jacred-go

Go-based multi-tracker torrent aggregator. Port of C# project jacred (mainly from https://github.com/jacred-fdb/jacred)

Collects torrent metadata from 20 Russian/Ukrainian trackers into a unified flat-file database with search, sync, and stats APIs.

Quick Start

# Build for current platform
go build -o ./Dist/jacred ./cmd
./Dist/jacred
# Listens on :9117 by default

# Health check
curl http://127.0.0.1:9117/health

# Parse first page of rutor
curl http://127.0.0.1:9117/cron/rutor/parse

# Search
curl "http://127.0.0.1:9117/api/v1.0/torrents?search=Interstellar"

Build for all platforms

chmod +x build_all.sh
./build_all.sh

Builds binaries for all supported platforms into Dist/ and creates a release archive:

Dist/
  jacred-linux-amd64
  jacred-linux-arm64
  jacred-linux-arm
  jacred-linux-386
  jacred-darwin-amd64
  jacred-darwin-arm64
  jacred-windows-amd64.exe
  jacred-windows-arm64.exe
  jacred-windows-386.exe
  jacred-freebsd-amd64
  jacred-freebsd-arm64

jacred-{version}-{gitSHA}.tar.gz   ← binaries + init.yaml + init.yaml.example
                                     (the web UI is embedded into each binary via //go:embed)

Requires Go 1.21+. All binaries are statically linked (CGO_ENABLED=0), no external dependencies.

Configuration

Configuration file: init.yaml in working directory.

Global Settings

listenip: "any"              # IP to bind ("any" = 0.0.0.0)
listenport: 9117             # Listening port
apikey: ""                   # API key required for /api/v1.0/* (empty = no auth)
devkey: ""                   # Key for dev endpoints (empty = local IP only)
log: true                    # Enable logging
logParsers: false            # Global gate for per-tracker parser logs (both must be true)
logFdb: false                # FDB audit log: JSON Lines per bucket change
logFdbRetentionDays: 7       # Delete FDB log files older than N days
logFdbMaxSizeMb: 0           # Max total FDB log size in MB (0 = unlimited)
logFdbMaxFiles: 0            # Max number of FDB log files (0 = unlimited)
fdbPathLevels: 2             # Directory nesting depth for bucket files
mergeduplicates: true        # Merge duplicate torrents from different trackers
mergenumduplicates: true     # Merge numeric ID variations
openstats: true              # Enable /stats/* endpoints (no auth)
opensync: true               # Enable /sync/fdb/torrents (V2 protocol, no auth)
opensync_v1: false           # Enable /sync/torrents (V1 protocol, no auth)
web: true                    # Serve web UI (index.html, stats.html, settings.html)
timeStatsUpdate: 90          # Rebuild stats.json every N minutes
memlimit: 0                  # Hard cap on Go heap in MB (0 = no limit)
gcpercent: 50                # GC frequency: lower = more GC, less peak RAM (default 50)

Memory Limits

Control Go runtime memory usage. Essential on VPS with limited RAM.

memlimit: 1500   # Hard cap on Go heap in MB; GC becomes very aggressive near the limit
gcpercent: 50    # Go's GOGC knob: 50 = GC at +50% heap growth (default 50, Go default is 100)

Recommended settings by RAM:

VPS RAM	`memlimit`	`gcpercent`	`evercache.maxOpenWriteTask`
1 GB	`700`	`20`	`200`
2 GB	`1500`	`30`	`300`
4 GB+	`0`	`50`	`500`

memlimit: 0 disables the hard cap (Go default behaviour).

CloudFlare Bypass (flaresolverr-go)

For trackers protected by CloudFlare (megapeer, bitru, anistar, anifilm, torrentby, mazepa), the system uses an embedded flaresolverr-go solver — no external Docker service required.

How it works:

The first request to a CF-gated domain spawns a real browser (Camoufox via geckodriver by default, or Chrome for Testing via chromedriver) on a virtual Xvfb display, lets it pass the CF managed challenge, and extracts the resulting cf_clearance cookies and User-Agent.
Solved cookies are cached in memory for 60 minutes per domain and persisted to Data/temp/flare/ so quick restarts reuse them.
Subsequent requests for the same domain use plain net/http with those cookies + the browser's UA (fast path). Only on a fresh CF challenge (403 or interstitial markers in the body) is the browser invoked again.
Idle browser sessions are reaped after 5 minutes of inactivity to free RAM (~800 MB per Camoufox process).

flaresolverr_go:
  headless: true               # true = headless mode, false = visible (still needs Xvfb on servers)
  browser_backend: geckodriver # geckodriver (Camoufox, recommended) | chromedriver (Chrome for Testing)
  browser_path: ""             # Browser binary. Empty + geckodriver = auto-download Camoufox (~680 MB, one-time) into ~/.cache. Empty + chromedriver = system Chrome.
  driver_path: ""              # Custom geckodriver/chromedriver path. Empty = library auto-downloads a matching driver.
  chrome_version: ""           # Pin Chrome for Testing to a major (e.g. "146") — only used with browser_backend: chromedriver.

Requires Xvfb on headless servers (apt install xvfb). jacred starts its own Xvfb instance on :99-:119 if DISPLAY is unset.

Auto-detection

CloudFlare detection is automatic — fetchmode in init.yaml is now an optional hint, not a requirement. When any parser's standard request returns a CF interstitial page (markers like <title>Just a moment, cf-browser-verification, window._cf_chl_opt), the Fetcher:

Logs cf-auto: detected CloudFlare on <domain> — future requests will use flaresolverr.
Flags the domain in an in-memory + on-disk registry (Data/temp/cf_auto.json).
Transparently retries the same request through flaresolverr-go in the same call — the caller gets the real response.
Routes every subsequent request for that domain through flaresolverr automatically (skipping the wasted standard roundtrip).

