siftlode/scripts/e2e_seed.py

"""Idempotent seed for the E2E characterization (golden-master) account.

Run INSIDE the api container — it needs the app's models + the argon2 hasher:

    docker exec -i siftlode-api-1 python - < scripts/e2e_seed.py

Resets two dedicated test accounts to a known baseline every run, so the Playwright
golden-master net is deterministic and never touches a real account: the admin the suite runs as,
and a plain member (`member.spec.ts`) that holds what a non-admin sees. Self-healing: a crashed
spec that left the account dirty is cleaned on the next run. Prints one JSON line describing
what it seeded (the wrapper / global-setup reads it).

Credentials come from env: E2E_EMAIL (default e2e@siftlode.local), E2E_MEMBER_EMAIL (default
e2e-member@siftlode.local) and E2E_PASSWORD (REQUIRED, no default — the ops wrapper exports it from
the secret store; both accounts share it); the accounts live only in the localdev DB.
"""

import json
import os
import sys
from datetime import datetime, timezone

sys.path.insert(0, "/app")

from sqlalchemy import delete, func, select

from app.db import SessionLocal
import app.tables  # noqa: F401 — every table, before the first query
from app.catalog.models import Channel, Subscription, Video
from app.models import AppState, User
from app.reader.models import Playlist, PlaylistItem, TakeoutItem, VideoState
from app.security import hash_password

EMAIL = os.environ.get("E2E_EMAIL", "e2e@siftlode.local")
MEMBER_EMAIL = os.environ.get("E2E_MEMBER_EMAIL", "e2e-member@siftlode.local")
PASSWORD = os.environ.get("E2E_PASSWORD", "")
if not PASSWORD:
    sys.exit(
        "e2e_seed: E2E_PASSWORD is not set (no default by design) — run it via `siftlode e2e-seed`"
    )
N_SUBS = 3


def main() -> None:
    db = SessionLocal()
    try:
        _mark_configured(db)
        # Admin so the E2E can reach the admin-only pages (scheduler / config / users / audit) and
        # thus the render ternary's admin branches — the primary real user (npeter83) is admin too.
        admin = _seed_account(db, EMAIL, "admin", "E2E Tester")
        member = _seed_account(db, MEMBER_EMAIL, "user", "E2E Member")
        # T201: this script's stdout IS its interface — the e2e runner parses this JSON.
        print(json.dumps({"ok": True, **admin, "member": member}))  # noqa: T201
    finally:
        db.close()


def _mark_configured(db) -> None:
    # 0. Mark the instance configured so the first-run setup wizard is skipped. A prod→dev
    #    reseed (siftlode seed-dev-from-prod) TRUNCATEs app_state — it's excluded from the
    #    data allow-list, being local/operational state — which would otherwise drop the box
    #    back into "Setup needed" and block login/UAT. Since e2e_seed always leaves a loginable
    #    admin below, the instance is legitimately set up. Idempotent (mirrors state._row +
    #    state.mark_configured). The running api process caches `configured` one-way from False,
    #    so it re-reads the DB and promotes on the next request — no restart needed.
    app_state = db.get(AppState, 1)
    if app_state is None:
        app_state = AppState(id=1, sync_paused=False)
        db.add(app_state)
    app_state.configured = True
    app_state.setup_token_hash = None
    db.commit()


def _seed_account(db, email: str, role: str, display_name: str) -> dict:
    # 1. Upsert the account and always reset it to a clean, loginable baseline
    #    (email_verified + is_active + not suspended + password set).
    user = db.scalar(select(User).where(User.email == email))
    if user is None:
        user = User(email=email)
        db.add(user)
    user.password_hash = hash_password(PASSWORD)
    user.email_verified = True
    user.is_active = True
    user.is_suspended = False
    user.google_sub = None
    user.role = role
    user.display_name = display_name
    user.preferences = {"language": "en"}
    db.commit()

