failed metadata grabber after latest updates

[laughing_man77@…]

Hi,

I'm running myth .26, updated to latest stable release in the repos:

2:0.26.1+fixes.20130916.7892e19-0ubuntu0mythbuntu

When I run the metadata grabber, it just fails with nothing more in the log, than:

E CoreContext xmlparsebase.cpp:350 (ParseChildren) XMLParseBase: Parent is NULL

Running frontend in verbose mode:

.
.
.
2013-09-18 18:04:48.853590 I  Bonjour: Service registration complete: name 'Mythfrontend on server' type '_mythfrontend._tcp.' domain: 'local.'
2013-09-18 18:04:51.693053 I  MythCoreContext: Connecting to backend server: 192.168.1.1:6543 (try 1 of 1)
2013-09-18 18:04:51.694493 I  Using protocol version 75
2013-09-18 18:05:00.373318 E  XMLParseBase: Parent is NULL
2013-09-18 18:05:07.572402 E  XMLParseBase: Parent is NULL
2013-09-18 18:05:11.219387 I  Running Grabber: /usr/share/mythtv/metadata/Movie/tmdb.py -l en -M The Dreamers
2013-09-18 18:05:12.149249 I  No results found for The Dreamers 0 0
2013-09-18 18:05:18.014327 I  Bonjour: De-registering service '_mythfrontend._tcp.' on 'Mythfrontend on server'
2013-09-18 18:05:18.014916 I  RAOP Device: Cleaning up.
.
.
.

I tried running the script from bash:

$ /usr/share/mythtv/metadata/Movie/tmdb.py -l en -M "The Dreamers"
! Error: Unknown error during a Title search (The Dreamers)
Error(Input object has no document: lxml.etree._ElementTree)

This has only started happening this week. It has run fine up until now.

Full pastebin is attached and available at ​http://pastebin.com/NZwU4AYB

Cheers

[laughing_man77@…]

verbose log when opening frontend to fetch movie metadata

[laughing_man77@…]

I should also mention that this a[p[ears to be happening for all movies, not just the one mentioned above.

[stuartm]

tmdb.py is dead, it uses the old TMDB API which has been discontinued, you should have been switched to using tmdb3.py in the latest 0.26-fixes.

Can you switch the grabber to tmdb v3 in the settings and see if that is working?

[laughing_man77@…]

Yes, that fixed it. Thanks.

Should it have automajically switched, or was it up to me?

[Raymond Wagner]

There was a schema update two months past on 0.27 to force the use of tmdb3.py, but there was no such change in 0.26 or previous.

9be80b77c163dd6575684e455dcac978e98d6214

0.26 will default to tmdb3.py, but it will not force the switch for users who have already configured their grabbers. As mentioned in the commit, the V2.1 API went dead this Sunday.

[laughing_man77@…]

cool, cool. Thanks for your time, and sorry - i thought it was a bug.

[Raymond Wagner]

Fixed by d198fd12cdb908c94c3aeb1b13279a0a9445a586