Setting fetchmode: "flaresolverr" explicitly still works; it just saves the very first wasted request. POST requests are auto-promoted too: when CF rejects a POST, the original body never reached the upstream API, so the retry through flaresolverr (which obtains cf_clearance and re-issues the POST via net/http + browser UA) is safe.

Registry entries persist across restarts and age out after 30 days (any domain that's gone quiet that long is rechecked from scratch). To inspect or clear the registry manually, use /admin/cf-domains (see Database Admin Endpoints).

Megapeer:
  fetchmode: "flaresolverr"   # optional: skip the first detection roundtrip for known CF sites
  host: "https://megapeer.vip"

If a previously-flagged site stops using CF, either wait 30 days or curl -X DELETE 'http://127.0.0.1:9117/admin/cf-domains?domain=megapeer.vip'. A circuit breaker also puts a domain into a 3-minute cooldown after a flaresolverr solve failure to prevent retry storms.

Evercache (In-Memory Bucket Cache)

Keeps recently opened buckets in RAM to reduce disk reads on repeated searches. The cache is hard-capped at maxOpenWriteTask entries; when full the oldest dropCacheTake entries are evicted immediately. Stale entries (older than validHour) are swept every 10 minutes by a background goroutine.

evercache:
  enable: true               # Enable in-memory caching of buckets
  validHour: 1               # Cache TTL in hours; entries older than this are evicted
  maxOpenWriteTask: 500      # Max buckets held in memory (hard cap)
  dropCacheTake: 100         # How many to evict when the cap is hit

Sync (Multi-Instance)

syncapi: "http://other-instance:9117"   # URL of remote jacred to sync from
timeSync: 60                            # Sync interval in seconds
synctrackers:                           # Sync only these trackers (empty = all)
  - "Rutor"
  - "Kinozal"
disable_trackers:                       # Never sync these trackers
  - "Mazepa"
syncsport: true                         # Sync sport torrents
syncspidr: true                         # Enable spider (metadata-only) sync
timeSyncSpidr: 60                       # Spider sync interval in seconds

Per-Tracker Settings

Each tracker section is optional — defaults are used if omitted.

Kinozal:
  host: "https://kinozal.tv"   # Override default host
  cookie: "uid=abc123; pass=..." # Session cookie (required for login-only trackers)
  login:
    u: "username"
    p: "password"
  reqMinute: 8                  # Max requests per minute (rate limiting)
  parseDelay: 7000              # Delay between category/page requests in ms
  log: false                    # Log this tracker's requests
  useproxy: false               # Route requests through globalproxy

Default hosts:

Tracker	Default Host
Rutor	`https://rutor.is`
Megapeer	`https://megapeer.vip`
TorrentBy	`https://torrent.by`
Kinozal	`https://kinozal.tv`
NNMClub	`https://nnmclub.to`
Bitru	`https://bitru.org`
Toloka	`https://toloka.to`
Mazepa	`https://mazepa.to`
Rutracker	`https://rutracker.org`
Selezen	`https://use.selezen.club`
Lostfilm	`https://www.lostfilm.tv`
Animelayer	`https://animelayer.ru`
Anidub	`https://tr.anidub.com`
Aniliberty	`https://aniliberty.top`
Knaben	`https://api.knaben.org`
Anistar	`https://anistar.org`
Anifilm	`https://anifilm.pro`
Leproduction	`https://www.le-production.tv`
Baibako	`http://baibako.tv`
Korsars	`https://korsars.pro`
Ultradox	`https://ultradox.top`

Proxy

globalproxy:
  - pattern: "\.onion"           # Regex: apply proxy when URL matches
    list:
      - "socks5://127.0.0.1:9050"
      - "http://proxy.example.com:8080"
    useAuth: false
    username: ""
    password: ""
    BypassOnLocal: true          # Skip proxy for 127.x / 192.168.x / 10.x

Parser Endpoints

All parser endpoints return:

{
  "status": "ok",
  "fetched": 150,
  "added": 12,
  "updated": 5,
  "skipped": 133,
  "failed": 0
}

Multi-category parsers also include "by_category": [...].

Parsing Strategies

There are five distinct parsing strategies across the 21 trackers:

1. Single-page (`page=N`)

Parse exactly one page. Default is page 0 (most recent) for most trackers; Bitru defaults to page 1.

Trackers: Rutor, Selezen, Bitru, Kinozal, NNMClub, RuTracker, TorrentBy, Toloka, Korsars, Ultradox

Note: Rutor, Bitru, Kinozal, NNMClub, RuTracker, TorrentBy, Toloka also support task-based parsing (see §4 below). The parse?page=N endpoint is available as a single-page fallback.

# Parse the latest page (default)
curl "http://127.0.0.1:9117/cron/rutor/parse"

# Parse a specific page
curl "http://127.0.0.1:9117/cron/kinozal/parse?page=3"

2. Multi-page (`maxpage=N` or `limit_page=N`)

Parse N pages starting from the first. 0 means unlimited (all pages).

