yu-zhi-ran/platform/backend/app/api/metrics.py

from fastapi import APIRouter, Depends, HTTPException, Query
from sqlalchemy.orm import Session
from sqlalchemy import func, desc
from typing import List, Optional
from datetime import datetime, timedelta, date
import logging
import re
import requests as http_requests
from pydantic import BaseModel

from ..database import get_db
from ..models import ContentMetrics, Topic, ContentCalendar
from ..schemas import (
    ContentMetricsCreate, ContentMetricsUpdate, ContentMetricsResponse,
    MetricsDashboard
)
from .auth import get_current_user, org_filter

router = APIRouter(prefix="/api/metrics", tags=["metrics"])


@router.get("/dashboard", response_model=MetricsDashboard)
def get_dashboard(
    days: int = Query(30, ge=1, le=365),
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    since = datetime.now() - timedelta(days=days)

    topic_base = db.query(Topic)
    of = org_filter(current_user, Topic)
    if of is not True:
        topic_base = topic_base.filter(of)

    total_topics = topic_base.count()

    raw_status = topic_base.with_entities(Topic.status, func.count()).group_by(Topic.status).all()
    topics_by_status = {}
    for s, cnt in raw_status:
        topics_by_status[s] = cnt

    metrics_q = db.query(ContentMetrics).join(Topic, ContentMetrics.topic_id == Topic.id)
    if of is not True:
        metrics_q = metrics_q.filter(of)
    total_published = metrics_q.filter(ContentMetrics.views > 0).count()

    all_metrics = metrics_q.filter(
        ContentMetrics.created_at >= since
    ).all()

    total_views = sum(m.views for m in all_metrics)
    total_likes = sum(m.likes for m in all_metrics)

    engagement_rates = [m.engagement_rate for m in all_metrics if m.views > 0]
    avg_engagement = sum(engagement_rates) / len(engagement_rates) if engagement_rates else 0

    top_q = db.query(
        ContentMetrics.topic_id,
        func.sum(ContentMetrics.views).label("total_views"),
        func.sum(ContentMetrics.likes).label("total_likes")
    ).join(Topic, ContentMetrics.topic_id == Topic.id)
    if of is not True:
        top_q = top_q.filter(of)
    top_topics_data = top_q.filter(
        ContentMetrics.created_at >= since
    ).group_by(ContentMetrics.topic_id).order_by(desc("total_views")).limit(10).all()

    top_topics = []
    for row in top_topics_data:
        topic = db.query(Topic).filter(Topic.id == row.topic_id).first()
        top_topics.append({
            "topic_id": row.topic_id,
            "title": topic.title if topic else row.topic_id,
            "total_views": row.total_views or 0,
            "total_likes": row.total_likes or 0,
        })

    recent_q = db.query(ContentMetrics).join(Topic, ContentMetrics.topic_id == Topic.id)
    if of is not True:
        recent_q = recent_q.filter(of)
    recent_metrics = recent_q.order_by(
        ContentMetrics.created_at.desc()
    ).limit(10).all()

    return MetricsDashboard(
        total_topics=total_topics,
        topics_by_status=topics_by_status,
        total_published=total_published,
        total_views=total_views,
        total_likes=total_likes,
        avg_engagement_rate=round(avg_engagement, 2),
        top_topics=top_topics,
        recent_metrics=[ContentMetricsResponse.model_validate(m) for m in recent_metrics]
    )


@router.get("/trend")
def get_trend(
    days: int = Query(30, ge=1, le=365),
    group_by: str = Query("day", enum=["day", "week", "month"]),
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    since = datetime.now() - timedelta(days=days)

    if group_by == "day":
        date_format = func.date(ContentMetrics.created_at)
    else:
        date_format = func.date_trunc(group_by, ContentMetrics.created_at)

    q = db.query(
        date_format.label("period"),
        func.sum(ContentMetrics.views).label("views"),
        func.sum(ContentMetrics.likes).label("likes"),
        func.sum(ContentMetrics.comments).label("comments"),
        func.count(ContentMetrics.id).label("count")
    ).join(Topic, ContentMetrics.topic_id == Topic.id)
    of_m = org_filter(current_user, Topic)
    if of_m is not True:
        q = q.filter(of_m)
    rows = q.filter(
        ContentMetrics.created_at >= since
    ).group_by(date_format).order_by(date_format).all()

    return [
        {
            "period": str(row.period),
            "views": row.views or 0,
            "likes": row.likes or 0,
            "comments": row.comments or 0,
            "count": row.count or 0
        }
        for row in rows
    ]


