[mythtvnz] Python 3 version of mhegepgsnoop.py
OpenMedia Support
support at openmedia.co.nz
Tue May 11 17:08:26 BST 2021
Thanks for the update
I run with a map file rather than the database and I thing you've got a
bug. If I specify "-f map_file" it still tries to connect to the database.
I think you need to change line 891
if py_bind:
to
elif py_bind:
Does that make sense?
> The Python 3 version of mhegepgsnoop.py is now available for download
> from my web server:
>
> http://www.jsw.gen.nz/mythtv/mhegepgsnoop-0.7.0.py
>
> This requires the Python 3 version of the MythTV Python bindings,
> which is the default in MythTV v31.
>
> I have been using it on my main MythTV box (Ubuntu 20.04, MythTV
> v31-fixes) for a while and it is working fine there, but it should be
> considered to be beta software until it has had wider testing.
>
> Please let me know if you have any problems or suggestions.
>
> Changes:
>
> 1) Converted from Python 2 to Python 3. This has allowed the code to
> be simplified in many places where it used to have to use nasty
> workarounds to handle binary data. Now "bytes" and "byte array" types
> are used to handle the binary data directly.
> 2) Fully UTF-8 - including the output from the -v option (so the a in
> Maori will have the macron over it instead of being converted to a ?
> symbol).
> 3) The -t and -T options were not working in my Python 2 version, so I
> have fixed them. It is no longer necessary to use an external tuner
> program such as dvbtune. I recommend using the -T <chanid> option if
> you have TV sources other than FreeviewHD as the -t <channum> can get
> confused if there is more than one channel with the same <channum>.
> 4) I replaced the code in build_modules with a much faster version
> (runs in 1/10th of the time).
> 5) The buffer overruns that used to occasionally happen when Python
> was not fast enough to read the tuner data from the system buffers
> should be fixed. This problem happened more on slower machines and
> when other programs were making the machine busy, and it was also data
> dependent - when there was more data for a particular channel, it was
> more likely to happen. Re-running mhegepgsnoop.py at a later time
> would often work when this bug was being triggered. I increased the
> size of the system buffers. The symptom was an unhandled exception
> that made you think that the Python buffer had overflowed when in fact
> it was the underlying system buffer where that had happened.
> 6) There is a new -b (both) option to do both map file and fuzzy
> mapping of channels. This allows the use of a map file to map
> channels that the fuzzy mapping does not map correctly, but the map
> file only needs to map the problem channels, rather than all channels.
> It works by first mapping all the channels from the map file, then
> removing any channels mapped in that process from both the channels
> read from the MHEG-5 data and the channels read from the database.
> Then the remaining channels from both sources are mapped to each other
> using the fuzzy mapping.
> 7) Detects the presence of the new "deleted" column in the channel
> table and if present, does not map deleted channels.
> 8) If you have more than one source (eg Sky TV), there used to be a
> problem that channels from other than FreeviewHD would be mapped. This
> is now prevented by only mapping channels that are received on
> multiplexes that have a dtv_multiplex.mod_sys value of "UNDEFINED" or
> where mod_sys starts with "DVB-T".
> 9) Removed the obsolete optparse code - argparse is always available
> in Python 3.
> 10) Prevented a hang condition on reading from the tuner for the first
> time. If it is not configured correctly there may never be any data
> to read, so now there is a timeout to prevent a hang.
>
> There are probably a few more little things that I have fixed/changed
> that I can not remember at present.
>
> _______________________________________________
> mythtvnz mailing list
> mythtvnz at lists.ourshack.com
> https://lists.ourshack.com/mailman/listinfo/mythtvnz
> Archives http://www.gossamer-threads.com/lists/mythtv/mythtvnz/
>
More information about the mythtvnz
mailing list