Initial Commit

backend/main.py

@@ -0,0 +1,173 @@
+import logging
+import os
+
+from apscheduler.schedulers.background import BackgroundScheduler
+from fastapi import Depends, FastAPI, Query
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.staticfiles import StaticFiles
+from fastapi.responses import FileResponse
+from sqlalchemy.orm import Session
+
+from backend import config
+from backend.database import get_db, init_db
+from backend.models import NewsItem
+from backend.news_service import scheduled_news_fetch
+from backend.repository import (
+    archive_old_news,
+    delete_archived_news,
+    get_latest_news,
+    get_news_paginated,
+    get_translation,
+    normalize_language,
+    resolve_news_content,
+)
+from backend.schemas import HealthResponse, NewsItemResponse, PaginatedNewsResponse
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(name)s: %(message)s",
+)
+logger = logging.getLogger(__name__)
+
+app = FastAPI(title="ClawFort News API", version="0.1.0")
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+static_dir = os.path.join(os.path.dirname(__file__), "static")
+app.mount("/static", StaticFiles(directory=static_dir), name="static")
+
+scheduler = BackgroundScheduler()
+
+
+def nightly_cleanup() -> None:
+    from backend.database import SessionLocal
+
+    db = SessionLocal()
+    try:
+        archived = archive_old_news(db, config.RETENTION_DAYS)
+        deleted = delete_archived_news(db, days_after_archive=60)
+        logger.info("Nightly cleanup: archived=%d, deleted=%d", archived, deleted)
+    finally:
+        db.close()
+
+
+@app.on_event("startup")
+async def startup_event() -> None:
+    if not config.PERPLEXITY_API_KEY:
+        logger.error("PERPLEXITY_API_KEY is not set — news fetching will fail")
+
+    os.makedirs("data", exist_ok=True)
+    init_db()
+    logger.info("Database initialized")
+
+    scheduler.add_job(scheduled_news_fetch, "interval", hours=1, id="news_fetch")
+    scheduler.add_job(nightly_cleanup, "cron", hour=3, minute=0, id="nightly_cleanup")
+    scheduler.start()
+    logger.info("Scheduler started: hourly news fetch + nightly cleanup")
+
+
+@app.on_event("shutdown")
+async def shutdown_event() -> None:
+    scheduler.shutdown(wait=False)
+    logger.info("Scheduler shut down")
+
+
+@app.get("/api/news", response_model=PaginatedNewsResponse)
+def api_get_news(
+    cursor: int | None = Query(None, description="Cursor for pagination (last item ID)"),
+    limit: int = Query(10, ge=1, le=50),
+    exclude_hero: int | None = Query(None, description="Hero item ID to exclude from feed"),
+    language: str = Query("en", description="Language code: en, ta, ml"),
+    db: Session = Depends(get_db),
+) -> PaginatedNewsResponse:
+    lang = normalize_language(language)
+    items = get_news_paginated(db, cursor=cursor, limit=limit + 1, exclude_id=exclude_hero)
+
+    has_more = len(items) > limit
+    if has_more:
+        items = items[:limit]
+
+    next_cursor = items[-1].id if items and has_more else None
+
+    response_items: list[NewsItemResponse] = []
+    for item in items:
+        translation = None
+        if lang != "en":
+            translation = get_translation(db, item.id, lang)
+        headline, summary = resolve_news_content(item, translation)
+        response_items.append(
+            NewsItemResponse(
+                id=item.id,
+                headline=headline,
+                summary=summary,
+                source_url=item.source_url,
+                image_url=item.image_url,
+                image_credit=item.image_credit,
+                published_at=item.published_at,
+                created_at=item.created_at,
+                language=lang if translation is not None else "en",
+            )
+        )
+
+    return PaginatedNewsResponse(
+        items=response_items,
+        next_cursor=next_cursor,
+        has_more=has_more,
+    )
+
+
+@app.get("/api/news/latest", response_model=NewsItemResponse | None)
+def api_get_latest_news(
+    language: str = Query("en", description="Language code: en, ta, ml"),
+    db: Session = Depends(get_db),
+) -> NewsItemResponse | None:
+    lang = normalize_language(language)
+    item = get_latest_news(db)
+    if not item:
+        return None
+
+    translation = None
+    if lang != "en":
+        translation = get_translation(db, item.id, lang)
+    headline, summary = resolve_news_content(item, translation)
+    return NewsItemResponse(
+        id=item.id,
+        headline=headline,
+        summary=summary,
+        source_url=item.source_url,
+        image_url=item.image_url,
+        image_credit=item.image_credit,
+        published_at=item.published_at,
+        created_at=item.created_at,
+        language=lang if translation is not None else "en",
+    )
+
+
+@app.get("/api/health", response_model=HealthResponse)
+def api_health(db: Session = Depends(get_db)) -> HealthResponse:
+    count = db.query(NewsItem).filter(NewsItem.archived.is_(False)).count()
+    return HealthResponse(status="ok", version="0.1.0", news_count=count)
+
+
+frontend_dir = os.path.join(os.path.dirname(os.path.dirname(__file__)), "frontend")
+
+
+@app.get("/")
+async def serve_frontend() -> FileResponse:
+    return FileResponse(os.path.join(frontend_dir, "index.html"))
+
+
+@app.get("/config")
+async def serve_config() -> dict:
+    return {
+        "umami_script_url": config.UMAMI_SCRIPT_URL,
+        "umami_website_id": config.UMAMI_WEBSITE_ID,
+        "supported_languages": config.SUPPORTED_LANGUAGES,
+        "default_language": "en",
+    }