"""Сквозной индекс контента: всё, что есть в сервисах, в одной базе. Зачем: искать по одному месту, а не по семи, и видеть «где что лежит». Индекс строится по расписанию и по команде; каждый источник опционален — если сервис недоступен, остальные всё равно попадают в базу. Выгрузки из баз делают маленькие скрипты-помощники (hub/helpers): они кладутся в контейнеры при развёртывании и печатают JSON. """ import datetime import json import os import sqlite3 from . import local, util from .sources import CFG, RADARR, SONARR, Jellyfin DB_PATH = os.environ.get("INDEX_DB", "/var/lib/hub-bots/index.db") HELPER_LOCAL = "/opt/hub-helpers" SCHEMA = """ create table if not exists items ( id integer primary key autoincrement, kind text not null, source text not null, title text not null, subtitle text, url text, external_id text, updated_at text, search_text text, unique(source, external_id) ); create index if not exists idx_items_title on items(title); create index if not exists idx_items_kind on items(kind); create table if not exists meta (key text primary key, value text); """ def _connect(): os.makedirs(os.path.dirname(DB_PATH), exist_ok=True) con = sqlite3.connect(DB_PATH, timeout=20) con.executescript(SCHEMA) # миграция для баз, созданных до появления поиска cols = {r[1] for r in con.execute("pragma table_info(items)")} if "search_text" not in cols: con.execute("alter table items add column search_text text") con.commit() return con def _norm(title, subtitle): """Приводим к нижнему регистру заранее: SQLite не умеет lower() для кириллицы.""" return f"{title or ''} {subtitle or ''}".strip().lower() def _year(*values): for v in values: if v: return str(v)[:4] return "" # ─────────────────────────── источники ─────────────────────────── def _jellyfin(): data, err = util.get( f"{CFG.jellyfin_url}/Items?Recursive=true&Limit=5000" f"&IncludeItemTypes=Movie,Series,Audio&Fields=ProductionYear&SortBy=SortName", {"Authorization": f'MediaBrowser Token="{CFG.jellyfin_key}"'}, ttl=600) if err: return [], err kinds = {"Movie": "фильм", "Series": "сериал", "Audio": "трек"} rows = [] for it in (data or {}).get("Items", []): rows.append({ "kind": kinds.get(it.get("Type"), "медиа"), "source": "jellyfin", "title": it.get("Name", ""), "subtitle": _year(it.get("ProductionYear")), "url": f"https://media.buryans.ru/web/index.html#!/details?id={it.get('Id')}", "external_id": str(it.get("Id")), }) return rows, None def _sonarr(): items, err = SONARR.series() if err: return [], err rows = [{ "kind": "сериал", "source": "sonarr", "title": s.get("title", ""), "subtitle": _year(s.get("year")), "url": "https://sonarr.buryans.ru/series", "external_id": str(s.get("id")), } for s in items or []] return rows, None def _radarr(): items, err = RADARR.movies() if err: return [], err rows = [{ "kind": "фильм", "source": "radarr", "title": m.get("title", ""), "subtitle": _year(m.get("year")), "url": "https://radarr.buryans.ru/movie", "external_id": str(m.get("id")), } for m in items or []] return rows, None def _navidrome(): out, err = local.pct_exec(125, ["docker", "exec", "navidrome", "sqlite3", "-separator", "|", "/data/navidrome.db", "select ifnull(album_artist,''), name, id from album limit 5000"]) if err: return [], err rows = [] for line in (out or "").splitlines(): parts = line.split("|") if len(parts) >= 3: rows.append({ "kind": "музыка", "source": "navidrome", "title": parts[1], "subtitle": parts[0], "url": "https://music.buryans.ru", "external_id": parts[2], }) return rows, None def _helper(ct, script, kind, source, make_row): out, err = local.pct_exec(ct, ["python3", f"{HELPER_LOCAL}/{script}"], timeout=40) if err: return [], err try: data = json.loads(out or "[]") except json.JSONDecodeError: return [], "не разобрал ответ помощника" rows = [] for rec in data: r = make_row(rec) if r and r.get("title"): r.update({"kind": kind, "source": source}) rows.append(r) return rows, None def _calibre(): return _helper(111, "calibre_dump.py", "книга", "calibre", lambda rec: { "title": rec[1], "subtitle": rec[2], "url": f"https://books.buryans.ru/book/{rec[0]}", "external_id": str(rec[0]), }) def _linkding(): return _helper(127, "linkding_dump.py", "закладка", "linkding", lambda rec: { "title": rec[1] or