aza/AzA march 2026 - Kopie (18)/aza_news_backend.py

"""Aggregator for open-access medical news and congress events."""

from __future__ import annotations

import json
import re
import threading
import urllib.parse
import urllib.request
import xml.etree.ElementTree as ET
from dataclasses import asdict, dataclass, field
from datetime import date, datetime, timedelta, timezone
from email.utils import parsedate_to_datetime
from pathlib import Path
from typing import Any


NEWS_CACHE_TTL_SECONDS = 30 * 60
EVENTS_CACHE_TTL_SECONDS = 45 * 60
EVENT_URL_CHECK_TTL_SECONDS = 12 * 60 * 60


@dataclass(frozen=True)
class NewsFilter:
    specialties: list[str]
    regions: list[str]
    language: str = "de"
    sort: str = "newest"
    limit: int = 30


@dataclass(frozen=True)
class EventFilter:
    specialties: list[str]
    regions: list[str]
    from_date: date
    to_date: date
    sort: str = "soonest"
    limit: int = 100


@dataclass(frozen=True)
class NewsItem:
    id: str
    source: str
    title: str
    url: str
    publishedAt: str
    tags: list[str]
    languageOriginal: str
    isOpenAccess: bool
    evidenceType: str
    summaryOriginal: str
    summaryTranslated: str | None = None
    regions: list[str] = field(default_factory=list)


@dataclass(frozen=True)
class EventItem:
    id: str
    name: str
    startDate: str
    endDate: str
    city: str
    country: str
    regions: list[str]
    tags: list[str]
    url: str
    description: str
    type: str
    cmeFlag: bool = False
    organizer: str = ""
    source: str = ""
    icsUrl: str | None = None


@dataclass(frozen=True)
class FeedSource:
    name: str
    url: str
    regions: list[str]
    language: str
    default_tags: list[str]
    evidence_type: str


NEWS_SOURCES: tuple[FeedSource, ...] = (
    FeedSource("WHO", "https://www.who.int/feeds/entity/news-room/releases/en/rss.xml", ["WORLD"], "en", ["public-health"], "official"),
    FeedSource("ECDC", "https://www.ecdc.europa.eu/en/rss.xml", ["EU"], "en", ["public-health"], "official"),
    FeedSource("CDC", "https://tools.cdc.gov/api/v2/resources/media/132608.rss", ["WORLD"], "en", ["public-health"], "official"),
    FeedSource("EMA", "https://www.ema.europa.eu/en/news-events/news/rss.xml", ["EU"], "en", ["drug-safety"], "official"),
    FeedSource("Swissmedic", "https://www.swissmedic.ch/swissmedic/en/home/news/rss-feed/_jcr_content/contentPar/rssfeed.rss", ["CH"], "en", ["drug-safety"], "official"),
    FeedSource("Cochrane", "https://www.cochrane.org/news/rss.xml", ["WORLD"], "en", ["evidence-based-medicine"], "review"),
    FeedSource("medRxiv", "https://connect.medrxiv.org/relate/feed/medrxiv.xml", ["WORLD"], "en", ["preprint", "dermatology"], "preprint"),
    FeedSource("bioRxiv", "https://connect.biorxiv.org/relate/feed/biorxiv.xml", ["WORLD"], "en", ["preprint", "research"], "preprint"),
)

DEFAULT_SPECIALTY = "dermatology"
DEFAULT_NEWS_REGIONS = ["CH", "EU"]
DEFAULT_EVENT_REGIONS = ["CH", "EU"]

_news_cache_lock = threading.Lock()
_events_cache_lock = threading.Lock()
_event_url_cache_lock = threading.Lock()
_news_cache: dict[str, Any] = {"payload": [], "expires_at": 0.0}
_events_cache: dict[str, Any] = {"payload": [], "expires_at": 0.0, "seed_mtime": 0.0}
_event_url_status_cache: dict[str, tuple[bool, float]] = {}


def _now_ts() -> float:
    return datetime.now(timezone.utc).timestamp()


def _clean_text(text: str) -> str:
    cleaned = re.sub(r"<[^>]+>", " ", text or "")
    cleaned = re.sub(r"\s+", " ", cleaned).strip()
    return cleaned


def _safe_summary(text: str) -> str:
    s = _clean_text(text)
    if len(s) > 520:
        return s[:517].rstrip() + "..."
    return s