    # 2. Subscriptions: the top-N organic channels by catalog size — deterministic, and
    #    >1 so a channel filter visibly narrows the feed. Reset each run.
    top = db.execute(
        select(Video.channel_id, func.count().label("n"))
        .join(Channel, Channel.id == Video.channel_id)
        .where(Channel.from_search.is_(False), Channel.from_explore.is_(False))
        .group_by(Video.channel_id)
        .order_by(func.count().desc())
        .limit(N_SUBS)
    ).all()
    channel_ids = [row[0] for row in top]

    db.execute(delete(Subscription).where(Subscription.user_id == user.id))
    for cid in channel_ids:
        db.add(
            Subscription(
                user_id=user.id,
                channel_id=cid,
                subscribed_at=datetime.now(timezone.utc),
            )
        )
    db.commit()

    # 3. Deterministic watch state (reset each run):
    #    - one in-progress video (drives the resume bar + "in progress" filter) from the
    #      first channel — status stays "new" so it still counts as unwatched;
    #    - one WATCHED video from the second channel, so the "watched" show-filter is a
    #      real, non-empty narrowing of the feed (and the "all" vs "unwatched" split exists).
    db.execute(delete(VideoState).where(VideoState.user_id == user.id))
    # A Takeout import's kept videos too: a settled row keeps its outcome over a re-upload, so the
    # import spec's counts would otherwise be the previous run's, not this one's.
    db.execute(delete(TakeoutItem).where(TakeoutItem.user_id == user.id))

    def _newest(channel_id: str, exclude: set[str]) -> Video | None:
        conds = [
            Video.channel_id == channel_id,
            Video.duration_seconds.is_not(None),
            Video.live_status == "none",
        ]
        if exclude:
            conds.append(Video.id.not_in(exclude))
        return db.scalar(select(Video).where(*conds).order_by(Video.published_at.desc()).limit(1))

    used: set[str] = set()
    in_progress = None
    watched = None
    if channel_ids:
        vid = _newest(channel_ids[0], used)
        if vid is not None:
            db.add(
                VideoState(
                    user_id=user.id,
                    video_id=vid.id,
                    status="new",
                    position_seconds=42,
                    progress_updated_at=datetime.now(timezone.utc),
                )
            )
            in_progress = vid.id
            used.add(vid.id)
    if len(channel_ids) > 1:
        wv = _newest(channel_ids[1], used)
        if wv is not None:
            db.add(
                VideoState(
                    user_id=user.id,
                    video_id=wv.id,
                    status="watched",
                    watched_at=datetime.now(timezone.utc),
                )
            )
            watched = wv.id
            used.add(wv.id)

    # 4. One "saved" video (Watch-later membership drives the saved marker) from the last
    #    subscribed channel. Reset the built-in playlist's items each run.
    wl = db.scalar(
        select(Playlist).where(Playlist.user_id == user.id, Playlist.kind == "watch_later")
    )
    if wl is None:
        wl = Playlist(user_id=user.id, name="Watch later", kind="watch_later")
        db.add(wl)
        db.commit()
    db.execute(delete(PlaylistItem).where(PlaylistItem.playlist_id == wl.id))
    saved = None
    if channel_ids:
        svid = _newest(channel_ids[-1], used)
        if svid is not None:
            db.add(PlaylistItem(playlist_id=wl.id, video_id=svid.id, position=0))
            saved = svid.id
    db.commit()

    # 5. Videos the Takeout import marks watched: from the subscribed channels, and already cleared by
    #    the Shorts probe — one not yet probed would wait for it, and the import spec would count
    #    nothing. One per browser project: they share this account, and a video imported once is not
    #    marked again.
    known = (
        list(
            db.scalars(
                select(Video.id)
                .where(
                    Video.channel_id.in_(channel_ids),
                    Video.shorts_probed.is_(True),
                    Video.is_short.is_(False),
                    Video.id.not_in(used | {saved} if saved else used),
                )
                .order_by(Video.published_at.desc())
                .limit(5)
            )
        )
        if channel_ids
        else []
    )

    return {
        "user_id": user.id,
        "email": email,
        "channels": channel_ids,
        "in_progress": in_progress,
        "watched": watched,
        "saved": saved,
        "takeout_known": known,
    }


if __name__ == "__main__":
    main()