* Add LibraryUpdateRanker
This class provides various functions to generate Comparators that can
be used to order the manga to update.
One such ordering is by relevance:
It prioritises manga that were updated more recently.
Another Ordering is by lexicographic order:
This is the default behaviour.
* Use relevanceRanking scheme
Instead of default(noRanking/lex ranking) now mangaList is sorted with
relevanceRanking.
* Add UI and associated variables & strings for Update Ranking.
* Use user preferences to determine update ranking scheme.
* Refactor relevanceRanking to latestFirstranking.
This name seems to better reflect the ranking scheme and frees up the
name relevanceRanking for future use.
* Set latestFirst scheme as default.
(Changing over from lexicographic scheme)
* Fix 1
[Convert LibraryUpdateRanker to a object.](./files/82f263749f0ae775385b23dd919f1865360db969#r287513539)
[Nitpick: Add lines](./files/82f263749f0ae775385b23dd919f1865360db969#r287540256)
[Replace Java comparator](./files/82f263749f0ae775385b23dd919f1865360db969#r287539976)
[Nitpick: Add local variable](./files/82f263749f0ae775385b23dd919f1865360db969#r287514805)
* Fix 2
[Weird import](./files/82f263749f0ae775385b23dd919f1865360db969#r287513709)
[Default value](./files/82f263749f0ae775385b23dd919f1865360db969#r287540064)
[Use existing Strings](./files/82f263749f0ae775385b23dd919f1865360db969#r287514476)
[Use Library update order](./files/82f263749f0ae775385b23dd919f1865360db969#r287540204)
* * fix cookieManager not clearing cookies properly
* manually clear tracking prefs when !isLogged (e.g. cookies were cleared)
* use full url for removing cookies
* add interceptor for all non-login network calls
* attempt auto login if cookies are missing
* move handling of csrf token to interceptor
* * move methods around to improve readability
* fix TrackSearchAdapter not updating other fields if cover_url is missing
* revert accidental removal of feature in https://github.com/inorichi/tachiyomi/issues/65
* avoid login if credentials are missing
* fix eol
* *separate login flow from rxjava for reuse in sync
* *use less expensive method of finding manga
* *move variable declaration
* formatting
* set total chapters in remote track
* Added second updateTrackLastChapterRead() called whenever a chapter has been read in the reader
* Removed old updateTrackLastChapterRead() so that it's not called twice.
* [Cloudflare] Fix recent CF JS Challenge error that calls DOM
* Replace `atob` to pure js version. (was node.js API which invalid)
* Use `atob` as native function `Base64.decode()``
* Use okio Base64 decoder instead of Android one.
