Adam Sutton [Thu, 14 Jun 2012 16:11:03 +0000 (17:11 +0100)]
Some very early attempts at duplicate detection. Really not sure if this is the best way to go and definitley needs more thought but it does at least stop series links from creating lots of duplicates.
Adam Sutton [Thu, 14 Jun 2012 09:10:02 +0000 (10:10 +0100)]
Rename some link variables to be a bit clearer and move to using anon structs to remove some of the code ugliness. May need to replace anon structures with MACRO based solution if people are not happy with using anon structs.
Adam Sutton [Tue, 12 Jun 2012 12:05:39 +0000 (13:05 +0100)]
Partial revert of previous multi-key update. I have decided to leave it to the upstream data providers to fix this (which still makes it my problem) but I have kept some of the simplifications added as part of the previous update.
Adam Sutton [Mon, 11 Jun 2012 16:18:11 +0000 (17:18 +0100)]
More significant changes to the EPG code because I want to include the concept of having multiple URIs per object (to cope with badly merged upstream data).
Adam Sutton [Fri, 8 Jun 2012 21:24:11 +0000 (22:24 +0100)]
Updated to include compat genre support (more will have to wait), DVR code fully updated to use new EPG code. Updated related code and added useful extra inputs to dvr event creation that should make XBMC devs happy.
Adam Sutton [Fri, 8 Jun 2012 14:34:43 +0000 (15:34 +0100)]
Internal automatic detection of the fact the epg objects have been updated. This should remove some burden from user code. However save flag is still returned and epg_updated() still needs to be called.
Andreas Öman [Fri, 8 Jun 2012 05:56:38 +0000 (07:56 +0200)]
htsp: Remove the broken signal status updates (causing crashes)
htsp->hs_s may only be accessed if global_lock is held and obtaining
global_lock in HTSP delivery code is not acceptable as it can take
a long time to acquire that lock
Adam Sutton [Thu, 7 Jun 2012 11:42:57 +0000 (12:42 +0100)]
Another significant reworking (cannot make my mind up ;) ) of the epggrab framework. This time its MUCH simpler. All clever stuff is defered to a unix domain socket interface which allows users to define schedules using system cron and do whatever the hell they like as long as they send the data to the socket (using nc etc.).
Adam Sutton [Wed, 6 Jun 2012 15:02:07 +0000 (16:02 +0100)]
Major update to the way the epggrab configuration is handled. I now have most features avilable via the UI (except the yet to be implemented external interface config).
Adam Sutton [Wed, 30 May 2012 10:01:29 +0000 (11:01 +0100)]
Started work on deleteing epg_channel_t and moving schedule info back into channel_t. This will make working with EIT much simpler, removes a fairly redundant object and will allow me to move epggrab channel mapping into each grabber which will be far more flexible and was what I had originally intended.
Adam Sutton [Mon, 28 May 2012 14:41:56 +0000 (15:41 +0100)]
Add global object hash similar to that used by andoma in original code. Although mine includes ALL objects, its currently only used for broadcast lookup as generally the others will probably be quicker (episodes?) to go via the normal URI based maps.
Adam Sutton [Thu, 24 May 2012 11:02:22 +0000 (12:02 +0100)]
Significant reworking of EPG code to use an OO style, its still rough and there are quite a few bodges to make it work since its not proper OO (but will be if included in TVH3).
Adam Sutton [Wed, 23 May 2012 10:18:32 +0000 (11:18 +0100)]
Simple demo to list alternative broadcasts, this is a bit of a hack but gives a useful demo of the capabilties of the new structure and helps in testing too.
Adam Sutton [Tue, 22 May 2012 15:20:25 +0000 (16:20 +0100)]
Now have full EPG save/load, I have included internal IDs in all objects though I think they are redundant (and except for broadcast they are not saved).