Trackers: Megapeer (maxpage, default 1), Animelayer (maxpage, default 1), Baibako (maxpage, default 10), Anistar (limit_page, default 0 = all), Leproduction (limit_page, default 0 = all)

# Megapeer/Animelayer/Baibako: default = 1 page
curl "http://127.0.0.1:9117/cron/megapeer/parse"

# Parse up to 5 pages
curl "http://127.0.0.1:9117/cron/megapeer/parse?maxpage=5"

# Anistar/Leproduction: default = all pages (limit_page=0)
curl "http://127.0.0.1:9117/cron/anistar/parse"

# Limit to 3 pages
curl "http://127.0.0.1:9117/cron/anistar/parse?limit_page=3"
curl "http://127.0.0.1:9117/cron/leproduction/parse?limit_page=3"

3. Page range (`parseFrom=N&parseTo=M`)

Parse pages from N to M inclusive.

Trackers: Anidub, Aniliberty, Selezen

# Parse pages 1 to 5
curl "http://127.0.0.1:9117/cron/anidub/parse?parseFrom=1&parseTo=5"

# Parse only page 1 (single page)
curl "http://127.0.0.1:9117/cron/selezen/parse?parseFrom=1&parseTo=1"

# Aniliberty also returns lastPage in response
curl "http://127.0.0.1:9117/cron/aniliberty/parse?parseFrom=1&parseTo=3"

4. Task-based (incremental, recommended for large trackers)

For large trackers with hundreds of category pages. Works in three steps:

Discover all pages by category and year → stores tasks in Data/{tracker}_tasks.json
Parse all discovered tasks (can be interrupted and resumed)
Parse latest — shortcut to parse only the most recent N pages

Trackers: Rutor, Selezen, Bitru, Kinozal, NNMClub, RuTracker, TorrentBy, Toloka, Korsars, Ultradox

# Step 1: Discover all pages and build task list (run once or periodically)
curl "http://127.0.0.1:9117/cron/kinozal/updatetasksparse"

# Step 2: Parse all discovered tasks (can take a long time)
curl "http://127.0.0.1:9117/cron/kinozal/parsealltask"
curl "http://127.0.0.1:9117/cron/kinozal/parsealltask?force=true"  # ignore "updated today" flag

# Or: Parse only the latest N pages (quick daily update)
curl "http://127.0.0.1:9117/cron/kinozal/parselatest"           # default pages=100 (kinozal); others default pages=5
curl "http://127.0.0.1:9117/cron/kinozal/parselatest?pages=10"

# Fallback: parse a single known page
curl "http://127.0.0.1:9117/cron/kinozal/parse?page=0"

Task state is persisted — interrupted parsealltask resumes from where it stopped.

5. Full vs. incremental (`fullparse=true/false`)

Tracker: Anifilm only

# Incremental: only new/updated since last run (default)
curl "http://127.0.0.1:9117/cron/anifilm/parse"
curl "http://127.0.0.1:9117/cron/anifilm/parse?fullparse=false"

# Full re-parse: all pages
curl "http://127.0.0.1:9117/cron/anifilm/parse?fullparse=true"

Per-Tracker Parse Reference

Rutor

GET /cron/rutor/parse
  page=N   (default 0) — parse page N across all 11 categories

GET /cron/rutor/updatetasksparse      — discover all pages per category
GET /cron/rutor/parsealltask          — parse all discovered tasks
GET /cron/rutor/parselatest
  pages=N   (default 5) — parse latest N pages per category

Parses all 11 categories: movies, music, serials, documentaries, cartoons, anime, sport, Ukrainian content. Without parameters parse fetches one page (page 0) from all categories simultaneously.

Megapeer

GET /cron/megapeer/parse
  maxpage=N   (default 1) — parse up to N pages

Anidub

GET /cron/anidub/parse
  parseFrom=N   (default 0) — page range start
  parseTo=M     (default 0) — page range end

Without parameters: parses only page 0.

Aniliberty

GET /cron/aniliberty/parse
  parseFrom=N   (default 0) — page range start
  parseTo=M     (default 0) — page range end

Response includes: { ..., "lastPage": 42 }

Without parameters: parses only page 0.

Animelayer

GET /cron/animelayer/parse
  maxpage=N   (default 1)

Anistar

GET /cron/anistar/parse
  limit_page=N   (default 0 = parse all pages)
  limitPage=N    (alias)

Anifilm

GET /cron/anifilm/parse
  fullparse=false   (default) — only new/updated since last run
  fullparse=true    — re-parse all pages

Baibako

GET /cron/baibako/parse
  maxpage=N   (default 10)

Bitru

GET /cron/bitru/parse
  page=N   (default 1) — parse single page

GET /cron/bitru/updatetasksparse      — discover all category pages
GET /cron/bitru/parsealltask          — parse all discovered tasks
GET /cron/bitru/parselatest
  pages=N   (default 5) — parse latest N pages only

BitruAPI

GET /cron/bitruapi/parse
  limit=N   (default 100) — number of recent items to fetch via API

GET /cron/bitruapi/parsefromdate
  lastnewtor=YYYY-MM-DD   — fetch items newer than this date
  limit=N                  (default 100)

Kinozal

GET /cron/kinozal/parse
  page=N   (default 0) — parse single page

GET /cron/kinozal/updatetasksparse
GET /cron/kinozal/parsealltask
  force=true   — re-parse all tasks even if marked updated today

GET /cron/kinozal/parselatest
  pages=N   (default 100, max 100) — scan unfiltered pages 0..N-1 per category, no year filter

Knaben

GET /cron/knaben/parse
  from=N            (default 0) — offset in results
  size=N            (default 300) — results per page
  pages=N           (default 1) — number of pages to fetch
  query=string      — search query
  hours=N           (0 = ignore time filter) — only items from last N hours
  orderBy=string    (default "date") — sort order
  categories=a,b,c  — comma-separated category filters

Leproduction

GET /cron/leproduction/parse
  limit_page=N   (default 0 = parse all pages)

