myYouTube/backend/app/api/feed.py

import base64
from datetime import datetime, timezone

from fastapi import APIRouter, Depends, HTTPException, Query
from sqlalchemy import func, select
from sqlalchemy.orm import Session

from app.core.auth_dependency import require_session
from app.db import get_db
from app.models.channel import Channel
from app.models.channel_category import channel_categories
from app.models.download_job import DownloadJob
from app.models.video import Video
from app.services.download_jobs import latest_jobs_map
from app.services.video_presentation import channel_categories_map, serialize_video

router = APIRouter(dependencies=[Depends(require_session)])

DEFAULT_LIMIT = 30
MAX_LIMIT = 100


def _encode_cursor(published_at: datetime, video_id: int) -> str:
    raw = f"{published_at.isoformat()}|{video_id}"
    return base64.urlsafe_b64encode(raw.encode()).decode()


def _decode_cursor(cursor: str) -> tuple[datetime, int]:
    try:
        raw = base64.urlsafe_b64decode(cursor.encode()).decode()
        published_at_raw, video_id_raw = raw.rsplit("|", 1)
        published_at = datetime.fromisoformat(published_at_raw)
        if published_at.tzinfo is None:
            published_at = published_at.replace(tzinfo=timezone.utc)
        return published_at, int(video_id_raw)
    except Exception:
        raise HTTPException(status_code=400, detail="Invalid cursor")


@router.get("/feed")
def get_feed(
    category_id: int | None = None,
    uncategorized: bool = False,
    channel_id: int | None = None,
    downloaded: bool = False,
    limit: int = Query(DEFAULT_LIMIT, ge=1, le=MAX_LIMIT),
    cursor: str | None = None,
    db: Session = Depends(get_db),
) -> dict:
    query = db.query(Video)

    if channel_id is not None:
        query = query.filter(Video.channel_id == channel_id)
    elif uncategorized:
        categorized_channel_ids = select(channel_categories.c.channel_id)
        query = query.filter(~Video.channel_id.in_(categorized_channel_ids))
    elif category_id is not None:
        channel_ids_in_category = select(channel_categories.c.channel_id).where(
            channel_categories.c.category_id == category_id
        )
        query = query.filter(Video.channel_id.in_(channel_ids_in_category))

    if downloaded:
        # A video may have several download_jobs over time (retries,
        # redownload after delete); only its most recent one counts.
        latest_job_ids = select(func.max(DownloadJob.id)).group_by(DownloadJob.video_id)
        completed_video_ids = select(DownloadJob.video_id).where(
            DownloadJob.id.in_(latest_job_ids), DownloadJob.status == "completed"
        )
        query = query.filter(Video.id.in_(completed_video_ids))

    if cursor:
        cursor_published_at, cursor_id = _decode_cursor(cursor)
        query = query.filter(
            (Video.published_at < cursor_published_at)
            | ((Video.published_at == cursor_published_at) & (Video.id < cursor_id))
        )

    query = query.order_by(Video.published_at.desc(), Video.id.desc())
    rows = query.limit(limit + 1).all()

    next_cursor = None
    if len(rows) > limit:
        last_kept = rows[limit - 1]
        next_cursor = _encode_cursor(last_kept.published_at, last_kept.id)
        rows = rows[:limit]

    channel_ids = list({v.channel_id for v in rows})
    channels = {c.id: c for c in db.query(Channel).filter(Channel.id.in_(channel_ids)).all()}
    categories_map = channel_categories_map(db, channel_ids)
    jobs_map = latest_jobs_map(db, [v.id for v in rows])

    items = []
    for video in rows:
        channel = channels.get(video.channel_id)
        if channel is None:
            continue
        items.append(
            serialize_video(video, channel, categories_map.get(channel.id, []), jobs_map.get(video.id))
        )

    return {"items": items, "next_cursor": next_cursor}