@router.get("/entries", response_model=List[ContentMetricsResponse])
def list_metrics(
    topic_id: Optional[str] = None,
    platform: Optional[str] = None,
    limit: int = Query(50, le=200),
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    query = db.query(ContentMetrics).join(Topic, ContentMetrics.topic_id == Topic.id)
    of_m = org_filter(current_user, Topic)
    if of_m is not True:
        query = query.filter(of_m)
    if topic_id:
        query = query.filter(ContentMetrics.topic_id == topic_id)
    if platform:
        query = query.filter(ContentMetrics.platform == platform)
    return query.order_by(ContentMetrics.created_at.desc()).limit(limit).all()


@router.post("/entries", response_model=ContentMetricsResponse)
def create_metric(
    data: ContentMetricsCreate,
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    topic = db.query(Topic).filter(Topic.id == data.topic_id).first()
    if not topic:
        raise HTTPException(status_code=404, detail="选题不存在")
    if current_user.role != "admin" and topic.org_id != current_user.org_id:
        raise HTTPException(status_code=404, detail="选题不存在")

    existing = db.query(ContentMetrics).filter(
        ContentMetrics.topic_id == data.topic_id,
        ContentMetrics.platform == data.platform
    ).first()

    if existing:
        for k, v in data.model_dump(exclude_unset=True).items():
            if k != "topic_id" and k != "platform":
                setattr(existing, k, v)
        existing.last_fetched = datetime.now()
        db.commit()
        db.refresh(existing)
        return existing

    metric = ContentMetrics(**data.model_dump(), last_fetched=datetime.now())
    db.add(metric)
    db.commit()
    db.refresh(metric)
    return metric


@router.put("/entries/{metric_id}", response_model=ContentMetricsResponse)
def update_metric(
    metric_id: int,
    data: ContentMetricsUpdate,
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    metric = db.query(ContentMetrics).filter(ContentMetrics.id == metric_id).first()
    if not metric:
        raise HTTPException(status_code=404, detail="数据记录不存在")

    for k, v in data.model_dump(exclude_unset=True).items():
        setattr(metric, k, v)
    metric.last_fetched = datetime.now()
    db.commit()
    db.refresh(metric)
    return metric


@router.delete("/entries/{metric_id}")
def delete_metric(
    metric_id: int,
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    metric = db.query(ContentMetrics).filter(ContentMetrics.id == metric_id).first()
    if not metric:
        raise HTTPException(status_code=404, detail="数据记录不存在")
    db.delete(metric)
    db.commit()
    return {"ok": True}


@router.get("/topics/{topic_id}", response_model=List[ContentMetricsResponse])
def get_topic_metrics(
    topic_id: str,
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    topic = db.query(Topic).filter(Topic.id == topic_id).first()
    if not topic:
        raise HTTPException(status_code=404, detail="选题不存在")
    if current_user.role != "admin" and topic.org_id != current_user.org_id:
        raise HTTPException(status_code=404, detail="选题不存在")
    return db.query(ContentMetrics).filter(
        ContentMetrics.topic_id == topic_id
    ).order_by(ContentMetrics.created_at.desc()).all()


@router.get("/by-platform")
def get_metrics_by_platform(
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    q = db.query(
        ContentMetrics.platform,
        func.sum(ContentMetrics.views).label("total_views"),
        func.sum(ContentMetrics.likes).label("total_likes"),
        func.sum(ContentMetrics.comments).label("total_comments"),
        func.count(ContentMetrics.id).label("count")
    ).join(Topic, ContentMetrics.topic_id == Topic.id)
    of_m = org_filter(current_user, Topic)
    if of_m is not True:
        q = q.filter(of_m)
    rows = q.group_by(ContentMetrics.platform).all()

    return [
        {
            "platform": row.platform,
            "total_views": row.total_views or 0,
            "total_likes": row.total_likes or 0,
            "total_comments": row.total_comments or 0,
            "count": row.count or 0,
            "avg_views": (row.total_views or 0) / (row.count or 1),
            "avg_likes": (row.total_likes or 0) / (row.count or 1),
        }
        for row in rows
    ]