Lostfilm

GET /cron/lostfilm/parse            — parse main catalog (latest releases)

GET /cron/lostfilm/parsepages
  pageFrom=N   (default 1)
  pageTo=N     (default 1)

GET /cron/lostfilm/parseseasonpacks
  series=SeriesName   — parse all season packs for a specific series

GET /cron/lostfilm/verifypage
  series=SeriesName   — verify parsed data for a series

GET /cron/lostfilm/stats            — Lostfilm-specific stats

Mazepa

GET /cron/mazepa/parse   — no parameters, parses current page

NNMClub

GET /cron/nnmclub/parse
  page=N   (default 0)

GET /cron/nnmclub/updatetasksparse
GET /cron/nnmclub/parsealltask
GET /cron/nnmclub/parselatest
  pages=N   (default 5)

RuTracker

GET /cron/rutracker/parse
  page=N   (default 0)

GET /cron/rutracker/updatetasksparse
GET /cron/rutracker/parsealltask
GET /cron/rutracker/parselatest
  pages=N   (default 5)

Korsars

GET /cron/korsars/parse
  page=N   (default 0)

GET /cron/korsars/updatetasksparse
GET /cron/korsars/parsealltask
GET /cron/korsars/parselatest
  pages=N   (default 5)

phpBB-mod tracker (films / series / cartoons), 24 forum IDs hardcoded. Login required (Korsars.login.u/p); magnets are inline in the listing so no extra dl.php round-trip per torrent.

Ultradox

GET /cron/ultradox/parse
  page=N   (default 0)

GET /cron/ultradox/updatetasksparse
GET /cron/ultradox/parsealltask
GET /cron/ultradox/parselatest
  pages=N   (default 5)

Listing-then-detail tracker, no login. Six sections: serial-hd, hd, rufilm, camrip, webrips, anime. Listing rows expose placeholder magnets with empty btih, so the parser follows each title link to the detail page where every quality variant has a full info-hash. One torrent record is stored per quality variant. Sid/pir are placeholder values (1) — the site doesn't expose peer counts. Upstream's TLS certificate is expired, so the default config carries insecureSkipVerify: true.

Selezen

GET /cron/selezen/parse
  parseFrom=N   (default 0) — page range start
  parseTo=M     (default 0) — page range end

GET /cron/selezen/updatetasksparse      — discover all pages
GET /cron/selezen/parsealltask          — parse all discovered tasks
GET /cron/selezen/parselatest
  pages=N   (default 5) — parse latest N pages only

Without parameters parse parses only page 1.

Toloka

GET /cron/toloka/parse
  page=N   (default 0)

GET /cron/toloka/updatetasksparse
GET /cron/toloka/parsealltask
GET /cron/toloka/parselatest
  pages=N   (default 5)

TorrentBy

GET /cron/torrentby/parse
  page=N   (default 0)

GET /cron/torrentby/updatetasksparse
GET /cron/torrentby/parsealltask
GET /cron/torrentby/parselatest
  pages=N   (default 5)

Search API

`GET /api/v1.0/torrents`

Full-text torrent search.

Query Parameters:

Parameter	Aliases	Type	Description
`search`	`q`	string	Full-text search in title and name fields
`altname`	`altName`	string	Search in original/alternative name
`exact`	—	bool	Exact match instead of fuzzy
`type`	—	string	Content type: `фильм`, `сериал`, `аниме`, `музыка`, etc.
`tracker`	`trackerName`	string	Filter by tracker name (e.g. `Kinozal`)
`voice`	`voices`	string	Filter by dubbing studio or voice
`videotype`	`videoType`	string	Video format filter (e.g. `hdr`, `sdr`)
`relased`	`released`	int	Release year (e.g. `2023`)
`quality`	—	int	Quality code: `480`, `720`, `1080`, `2160`
`season`	—	int	Season number
`sort`	—	string	Sort order: `date`, `size`, `sid`

# Basic search
curl "http://127.0.0.1:9117/api/v1.0/torrents?search=Interstellar"

# Filter by tracker and quality
curl "http://127.0.0.1:9117/api/v1.0/torrents?search=Dune&tracker=Kinozal&quality=1080"

# Anime by season
curl "http://127.0.0.1:9117/api/v1.0/torrents?search=Naruto&type=аниме&season=5"

# Exact title match
curl "http://127.0.0.1:9117/api/v1.0/torrents?search=Inception&exact=true"

Response:

[
  {
    "tracker": "Kinozal",
    "url": "https://kinozal.tv/details.php?id=123456",
    "title": "Дюна / Dune (2021) BDRip 1080p",
    "size": 15032385536,
    "sizeName": "14.0 GB",
    "createTime": "2021-11-15 12:30:00",
    "updateTime": "2021-11-15 12:30:00",
    "sid": 350,
    "pir": 12,
    "magnet": "magnet:?xt=urn:btih:...",
    "name": "дюна",
    "originalname": "dune",
    "relased": 2021,
    "videotype": "sdr",
    "quality": 1080,
    "voices": "Дублированный",
    "seasons": "",
    "types": ["фильм", "зарубежный"]
  }
]

`GET /api/v1.0/qualitys`

Paginated listing of quality metadata.

Parameter	Default	Description
`name`	—	Filter by name
`originalname` / `originalName`	—	Filter by original name
`type`	—	Filter by type
`page`	`1`	Page number
`take`	`1000`	Items per page

`GET /api/v2.0/indexers/[indexer]/results`

Jackett-compatible API for use with Sonarr, Radarr, etc.

