Automatic Video Library Manager for TV Shows

While saving subtitles there was not specified encoding. As a result saved subtitles with specials characters (e.g. polish signs: ąężźćńół) were not correctly encoded.

This function check for episode on torrent/nzb and set status on available if found. Tested only with torrent. For color code of new episode no preference, if you want to change it its ok.

• [x] if not curShow.paused chenge check
• [x] move log inside markAvailable procedure

I added the two following methods to SickRage's API:

• sb.checkversion:
  • Get information about the current version
  • Get information about the latest version
  • Check if an update is required
  • Get the number of commits between the current version and the latest version
• sb.update: update SickRage if necessary

Usage

sb.checkversion

Request:

GET /?cmd=sb.checkversion

Response:

{
    "data": {
        "commits_offset": 0, 
        "current_version": {
            "branch": "develop", 
            "commit": "ddc0fca21d9030612a5c810ecc24463a7f64112b", 
            "version": "v4.0.24"
        }, 
        "latest_version": {
            "branch": "develop", 
            "commit": "3565ec70dc924d8117f506983fc18a9351f487e4", 
            "version": "v4.0.24"
        }, 
        "needs_update": false
    }, 
    "message": "", 
    "result": "success"
}

sb.update

Request:

GET /?cmd=sb.update

Response:

// If no update is necessary
{
    "data": {}, 
    "message": "SickRage is already up to date", 
    "result": "failure"
}

// If the config could not be backed up
{
    "data": {}, 
    "message": "SickRage could not backup config ...", 
    "result": "failure"
}

// If an update was started
{
    "data": {}, 
    "message": "SickRage is updating ...", 
    "result": "success"
}

I'm not an expert in Python, but I think that my changes in sickbeard/versionChecker.py can be improved.

Reference issue: SiCKRAGETV/sickrage-issues#1780

PR's helping me do this conversion are most welcome. The changes in https://github.com/SiCKRAGETV/SickRage/commit/08ea0ec030c64b2a479604747e4f7016c14d7a57 are a good example of a starting point, although it is not complete and bugs will need worked out once everything is changed over.

This will stay in this branch until complete and production ready.

It needs to be tested for potential performance issue before being included in develop/master. I only have 2 shows with files in my dev environment, so I am not sure about the result.

Check if show are updating and/or SR is searching for episode before auto-update or manual update runs. Previous checks were only if PP were running and showsupdating (this PR fixes the forced-update check_

I am still working on this PR. But as I found a couple of differences between the API and the web UI, I decided to open this PR to track the changes and see which version is the best and validate the choices.

• Qualities definition:
  • API (source): qualList = Quality.DOWNLOADED + Quality.SNATCHED + [ARCHIVED, IGNORED]
  • Web UI (source): qualList = Quality.DOWNLOADED + Quality.SNATCHED + [ARCHIVED, IGNORED]
  • Solution: @miigotu confirmed that Quality.ARCHIVED should be used (https://github.com/SiCKRAGETV/sickrage-issues/issues/2567#issuecomment-138367874)
• Sorting methods:
  • API (source): only sort by the required field
  • Web UI (source): sort by the required field and then by time
  • Solution: I chose the web UI option. However the date sort might need to use airdate rather than localtime. What do you think?
• Date time format:
  • API: %Y-%m-%d %H:%M (for future and show commands)
  • Web UI: %Y-%m-%d %H:%M:%S
  • Solution: Use the Web UI format (%Y-%m-%d %H:%M:%S) everywhere (https://github.com/SiCKRAGETV/SickRage/pull/2438#issuecomment-139912848)

The rest of the logic looks similar, except in the SQL queries. The API only select what is needed (at least it looks like it does), while the web UI selects everything. I will try to improve that to select only what we need.

https://github.com/SiCKRAGETV/sickrage-issues/issues/2567

This pull request is intended to take in sync sickrage with wathclist. It adds Show to watch list when added to sickrage and remove from watchlist when deleted. It add Episode to watchlist when the episode is set to wanted or snetched on sickrage and will be removed from watchlist when the episode will be downloaded.

For now it works with an empty search string In the future real RSS must be used

@miigotu There's still a bug - that I thought I had fixed - about HDT only scrapping the first torrent in the table of torrents. the 'torrents' variable contains only one torrent.

@duramato

https://github.com/SiCKRAGETV/sickrage-issues/issues/292

I've added support for parsing specials from the TVRage API. There are a few limitations due to the results returned from TVRage:

• Each episode listed has a season listed corresponding to the season during which the special was released
• There are no episode numbers associated with each episode

Right now I use the season information to put the special in the correct season, but I must come up with the episode number myself. I simply keep a counter for each season and increment it (starting from 1) every time a special is seen, and use that as the episode number. This is dependent upon the order in which TVRage supplies us the episodes, but I believe this is what they do themselves as I get the same results, which is independent of air date. Because I use the season information, and I don't want to overwrite any episodes, I use 'S<special_num>' as the episode number. This isn't very desirable and I'd love to hear alternative ways to get around this, especially if there is some reason this must be an integer. I don't think stuffing all the episodes into Season 0 (i.e. Specials) is the way to go considering TVRage gives us season information, but I'm not picky on this point.

I've tested this on Top Gear, and get results agreeing with TVRage.

• [x] Pull metadata from TVRage
• [ ] Modify search behaviour for each provider to ignore episode number and search by name

HD-Torrents has some invalid html on the page with search results. Using the default html parser wasn't returning the correct data. Substituted it with html5 parser to fix the problem.

P.S. I've also created another PR with a different fix to that issue. Not sure which one would you prefere.