rec[2], "subtitle": (rec[3] or "")[:80], "url": rec[2], "external_id": str(rec[0]), }) def _paperless(): data, err = util.get( f"{CFG.paperless_url}/api/documents/?page_size=500&ordering=-created", {"Authorization": f"Token {CFG.paperless_token}"}, ttl=600) if err: return [], err rows = [{ "kind": "документ", "source": "paperless", "title": d.get("title", ""), "subtitle": (d.get("created") or "")[:10], "url": f"https://docs.buryans.ru/documents/{d.get('id')}", "external_id": str(d.get("id")), } for d in (data or {}).get("results", [])] return rows, None def _immich_albums(): """Альбомы Immich: сами снимки названий не имеют, а альбомы имеют.""" out, err = local.pct_exec(151, ["sh", f"{HELPER_LOCAL}/immich_albums.sh"], timeout=40) if err: return [], err rows = [] for line in (out or "").splitlines(): parts = line.split("|") if len(parts) >= 3: rows.append({ "kind": "фотоальбом", "source": "immich", "title": parts[1], "subtitle": parts[2], "url": "https://immich.buryans.ru/albums", "external_id": parts[0], }) return rows, None def _seerr_requests(): """Заявки: что просили, даже если ещё не скачано.""" data, err = util.get( f"{CFG.seerr_url}/api/v1/request?take=200&filter=all&sort=added", {"X-Api-Key": CFG.seerr_key}, ttl=300) if err: return [], err status = {1: "ожидает", 2: "одобрено", 3: "отклонено", 4: "частично"} rows = [] for r in (data or {}).get("results", []): media = r.get("media") or {} title = media.get("title") or media.get("name") or "" if not title: continue rows.append({ "kind": "заявка", "source": "seerr", "title": title, "subtitle": status.get(r.get("status"), "?"), "url": "https://ombi.buryans.ru", "external_id": str(r.get("id")), }) return rows, None SOURCES = { "jellyfin": _jellyfin, "sonarr": _sonarr, "radarr": _radarr, "navidrome": _navidrome, "calibre": _calibre, "linkding": _linkding, "paperless": _paperless, "immich": _immich_albums, "seerr": _seerr_requests, } # ─────────────────────────── работа с индексом ─────────────────────────── def rebuild(only=None): """Перестраиваем индекс. Возвращает отчёт по источникам.""" con = _connect() report = {} now = datetime.datetime.now().isoformat(timespec="seconds") for name, fn in SOURCES.items(): if only and name not in only: continue try: rows, err = fn() except Exception as e: rows, err = [], f"{type(e).__name__}: {e}" if err: report[name] = {"error": str(err)[:120], "count": 0} continue con.execute("delete from items where source = ?", (name,)) con.executemany( "insert or replace into items " "(kind, source, title, subtitle, url, external_id, updated_at, search_text) " "values (:kind, :source, :title, :subtitle, :url, :external_id, :updated_at, :search_text)", [{**r, "updated_at": now, "search_text": _norm(r["title"], r.get("subtitle"))} for r in rows]) report[name] = {"count": len(rows), "error": None} con.execute("insert or replace into meta (key, value) values ('built_at', ?)", (now,)) con.commit() con.close() return report def stats(): con = _connect() total = con.execute("select count(*) from items").fetchone()[0] by_kind = con.execute("select kind, count(*) from items group by kind order by 2 desc").fetchall() by_source = con.execute("select source, count(*) from items group by source order by 2 desc").fetchall() built = con.execute("select value from meta where key='built_at'").fetchone() con.close() return {"total": total, "kinds": by_kind, "sources": by_source, "built_at": built[0] if built else None, "path": DB_PATH} def search(term, limit=15): con = _connect() like = f"%{term.strip().lower()}%" rows = con.execute( "select kind, source, title, subtitle, url from items " "where search_text like ? order by kind, title limit ?", (like, limit)).fetchall() con.close() return [{"kind": k, "source": s, "title": t, "subtitle": sub, "url": u} for k, s, t, sub, u in rows] def links(term, limit=30): """Кросс-связи: одно и то же название в разных видах контента.""" con = _connect() like = f"%{term.strip().lower()}%" rows = con.execute( "select kind, title, subtitle, url from items where search_text like ? " "order by kind limit ?", (like, limit)).fetchall() con.close() grouped = {} for kind, title, subtitle, url in rows: grouped.setdefault(kind, []).append({"title": title, "subtitle": subtitle, "url": url}) return grouped