Parameter	Aliases	Description
`query`	`q`	Search query
`title`	—	Title to match
`title_original`	—	Original title
`year`	—	Release year
`is_serial`	—	`1` for series, `0` for movies
`category`	—	Category prefix: `mov_`, `tv_`, `anime_`, etc.
`apikey`	`apiKey`	API key (if configured)

curl "http://127.0.0.1:9117/api/v2.0/indexers/jacred/results?q=Dune&year=2021"

Response: { "Results": [...], "jacred": true }

`GET /api/v1.0/conf`

Validate API key.

curl "http://127.0.0.1:9117/api/v1.0/conf?apikey=your-key"
# Response: {"apikey": true}

Stats API

Stats endpoints are open (no API key required) when openstats: true.

`GET /stats/refresh`

Refresh Data/temp/stats.json

`GET /stats/torrents`

Returns pre-computed stats from Data/temp/stats.json (rebuilt every timeStatsUpdate seconds).

Parameter	Aliases	Default	Description
`trackerName`	—	—	If set: compute on-demand for this tracker
`newtoday`	`newToday`	`0`	`1` = only torrents added today
`updatedtoday`	`updatedToday`	`0`	`1` = only torrents updated today
`limit`	`take`	`200`	Max items to return

# Full stats (from cache)
curl "http://127.0.0.1:9117/stats/torrents"

# Today's new torrents from Kinozal
curl "http://127.0.0.1:9117/stats/torrents?trackerName=Kinozal&newtoday=1"

`GET /stats/trackers`

Per-tracker statistics summary.

Parameter	Default	Description
`newtoday` / `newToday`	`0`	Filter to today's new torrents
`updatedtoday` / `updatedToday`	`0`	Filter to today's updated torrents
`limit` / `take`	`200`	Max items

curl "http://127.0.0.1:9117/stats/trackers"
curl "http://127.0.0.1:9117/stats/trackers?newtoday=1&limit=50"

`GET /stats/trackers/{trackerName}`

Stats for a specific tracker.

curl "http://127.0.0.1:9117/stats/trackers/Rutor"
curl "http://127.0.0.1:9117/stats/trackers/Rutor?newtoday=1"

`GET /stats/trackers/{trackerName}/new`

Torrents added today from the specified tracker.

curl "http://127.0.0.1:9117/stats/trackers/Kinozal/new?limit=100"

`GET /stats/trackers/{trackerName}/updated`

Torrents updated today from the specified tracker.

curl "http://127.0.0.1:9117/stats/trackers/Kinozal/updated"

Sync API

Multi-instance synchronization. Enabled by opensync: true in config.

`GET /sync/conf`

Discovery endpoint — check protocol version.

{ "fbd": true, "spidr": true, "version": 2 }

V2 Protocol

`GET /sync/fdb`

List database bucket keys (low-level, for replication).

Parameter	Default	Description
`key`	—	Substring filter on bucket key (e.g. `matrix`)
`limit` / `take`	`20`	Max entries to return

curl "http://127.0.0.1:9117/sync/fdb?key=matrix&limit=5"

`GET /sync/fdb/torrents`

Incremental sync — returns torrents modified after a timestamp.

Parameter	Aliases	Required	Description
`time`	`fileTime`	Yes	Return only buckets with fileTime > this value (Windows FILETIME format)
`start`	`startTime`	No	Return only torrents with updateTime > this value
`spidr`	—	No	`true` = return only url/sid/pir metadata (lighter payload)
`take`	`limit`	No	Batch size (default 2000)

# Initial sync (time=0 = get everything)
curl "http://127.0.0.1:9117/sync/fdb/torrents?time=0&take=2000"

# Incremental sync using fileTime from previous response
curl "http://127.0.0.1:9117/sync/fdb/torrents?time=133476543210000000"

# Spider mode (metadata only, faster)
curl "http://127.0.0.1:9117/sync/fdb/torrents?time=0&spidr=true"

Response:

{
  "nextread": true,
  "countread": 2000,
  "take": 2000,
  "collections": [
    {
      "Key": "матрица:the matrix",
      "Value": {
        "time": "2024-01-15 10:30:45",
        "fileTime": 133476543210000000,
        "torrents": {
          "https://kinozal.tv/details.php?id=123": { ... }
        }
      }
    }
  ]
}

When nextread: true, call again with the last received fileTime to get more data.

V1 Protocol (Legacy)

Enabled with opensync_v1: true.

`GET /sync/torrents`

Parameter	Aliases	Required	Description
`time`	`fileTime`	Yes	Timestamp filter
`trackerName`	`tracker`	No	Filter by tracker
`take`	`limit`	No	Batch size (default 2000)

curl "http://127.0.0.1:9117/sync/torrents?time=0&trackerName=Rutor"

Response: flat array of { "key": "name:originalname", "value": {...} }

Database Admin Endpoints

`GET /jsondb/save`

Manually flush in-memory database to disk (Data/masterDb.bz).

curl "http://127.0.0.1:9117/jsondb/save"
# "work" — already saving
# "ok"   — saved successfully

A daily backup Data/masterDb_DD-MM-YYYY.bz is created on each save. Backups older than 3 days are auto-deleted.

Dev/Maintenance Endpoints

Available from local IP only (127.0.0.1, ::1, fe80::/10 link-local, fc00::/7 ULA, IPv4-mapped IPv6).

Data Integrity

`GET /dev/findcorrupt`

Scans all buckets for corrupted entries.