def _parse_datetime(raw: str) -> datetime:
    if not raw:
        return datetime.now(timezone.utc)
    try:
        dt = parsedate_to_datetime(raw)
        return dt if dt.tzinfo else dt.replace(tzinfo=timezone.utc)
    except Exception:
        pass
    try:
        dt2 = datetime.fromisoformat(raw.replace("Z", "+00:00"))
        return dt2 if dt2.tzinfo else dt2.replace(tzinfo=timezone.utc)
    except Exception:
        return datetime.now(timezone.utc)


def _read_url(url: str, timeout: int = 12) -> bytes:
    req = urllib.request.Request(url, headers={"User-Agent": "AZA-News-Aggregator/1.0"})
    with urllib.request.urlopen(req, timeout=timeout) as resp:
        return resp.read()


def _is_live_event_url(url: str) -> bool:
    u = (url or "").strip()
    if not u:
        return False
    now = _now_ts()
    with _event_url_cache_lock:
        cached = _event_url_status_cache.get(u)
        if cached and cached[1] > now:
            return bool(cached[0])
    ok = False
    try:
        req = urllib.request.Request(
            u,
            headers={"User-Agent": "AZA-News-Aggregator/1.0"},
            method="HEAD",
        )
        with urllib.request.urlopen(req, timeout=10) as resp:
            ok = int(getattr(resp, "status", 200)) < 400
    except Exception:
        try:
            req = urllib.request.Request(
                u,
                headers={"User-Agent": "AZA-News-Aggregator/1.0"},
                method="GET",
            )
            with urllib.request.urlopen(req, timeout=12) as resp:
                ok = int(getattr(resp, "status", 200)) < 400
        except Exception:
            ok = False
    with _event_url_cache_lock:
        _event_url_status_cache[u] = (ok, now + EVENT_URL_CHECK_TTL_SECONDS)
    return ok


def _rss_items(source: FeedSource, limit: int = 25) -> list[NewsItem]:
    try:
        payload = _read_url(source.url)
        root = ET.fromstring(payload)
    except Exception:
        return []

    channel = root.find("channel")
    if channel is not None:
        entries = channel.findall("item")
    else:
        entries = root.findall(".//{http://www.w3.org/2005/Atom}entry")

    items: list[NewsItem] = []
    for idx, node in enumerate(entries):
        if idx >= limit:
            break
        title = _clean_text((node.findtext("title") or node.findtext("{http://www.w3.org/2005/Atom}title") or "Ohne Titel"))
        link = _clean_text(node.findtext("link") or "")
        if not link:
            atom_link = node.find("{http://www.w3.org/2005/Atom}link")
            if atom_link is not None:
                link = _clean_text(atom_link.attrib.get("href") or source.url)
        summary = node.findtext("description") or node.findtext("{http://www.w3.org/2005/Atom}summary") or ""
        pub_raw = node.findtext("pubDate") or node.findtext("{http://www.w3.org/2005/Atom}updated") or ""
        published_at = _parse_datetime(pub_raw).isoformat()
        items.append(
            NewsItem(
                id=f"{source.name.lower()}-{abs(hash((title, link, published_at)))}",
                source=source.name,
                title=title or "Ohne Titel",
                url=link or source.url,
                publishedAt=published_at,
                tags=list(source.default_tags),
                languageOriginal=source.language,
                isOpenAccess=True,
                evidenceType=source.evidence_type,
                summaryOriginal=_safe_summary(summary) or "Kurz-Zusammenfassung in der Quelle nicht verfügbar.",
                regions=list(source.regions),
            )
        )
    return items


def _pubmed_open_access_news(limit: int = 12) -> list[NewsItem]:
    term = '(dermatology[Title/Abstract]) AND ("open access"[Filter])'
    query = urllib.parse.urlencode(
        {
            "db": "pubmed",
            "retmode": "json",
            "retmax": str(limit),
            "sort": "pub+date",
            "term": term,
        }
    )
    search_url = f"https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?{query}"
    try:
        ids_payload = json.loads(_read_url(search_url).decode("utf-8", errors="ignore"))
        ids = ids_payload.get("esearchresult", {}).get("idlist", [])
    except Exception:
        return []
    if not ids:
        return []

