast_default][epg-episode_adjustment]
Process Process-1:
Traceback (most recent call last):
File "/usr/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/usr/lib/python3.8/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/me/locast/tvheadend-locast-0.8.8a/lib/clients/web_admin.py", line 205, in start
WebAdminHttpHandler.start_httpserver(
File "/home/me/locast/tvheadend-locast-0.8.8a/lib/clients/web_handler.py", line 197, in start_httpserver
server_socket.bind((_plugins.config_obj.data['web']['bind_ip'], _port))
OSError: [Errno 98] Address already in use
Process Process-2:
Traceback (most recent call last):
File "/usr/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/usr/lib/python3.8/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/me/locast/tvheadend-locast-0.8.8a/lib/clients/web_tuner.py", line 237, in start
TunerHttpHandler.start_httpserver(
File "/home/me/locast/tvheadend-locast-0.8.8a/lib/clients/web_handler.py", line 197, in start_httpserver
server_socket.bind((_plugins.config_obj.data['web']['bind_ip'], _port))
OSError: [Errno 98] Address already in use

Can anyone help me out? Kudos to the developer for writting such a brilliant script.

Thanks,
Lisa

Thanks for replying.

I rebooted my server and ran the script again. The terminal displayed the following: 2021-08-01 20:55:38,079-WARNING:main #########################################
2021-08-01 20:55:38,079-WARNING:main MIT License, Copyright (C) 2021 ROCKY4546
2021-08-01 20:55:39,317-WARNING:user_config INVALID VALUE (0) FOR CONFIG ITEM [locast_default][epg-episode_adjustment]
Process Process-1:
Traceback (most recent call last):
File "/usr/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/usr/lib/python3.8/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/hts/locast/lib/clients/web_admin.py", line 205, in start
WebAdminHttpHandler.start_httpserver(
File "/home/hts/locast/lib/clients/web_handler.py", line 197, in start_httpserver
server_socket.bind((_plugins.config_obj.data['web']['bind_ip'], _port))
OSError: [Errno 98] Address already in use
Process Process-2:
Traceback (most recent call last):
File "/usr/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/usr/lib/python3.8/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/hts/locast/lib/clients/web_tuner.py", line 237, in start
TunerHttpHandler.start_httpserver(
File "/home/hts/locast/lib/clients/web_handler.py", line 197, in start_httpserver
server_socket.bind((_plugins.config_obj.data['web']['bind_ip'], _port))
OSError: [Errno 98] Address already in use

I logged into port 6077 and now the m3u and epg is listed, but the m3u files will not play in vlc. I loaded the m3u list into tvheadend and the muxes appear but they will not save for mapping when I run the automatic iptv. I tried to enter the m3u manually but I am in the same boat as the vlc with the m3u not playing.

Is there a log that can be reviewed and where would the file be located in linux? I am almost there and determined to get this baby running.

Thanks,
Lisa

You are correct and I do have 5004 running tvhproxy. Where did you catch that info about 5004 in the terminal? I did not see anything about 5004.

Thanks
Lisa