Parameter	Aliases	Default	Description
`sampleSize`	`sample`, `limit`	`20`	Max examples per issue type

curl "http://127.0.0.1:9117/dev/findcorrupt"
curl "http://127.0.0.1:9117/dev/findcorrupt?sampleSize=50"

Response:

{
  "ok": true,
  "totalFdbKeys": 12500,
  "totalTorrents": 480000,
  "corrupt": {
    "nullValue":           { "count": 0, "sample": [] },
    "missingName":         { "count": 12, "sample": [...] },
    "missingOriginalname": { "count": 3, "sample": [...] },
    "missingTrackerName":  { "count": 0, "sample": [] }
  }
}

`GET /dev/removeNullValues`

Deletes entries where torrent value is null.

curl "http://127.0.0.1:9117/dev/removeNullValues"
# Response: { "ok": true, "removed": 5, "files": 3 }

`GET /dev/findDuplicateKeys`

Finds bucket keys where name == originalname (potential duplicates).

Parameter	Aliases	Default	Description
`tracker`	`trackerName`	—	Filter: only keys containing this tracker's torrents
`excludeNumeric`	—	`true`	Exclude purely numeric keys

curl "http://127.0.0.1:9117/dev/findDuplicateKeys"
curl "http://127.0.0.1:9117/dev/findDuplicateKeys?tracker=Kinozal&excludeNumeric=false"

`GET /dev/findEmptySearchFields`

Finds torrents with empty _sn (search name) or _so (search original) fields.

Parameter	Aliases	Default	Description
`sampleSize`	`sample`, `limit`	`20`	Max examples per category

curl "http://127.0.0.1:9117/dev/findEmptySearchFields"

Data Updates

`GET /dev/updateDetails`

Recomputes quality, videotype, voices, languages, seasons for all stored torrents.

curl "http://127.0.0.1:9117/dev/updateDetails"

`GET /dev/updateSearchName`

Update search fields for a specific bucket.

Parameter	Required	Description
`bucket`	Yes	Bucket key in format `name:originalname`
`fieldName`	No	Field to update
`value`	No	New value

curl "http://127.0.0.1:9117/dev/updateSearchName?bucket=матрица:the+matrix&fieldName=_sn&value=матрица"

`GET /dev/updateSize`

Update torrent size in a bucket.

Parameter	Required	Description
`bucket`	Yes	Bucket key
`value`	Yes	Size with suffix: `500MB`, `14.7GB`, `1.2TB`, or bytes

curl "http://127.0.0.1:9117/dev/updateSize?bucket=матрица:the+matrix&value=14.7GB"

`GET /dev/resetCheckTime`

Resets checkTime to yesterday for all torrents (forces re-check on next parse).

curl "http://127.0.0.1:9117/dev/resetCheckTime"

Tracker-Specific Fixes

`GET /dev/fixKnabenNames`

Normalizes Knaben torrent names/years/titles and migrates to correct bucket keys.

curl "http://127.0.0.1:9117/dev/fixKnabenNames"

`GET /dev/fixBitruNames`

Cleans Bitru torrent titles (strips quality tags, codec info, release groups).

curl "http://127.0.0.1:9117/dev/fixBitruNames"

`GET /dev/fixEmptySearchFields`

Auto-populates empty _sn/_so search fields and migrates torrents to correct buckets.

curl "http://127.0.0.1:9117/dev/fixEmptySearchFields"

`GET /dev/migrateAnilibertyUrls`

Appends ?hash=<btih> to Aniliberty URLs using magnet link hashes.

curl "http://127.0.0.1:9117/dev/migrateAnilibertyUrls"

`GET /dev/removeDuplicateAniliberty`

Deduplicates Aniliberty torrents by magnet hash, keeping the most recent.

curl "http://127.0.0.1:9117/dev/removeDuplicateAniliberty"

`GET /dev/fixAnimelayerDuplicates`

Normalizes http→https for Animelayer URLs and removes duplicates by hex ID.

curl "http://127.0.0.1:9117/dev/fixAnimelayerDuplicates"

`GET /dev/fixKinozalUrls`

Normalizes http→https for Kinozal URLs and removes duplicates by id=NNN, keeping the most recent.

curl "http://127.0.0.1:9117/dev/fixKinozalUrls"

`GET /dev/fixSelezenUrls`

Normalizes Selezen host (old hosts like open.selezen.org → configured Selezen.Host) and removes duplicates by numeric item ID, keeping the most recent.

curl "http://127.0.0.1:9117/dev/fixSelezenUrls"

Bucket Management

`GET /dev/removeBucket`

Delete an entire bucket (all torrents under a key).

Parameter	Required	Description
`key`	Yes	Bucket key in format `name:originalname`
`migrateName`	No	If set: move torrents to this new name instead of deleting
`migrateOriginalname`	No	New originalname for migration target

# Delete bucket
curl "http://127.0.0.1:9117/dev/removeBucket?key=матрица:the+matrix"

# Rename/migrate bucket to new key
curl "http://127.0.0.1:9117/dev/removeBucket?key=матрица:the+matrix&migrateName=the+matrix&migrateOriginalname=the+matrix"

Web UI

When web: true, three pages are served. The UI assets (HTML, icons, manifest) are embedded into the binary at build time via //go:embed all:wwwroot in server/embed.go — no external wwwroot/ directory is required at runtime. If a wwwroot/ directory exists alongside the binary, individual files in it override the embedded copy (handy for live-editing during development).

Path	Purpose
`/`	`index.html` — torrent search (Kinopoisk / IMDB / title), filters, magnet links, TorrServer launcher
`/stats`	`stats.html` — per-tracker statistics dashboard (new/updated/checked counts, last run time)
`/settings`	`settings.html` — editor for the full `init.yaml` config (server, logging, sync, trackers, proxies)

The Settings page talks to /admin/config (GET to load, POST to save) and is local-only (127.0.0.1 / RFC1918 / link-local / ULA). On save the server writes init.yaml atomically, re-parses it with the same loader used on startup, and calls UpdateConfig to apply new values to the running server, DB, and all 21 parsers — no restart needed.