    summary_url = (
        "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?"
        + urllib.parse.urlencode({"db": "pubmed", "retmode": "json", "id": ",".join(ids)})
    )
    try:
        summary_payload = json.loads(_read_url(summary_url).decode("utf-8", errors="ignore"))
    except Exception:
        return []

    out: list[NewsItem] = []
    for pmid in ids:
        rec = summary_payload.get("result", {}).get(pmid) or {}
        title = _clean_text(str(rec.get("title") or ""))
        if not title:
            continue
        pubdate = _clean_text(str(rec.get("pubdate") or ""))
        dt = _parse_datetime(pubdate)
        url = f"https://pubmed.ncbi.nlm.nih.gov/{pmid}/"
        out.append(
            NewsItem(
                id=f"pubmed-{pmid}",
                source="PubMed",
                title=title,
                url=url,
                publishedAt=dt.isoformat(),
                tags=["dermatology", "open-access"],
                languageOriginal="en",
                isOpenAccess=True,
                evidenceType="peer-reviewed",
                summaryOriginal="Open-Access Eintrag aus PubMed. Volltext/Abstract je nach Journal frei verfügbar.",
                regions=["WORLD"],
            )
        )
    return out


def _normalize_tokens(values: list[str] | None, default: list[str]) -> list[str]:
    out = [str(v).strip().lower() for v in (values or []) if str(v).strip()]
    return out or list(default)


def _match_regions(item_regions: list[str], selected_regions: list[str]) -> bool:
    if not selected_regions:
        return True
    selected = {r.lower() for r in selected_regions}
    if "world" in selected or "worldwide" in selected:
        return True
    item_norm = {r.lower() for r in item_regions}
    # Strikter Regionsabgleich: EU zeigt nur EU/CH-Events, nicht WORLD/US/CA.
    if "eu" in selected and ("eu" in item_norm or "ch" in item_norm):
        return True
    if "ch" in selected and "ch" in item_norm:
        return True
    return bool(item_norm.intersection(selected))


def _match_specialties(item_tags: list[str], selected_specialties: list[str]) -> bool:
    if "all" in selected_specialties:
        return True
    tags = {t.lower() for t in item_tags}
    selected = set(selected_specialties)
    if "dermatology" in selected:
        selected.update({"skin"})
    if "infectiology" in selected:
        selected.update({"public-health", "infectiology"})
    if "general-medicine" in selected:
        selected.update({"public-health", "evidence-based-medicine", "internal-medicine"})
    if "internal-medicine" in selected:
        selected.update({"general-medicine", "internal-medicine"})
    return bool(tags.intersection(selected))


def _translate_summary_stub(summary: str, target_language: str, source_language: str) -> str | None:
    target = (target_language or "").strip().lower()
    source = (source_language or "").strip().lower()
    if not target or target in {"system", "auto", source}:
        return None
    return f"[Übersetzung nicht konfiguriert: {source}->{target}] {summary}"


def get_news(filters: NewsFilter) -> list[NewsItem]:
    now = _now_ts()
    with _news_cache_lock:
        if float(_news_cache["expires_at"]) > now:
            all_items = list(_news_cache["payload"])
        else:
            fetched: list[NewsItem] = []
            for src in NEWS_SOURCES:
                fetched.extend(_rss_items(src, limit=24))
            fetched.extend(_pubmed_open_access_news(limit=16))
            fetched.sort(key=lambda x: x.publishedAt, reverse=True)
            _news_cache["payload"] = fetched
            _news_cache["expires_at"] = now + NEWS_CACHE_TTL_SECONDS
            all_items = fetched

    specialties = _normalize_tokens(filters.specialties, [DEFAULT_SPECIALTY])
    regions = _normalize_tokens(filters.regions, DEFAULT_NEWS_REGIONS)
    filtered = [item for item in all_items if _match_specialties(item.tags, specialties) and _match_regions(item.regions, regions)]
    if filters.sort == "oldest":
        filtered.sort(key=lambda x: x.publishedAt)
    else:
        filtered.sort(key=lambda x: x.publishedAt, reverse=True)

