Improvements made to Indexer's API modules, better error handling for missing or incomplete shows, shows that don't have actor or banner info are now properly error handled as well.
I get this error when using API:
{"result":"error", "message": "error while composing output: "utf8 : terça 9:00 : error 3 : error 4 : invalid continuation byte"}
There's a special char "TERÇA"
Don't know why this happens. I had a special date style and even after change it to "yyyy-mm-dd" this error still happens. Don't know why the api builder is still using the translated week day. Also timezone is set to "network"
This error only happens when missing/future has "terça" - Tuesday and "sábado" - saturday weekdays. The other week days there's no special char.
Move local/network setting datetime convert into it's own function.
Function parse_date_time() now returns when possible a correct timezone aware datetime.
Change webapi to use new converter.
Fix Daily Searcher.
Fix saving old DateTime setting.
Add safety check if network_dict is already loaded.
The API should return application/json Content-Type for JSON data instead of application/html. Tornado expects a dict in order to JSON encode and send header "Content-Type:application/json". SR already encodes data plus it supports JSONP. So, the encoded string data is wrapped into a dict, and a test is added at Tornado def write() to unwrap and skip encoding, in order to set the correct content-header and also use JSONP.
Added file HACKS.txt to serve as a reminder for anyone updating the library.
Add anonymous redirect to external links on pages... Add Trending Show, Add Existing Shows, Add New Shows, Coming Episodes, Display Show, Config Notifications, Config Anime, Subtitle Providers, ASCII chart on General Config Advanced, links of newly created Usenet and Torrent providers.
Update xbmc link to Kodi at Config Notifications.
Fix irc protocol link to not use http protocol linkage on Config/Help and align lower right side column to the upper right side text.
Update CHANGES.md
Added line to growl registration which adds common.NOTIFY_GIT_UPDATE as a sickrage notification type.
growl on some systems will not show notification unless it is registered.
Fix white screen of death when trying to add a show that is already in library by replacing "Add Show" button with "In library".
Add show sorting options.
Add handler for when Trakt returns no results.
Fix image links when anchor child images are not found at Trakt.
Add image to be used when Trakt posters are void.
Add send2trash, a small package that sends files to the Trash (or Recycle Bin) natively and on all platforms. If send2trash is found not compatible, the user can use the default delete and manually delete failed send2trash files.
Add General Config/Misc/Send to trash for actions that involve removing shows and log rotation.
Add handling for the exception raised while deleting a show and show folder no longer exists.
Change to rename General Configuration -> "Web Interface" tab as "Interface".
Add a "User Interface" section to the "Interface" tab.
Change consolidate "Date and Time" and "Theme" tab content to "User Interface" section.
Add field in Advanced setting for a custom remote name used to populate branch versions.
Change theme name "original" to "light".
Improve text wording on all UI options under General Configuration.
Improve reduce over use of capitals.
Improve streamline UI layout, mark-up and some CSS styling of all General Configuration tabs.
Fix default top navbar background white-out behaviour on browsers that don't support gradients.
Improve top navbar gradient use for greater cross browser compatibility (e.g. Safari).
Fix dark theme divider between Season numbers on display show page.
Fix main background and border colour of logs on log page.
Fix imdb and three other images rejected by IExplorer because they were corrupt. Turns out that they were .ico files renamed to either .gif or .png instead of being properly converted.
Fix "Subtitle Language" drop down font colour when entering text on the Subtitles Search settings.
Clean up text, correct quotations, use spaces for code lines, tabs for html.
Update changelog.
The main core of this code was lifted from Trakt code and not developed enough to be suitable for purpose given that the Trakt code isn't fully developed either. An issue found leading to this decision was that show statuses were set false instead of the expected default, however, worse was that users had no way of choosing shows to add, was not informed of what was actually going on once a URL was added to the UI, and was not offered options prior to adding shows like the normal add new show work-flow does. Basically, there are too many issues and this feature is too immature for inclusion.
Used regex checks with javascript popup, for checking if valid imdb csv watchlist.
Added test/test_tvdb_api.py for testing the tvdb api for retrieving shows by IMDB id.
TODO: Help for assisting users in getting the correct IMDB url.
maybe, make interval configurable
First changes for imdb watchlist checker
Thrown the couch code overboard and created my own version of an imdb watchlist scraper
Added IMDB thread and test values with an IMDB csv url. These values should be replaced by front end UI settings.
- Removed some tracelogs when unable to download watchlist
- Added settings for enabling and setting of the watchlist csv url
TODO:
- create UI for setting from frontend
- Implement frontend and backend checks for validating the csv url
Default interval is 1 hour.
Unfortunately some of the old code I modified is still not 100% clear to me.. so I can't really oversee if this will work. That's why I suggest, this should be extensively tested, before merging into master!
I broke backlog searching.. Should be working again.
Tested, and seems to be working fine!
Added addiontional check for checking of the self.status is filled with the show status. If not, debug message is generated and the show is not updated.
The function is_manualsearch_in_progress() is introduced for this purpose and only reference in webserve.py.
Previously when a daily search is running, and a manual search is queued behind it, it will wrongly give back that the manual search is running, this could result in unexpected behavior.