`GET /admin/config`

Returns the current in-memory config as JSON. Local IPs only.

`POST /admin/config`

Accepts a full config JSON body, overlays it on top of the current config, writes init.yaml (atomic temp-file + rename), then reloads. Local IPs only.

# Example: dump current config
curl -s "http://127.0.0.1:9117/admin/config" > current.json

# Example: tweak and push back
jq '.log = true' current.json | curl -X POST -H 'Content-Type: application/json' \
  --data-binary @- "http://127.0.0.1:9117/admin/config"

Responses: {"ok": true} on success, {"ok": false, "error": "..."} on failure. The JSON writer normalizes to the structure init.yaml parser expects — existing comments in init.yaml are lost on save (the file is regenerated from the config struct).

`GET /admin/cf-domains`

Returns the auto-detected CloudFlare domain registry — domains that returned a CF challenge during a standard fetch and were promoted to flaresolverr routing automatically. Local IPs only.

curl -s "http://127.0.0.1:9117/admin/cf-domains"

{
  "ok": true,
  "count": 2,
  "domains": [
    {"domain": "megapeer.vip", "detected": "2026-04-30T12:34:56Z", "ageHours": 21},
    {"domain": "bitru.org",    "detected": "2026-05-01T08:10:00Z", "ageHours": 2}
  ]
}

`DELETE /admin/cf-domains`

Clears one entry (when ?domain= is supplied) or the whole registry (when omitted). Useful when a site stops using CloudFlare. Returns {"ok": true, "removed": N}. Local IPs only.

# Clear a single domain
curl -X DELETE "http://127.0.0.1:9117/admin/cf-domains?domain=megapeer.vip"
# Clear all
curl -X DELETE "http://127.0.0.1:9117/admin/cf-domains"

Config Hot-Reload

The init.yaml file is checked for changes every 10 seconds. When the file modification time changes, the config is reloaded automatically — no restart needed. Saves from the Settings page use the same reload path.

Hot-reloadable settings include: API keys, logging flags, sync settings, stats update interval, tracker hosts/cookies/credentials, rate limits.

Search Result Caching

Search endpoints (/api/v1.0/torrents and /api/v2.0/indexers/*/results) cache results in memory with a 5-minute TTL. Cache hits return an X-Cache: HIT header. The cache is keyed by the full query string.

FDB Audit Log

When logFdb: true, every bucket change is logged to Data/log/fdb.YYYY-MM-dd.log in JSON Lines format. Each line records the incoming and existing torrent data for the changed entry.

Retention is controlled by logFdbRetentionDays, logFdbMaxSizeMb, and logFdbMaxFiles. Cleanup runs automatically after each masterDb save.

Parser Logging

Two-level logging control:

Global gate: logParsers: true must be set to enable any parser logging
Per-tracker: each tracker's log: true enables logging for that specific tracker

Both must be true for logs to be written. Log files are stored in Data/log/{tracker}.log.

Cron Examples

Typical external crontab (/etc/cron.d/jacred or Data/crontab):

# Lightweight daily update — latest pages only
0  6 * * *  root  curl -s "http://127.0.0.1:9117/cron/rutor/parselatest?pages=3" >/dev/null
5  6 * * *  root  curl -s "http://127.0.0.1:9117/cron/kinozal/parselatest?pages=3" >/dev/null
10 6 * * *  root  curl -s "http://127.0.0.1:9117/cron/rutracker/parselatest?pages=3" >/dev/null
15 6 * * *  root  curl -s "http://127.0.0.1:9117/cron/nnmclub/parselatest?pages=3" >/dev/null
20 6 * * *  root  curl -s "http://127.0.0.1:9117/cron/torrentby/parselatest?pages=3" >/dev/null

# Anime trackers — multiple pages
30 6 * * *  root  curl -s "http://127.0.0.1:9117/cron/animelayer/parse?maxpage=3" >/dev/null
35 6 * * *  root  curl -s "http://127.0.0.1:9117/cron/anidub/parse?parseFrom=1&parseTo=3" >/dev/null
40 6 * * *  root  curl -s "http://127.0.0.1:9117/cron/anistar/parse?limit_page=3" >/dev/null
45 6 * * *  root  curl -s "http://127.0.0.1:9117/cron/anifilm/parse" >/dev/null

# Weekly full re-parse (task-based trackers)
0  2 * * 0  root  curl -s "http://127.0.0.1:9117/cron/kinozal/updatetasksparse" >/dev/null
30 2 * * 0  root  curl -s "http://127.0.0.1:9117/cron/kinozal/parsealltask" >/dev/null

# Force DB save after heavy parse
0  8 * * *  root  curl -s "http://127.0.0.1:9117/jsondb/save" >/dev/null

Database Structure

Data/
  masterDb.bz               # Gzipped JSON index: key → {fileTime, updateTime, path}
  masterDb_DD-MM-YYYY.bz    # Daily backups (auto-created, kept 3 days)
  fdb/
    ab/cdef012...            # Bucket files (gzipped JSON): url → torrent object
  temp/
    stats.json               # Pre-computed stats cache
  log/
    YYYY-MM-DD.log           # Application log (when log: true)
    kinozal.log              # Per-tracker add/update/skip/fail logs
    fdb.YYYY-MM-dd.log       # FDB audit log: JSON Lines per bucket change
  {tracker}_tasks.json       # Task state for incremental parsers

Each bucket file key is derived from MD5 of the bucket key (name:originalname) split into 2-char prefix directory.