    out: list[NewsItem] = []
    for item in filtered[: max(1, min(filters.limit, 120))]:
        out.append(
            NewsItem(
                id=item.id,
                source=item.source,
                title=item.title,
                url=item.url,
                publishedAt=item.publishedAt,
                tags=item.tags,
                languageOriginal=item.languageOriginal,
                isOpenAccess=item.isOpenAccess,
                evidenceType=item.evidenceType,
                summaryOriginal=item.summaryOriginal,
                summaryTranslated=_translate_summary_stub(item.summaryOriginal, filters.language, item.languageOriginal),
                regions=item.regions,
            )
        )
    return out


def _seed_events_path() -> Path:
    return Path(__file__).resolve().parent / "news_events_seed.json"


def _load_seed_events() -> list[EventItem]:
    try:
        with open(_seed_events_path(), "r", encoding="utf-8") as f:
            payload = json.load(f)
    except Exception:
        return []
    rows = payload.get("events") if isinstance(payload, dict) else None
    if not isinstance(rows, list):
        return []
    out: list[EventItem] = []
    for row in rows:
        if not isinstance(row, dict):
            continue
        try:
            out.append(
                EventItem(
                    id=str(row["id"]),
                    name=str(row["name"]),
                    startDate=str(row["startDate"]),
                    endDate=str(row["endDate"]),
                    city=str(row.get("city") or ""),
                    country=str(row.get("country") or ""),
                    regions=[str(r).upper() for r in row.get("regions", []) if str(r).strip()],
                    tags=[str(t).lower() for t in row.get("tags", []) if str(t).strip()],
                    url=str(row.get("url") or ""),
                    description=str(row.get("description") or ""),
                    type=str(row.get("type") or "kongress"),
                    cmeFlag=bool(row.get("cmeFlag", False)),
                    organizer=str(row.get("organizer") or ""),
                    source=str(row.get("source") or ""),
                    icsUrl=(str(row.get("icsUrl")).strip() if row.get("icsUrl") else None),
                )
            )
        except Exception:
            continue
    return out


def get_events(filters: EventFilter) -> list[EventItem]:
    now = _now_ts()
    try:
        seed_mtime = _seed_events_path().stat().st_mtime
    except Exception:
        seed_mtime = 0.0
    with _events_cache_lock:
        cache_mtime = float(_events_cache.get("seed_mtime", 0.0))
        if float(_events_cache["expires_at"]) > now and cache_mtime == seed_mtime:
            source_items = list(_events_cache["payload"])
        else:
            source_items = _load_seed_events()
            _events_cache["payload"] = source_items
            _events_cache["expires_at"] = now + EVENTS_CACHE_TTL_SECONDS
            _events_cache["seed_mtime"] = seed_mtime

    specialties = _normalize_tokens(filters.specialties, [DEFAULT_SPECIALTY])
    regions = _normalize_tokens(filters.regions, DEFAULT_EVENT_REGIONS)
    out: list[EventItem] = []
    for item in source_items:
        try:
            start = date.fromisoformat(item.startDate)
            end = date.fromisoformat(item.endDate)
        except Exception:
            continue
        if end < filters.from_date or start > filters.to_date:
            continue
        if not _match_specialties(item.tags, specialties):
            continue
        if not _match_regions(item.regions, regions):
            continue
        if not _is_live_event_url(item.url):
            continue
        out.append(item)

    if filters.sort == "latest":
        out.sort(key=lambda x: x.startDate, reverse=True)
    else:
        out.sort(key=lambda x: x.startDate)
    return out[: max(1, min(filters.limit, 300))]


# Backward-compatible wrappers used by backend_main.py
def get_news_items(specialties: list[str] | None, lang: str = "de", region: str = "CH", limit: int = 30) -> list[dict[str, Any]]:
    region_values = [r.strip() for r in str(region or "CH").split(",") if r.strip()]
    rows = get_news(NewsFilter(specialties=specialties or [DEFAULT_SPECIALTY], regions=region_values, language=lang, limit=limit))
    return [asdict(x) for x in rows]


def get_event_items(
    specialties: list[str] | None,
    regions: list[str] | None,
    from_date: date | None,
    to_date: date | None,
    limit: int = 100,
) -> list[dict[str, Any]]:
    rows = get_events(
        EventFilter(
            specialties=specialties or [DEFAULT_SPECIALTY],
            regions=regions or list(DEFAULT_EVENT_REGIONS),
            from_date=from_date or date.today(),
            to_date=to_date or (date.today() + timedelta(days=396)),
            limit=limit,
        )
    )
    return [asdict(x) for x in rows]