Indexers

Indexers

The Indexers page (sidebar → Indexers, /indexers) keeps trackers and aggregators in wranglarr and pushes them to the arrs you pick (src/indexers.rs).

Prowlarr in the stack first

With the prowlarr profile enabled (Service: Prowlarr), the simplest path is to add your trackers once in Prowlarr. wranglarr registers every arr in it as an application (App Connections), and Prowlarr pushes its indexers to them itself. The page's Prowlarr in this stack panel shows, per arr, whether it is registered.

The sources below are for anything you'd rather not run through Prowlarr: a tracker's own Torznab feed, a usenet indexer, Jackett, NZBHydra2, or a Prowlarr elsewhere.

Sources

Add tracker or aggregator opens the form. A source is one of:

Type	What each picked arr gets
Tracker (Torznab)	one Torznab indexer, named after the source. The API key may be blank.
Usenet indexer (Newznab)	one Newznab indexer, e.g. https://api.nzbgeek.info
Jackett	one Torznab indexer per tracker configured in Jackett, named <tracker> (<source name>). The list comes from Jackett's t=indexers&configured=true, and is read again on every sync.
NZBHydra2	one Newznab indexer; with Torrents too ticked, also a Torznab indexer <source name> (torrents) for Hydra's /torznab feed
Prowlarr (outside this stack)	no indexers from wranglarr: the picked arrs are registered in that Prowlarr as applications (Full Sync) and it pushes its own indexers. Pointing it at the Prowlarr in this stack is refused, because that one is wired already.

Fields:

Field
Name	1–60 letters, digits, spaces and . _ ( ) - + &; unique
URL	as the arrs reach it. A container name such as http://jackett:9117 works inside the stack. For Torznab/Newznab a URL ending in /api is split into base URL + API path /api.
API key	required except for Torznab. Blank when editing keeps the stored key.
Push to	the arrs that get it. All ticked means every arr, including ones enabled later. Prowlarr is never a target.
Priority	1–50, 1 most preferred (default 25, the arrs' own default)
Min. seeders	Torznab indexers only (default 1)
Enabled	off removes the source's indexers from the arrs on the next sync, and keeps the source

Each indexer is created with RSS, automatic and interactive search on, and the arr's default Newznab/Torznab categories:

Arr	Categories
Radarr	2000, 2010, 2020, 2030, 2040, 2045, 2050, 2060, 2070, 2080
Sonarr	5000, 5010, 5020, 5030, 5040, 5045, 5050, 5090; anime 5070
Lidarr	3000, 3010, 3030, 3040, 3050, 3060
Readarr	3030, 7000, 7010, 7020, 7030, 7040, 7050, 7060
Whisparr	6000, 6010, 6020, 6030, 6040, 6045, 6050, 6070, 6080, 6090

Sources, API keys included, are stored in wranglarr's state db (appdata/wranglarr/wranglarr.db, kv indexers:sources), not in wranglarr.toml. The API never returns the keys, only whether one is set.

Syncing

Save & sync saves the source and starts a sync run; deleting a source starts one too; Sync to arrs runs one by hand. A sync, per arr:

• creates each indexer that should be there and isn't, and updates one whose URL, API path, categories, seeders or priority changed. Indexers are matched by name. The arr tests the feed before it saves.
• removes the indexers wranglarr created that are no longer wanted: the source was deleted or disabled, the arr was un-picked, or the tracker is gone from Jackett.

wranglarr records what it created in each arr (kv indexers:pushed), so indexers you added by hand are never touched. If one of yours already has the name a source would use, the sync reports it as a failure and leaves it alone; rename the source to manage it here. If a source can't be read during a sync (Jackett down, say), its indexers are left as they are rather than removed.

For an external Prowlarr, a sync creates or updates an application per picked arr, using the arr's url from wranglarr.toml (for example http://radarr:7878), so that Prowlarr must be able to reach the arrs at that address. Prowlarr tests the application before saving. Un-picking an arr removes its application. Deleting the source leaves the applications in that Prowlarr; the run log says so, and you remove them there.

At startup, if any sources exist, a background thread waits (up to 10 minutes) for the arrs to answer and runs a sync if some arr lacks indexers it should have, for example an arr you just enabled.

The table shows each source against each arr: ✓ (all its indexers present), some, missing, ? (the arr couldn't be read), – (not picked). Below it, the number of indexers each arr has right now.

API

GET    /api/indexers                                 sources (without keys), per-arr state, targets, types, Prowlarr in the stack
POST   /api/indexers/sources                         SourceIn; saves, then starts a sync -> {"source": …, "run": id}
DELETE /api/indexers/sources/{id}                    removes the source, then starts a sync -> {"run": id}
POST   /api/indexers/sync                            -> {"run": id}

SourceIn: {"id": "", "type": "torznab"|"newznab"|"jackett"|"nzbhydra"|"prowlarr", "name": "…", "url": "…", "api_key": "", "apply_to": [], "enabled": true, "priority": 25, "min_seeders": 1, "torrents": false}. An empty id adds a source, an existing one edits it; apply_to is a list of arr app names, [] for every arr. A validation error is a 400. If another system run is in progress, save and delete still apply and return "run": null; the sync starts by itself once that run ends. POST /api/indexers/sync returns 409.