Torrent Object Fields

Field	Type	Description
`url`	string	Tracker page URL (primary key within bucket)
`title`	string	Full torrent title as shown on tracker
`name`	string	Normalized Russian/English name
`originalname`	string	Original (usually English) name
`trackerName`	string	Source tracker name
`relased`	int	Release year
`size`	int64	Size in bytes
`sizeName`	string	Human-readable size (e.g. `14.0 GB`)
`sid`	int	Seeders
`pir`	int	Peers/leechers
`magnet`	string	Magnet link
`btih`	string	Info hash
`quality`	int	`480`, `720`, `1080`, `2160`
`videotype`	string	`sdr`, `hdr`
`voices`	string	Dubbing studios
`languages`	string	`rus`, `ukr`
`seasons`	string	Season list, e.g. `1-5` or `1,3,7`
`types`	[]string	Content type tags
`createTime`	string	First seen timestamp
`updateTime`	string	Last updated timestamp
`checkTime`	string	Last checked/parsed timestamp
`_sn`	string	Search-normalized name
`_so`	string	Search-normalized original name

Utility Endpoints

GET /health          → {"status": "OK"}
GET /version         → {"version": "...", "gitSha": "...", "gitBranch": "...", "buildDate": "..."}
GET /lastupdatedb    → last database write timestamp
GET /                → index.html (if web: true)
GET /stats           → stats.html (if web: true)
GET /settings        → settings.html (if web: true)

Name		Name	Last commit message	Last commit date
Latest commit History 164 Commits
app		app
background		background
cmd		cmd
core		core
cron		cron
filedb		filedb
server		server
tracks		tracks
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
README.ru.md		README.ru.md
build_all.sh		build_all.sh
crontab		crontab
go.mod		go.mod
go.sum		go.sum
init.yaml		init.yaml
init.yaml.example		init.yaml.example
jacred.service		jacred.service

Folders and files

Latest commit

History

Repository files navigation

jacred-go

Table of Contents

Quick Start

Build for all platforms

Configuration

Global Settings

Memory Limits

CloudFlare Bypass (flaresolverr-go)

Auto-detection

Evercache (In-Memory Bucket Cache)

Sync (Multi-Instance)

Per-Tracker Settings

Proxy

Parser Endpoints

Parsing Strategies

1. Single-page (page=N)

2. Multi-page (maxpage=N or limit_page=N)

3. Page range (parseFrom=N&parseTo=M)

4. Task-based (incremental, recommended for large trackers)

5. Full vs. incremental (fullparse=true/false)

Per-Tracker Parse Reference

Rutor

Megapeer

Anidub

Aniliberty

Animelayer

Anistar

Anifilm

Baibako

Bitru

BitruAPI

Kinozal

Knaben

Leproduction

Lostfilm

Mazepa

NNMClub

RuTracker

Korsars

Ultradox

Selezen

Toloka

TorrentBy

Search API

GET /api/v1.0/torrents

GET /api/v1.0/qualitys

GET /api/v2.0/indexers/[indexer]/results

GET /api/v1.0/conf

Stats API

GET /stats/refresh

GET /stats/torrents

GET /stats/trackers

GET /stats/trackers/{trackerName}

GET /stats/trackers/{trackerName}/new

GET /stats/trackers/{trackerName}/updated

Sync API

GET /sync/conf

V2 Protocol

GET /sync/fdb

GET /sync/fdb/torrents

V1 Protocol (Legacy)

GET /sync/torrents

Database Admin Endpoints

GET /jsondb/save

Dev/Maintenance Endpoints

Data Integrity

GET /dev/findcorrupt

GET /dev/removeNullValues

GET /dev/findDuplicateKeys

GET /dev/findEmptySearchFields

Data Updates

GET /dev/updateDetails

GET /dev/updateSearchName

GET /dev/updateSize

GET /dev/resetCheckTime

Tracker-Specific Fixes

1. Single-page (`page=N`)

2. Multi-page (`maxpage=N` or `limit_page=N`)

3. Page range (`parseFrom=N&parseTo=M`)

5. Full vs. incremental (`fullparse=true/false`)

`GET /api/v1.0/torrents`

`GET /api/v1.0/qualitys`

`GET /api/v2.0/indexers/[indexer]/results`

`GET /api/v1.0/conf`

`GET /stats/refresh`

`GET /stats/torrents`

`GET /stats/trackers`

`GET /stats/trackers/{trackerName}`

`GET /stats/trackers/{trackerName}/new`

`GET /stats/trackers/{trackerName}/updated`

`GET /sync/conf`

`GET /sync/fdb`

`GET /sync/fdb/torrents`

`GET /sync/torrents`

`GET /jsondb/save`

`GET /dev/findcorrupt`

`GET /dev/removeNullValues`

`GET /dev/findDuplicateKeys`

`GET /dev/findEmptySearchFields`

`GET /dev/updateDetails`

`GET /dev/updateSearchName`

`GET /dev/updateSize`

`GET /dev/resetCheckTime`

`GET /dev/fixKnabenNames`

`GET /dev/fixBitruNames`

`GET /dev/fixEmptySearchFields`

`GET /dev/migrateAnilibertyUrls`

`GET /dev/removeDuplicateAniliberty`

`GET /dev/fixAnimelayerDuplicates`

`GET /dev/fixKinozalUrls`

`GET /dev/fixSelezenUrls`

`GET /dev/removeBucket`

`GET /admin/config`

`POST /admin/config`

`GET /admin/cf-domains`

`DELETE /admin/cf-domains`

Packages