@router.get("/recommend-topics")
def recommend_topics_from_metrics(
    limit: int = Query(10, le=50),
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    try:
        hp_q = db.query(
            ContentMetrics.topic_id,
            func.avg(ContentMetrics.engagement_rate).label("avg_engagement"),
            func.max(ContentMetrics.views).label("max_views")
        ).join(Topic, ContentMetrics.topic_id == Topic.id)
        of_m = org_filter(current_user, Topic)
        if of_m is not True:
            hp_q = hp_q.filter(of_m)
        high_performing = hp_q.group_by(ContentMetrics.topic_id).order_by(desc("avg_engagement")).limit(20).all()

        recommendations = []
        for row in high_performing:
            topic = db.query(Topic).filter(Topic.id == row.topic_id).first()
            if not topic:
                continue
            metrics = db.query(ContentMetrics).filter(
                ContentMetrics.topic_id == row.topic_id
            ).all()
            recommendations.append({
                "topic_id": row.topic_id,
                "title": topic.title,
                "field": topic.field_name,
                "status": topic.status,
                "avg_engagement": round(float(row.avg_engagement), 2) if row.avg_engagement else 0,
                "max_views": int(row.max_views) if row.max_views else 0,
                "platforms": list(set(m.platform for m in metrics)),
                "reason": f"平均互动率 {round(float(row.avg_engagement), 2) if row.avg_engagement else 0}%，最高阅读 {int(row.max_views) if row.max_views else 0}"
            })

        return recommendations[:limit]
    except Exception as e:
        logging.getLogger(__name__).exception(f"推荐选题失败: {e}")
        return []


_ZHIHU_UA = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"


def _extract_zhihu_post_id(url: str) -> str:
    m = re.search(r'zhuanlan\.zhihu\.com/p/(\d+)', url)
    if m:
        return m.group(1)
    m = re.search(r'zhihu\.com/question/\d+/answer/(\d+)', url)
    if m:
        return m.group(1)
    raise ValueError("无法从URL中提取知乎文章/回答ID")


class ZhihuFetchRequest(BaseModel):
    topic_id: str
    zhihu_url: str


@router.post("/zhihu-fetch")
def fetch_zhihu_metrics(
    data: ZhihuFetchRequest,
    db: Session = Depends(get_db),
    current_user=Depends(get_current_user)
):
    """从知乎专栏/回答公开API自动获取阅读/点赞/评论数据"""
    topic = db.query(Topic).filter(Topic.id == data.topic_id).first()
    if not topic:
        raise HTTPException(status_code=404, detail="选题不存在")

    post_id = _extract_zhihu_post_id(data.zhihu_url)
    api_url = f"https://zhuanlan.zhihu.com/api/posts/{post_id}"

    try:
        resp = http_requests.get(api_url, headers={"User-Agent": _ZHIHU_UA}, timeout=15)
        if resp.status_code == 404:
            api_url = f"https://www.zhihu.com/api/v4/answers/{post_id}"
            resp = http_requests.get(api_url, headers={"User-Agent": _ZHIHU_UA}, timeout=15)
        if resp.status_code != 200:
            raise HTTPException(status_code=502, detail=f"知乎API返回 {resp.status_code}")

        raw = resp.json()
        platform = "zhihu"

        existing = db.query(ContentMetrics).filter(
            ContentMetrics.topic_id == data.topic_id,
            ContentMetrics.platform == platform
        ).first()

        metric_data = {
            "views": raw.get("voteup_count", raw.get("views_count", 0)),
            "likes": raw.get("voteup_count", 0),
            "favorites": raw.get("favorite_count", 0),
            "comments": raw.get("comment_count", raw.get("comments_count", 0)),
            "shares": raw.get("share_count", 0) or raw.get("shared_count", 0),
            "publish_url": data.zhihu_url,
            "data_snapshot": raw,
        }

        if existing:
            for k, v in metric_data.items():
                setattr(existing, k, v)
            existing.last_fetched = datetime.now()
            db.commit()
            db.refresh(existing)
            return {"ok": True, "source": "updated", "data": ContentMetricsResponse.model_validate(existing)}
        else:
            metric = ContentMetrics(
                topic_id=data.topic_id,
                platform=platform,
                **metric_data,
                last_fetched=datetime.now()
            )
            db.add(metric)
            db.commit()
            db.refresh(metric)
            return {"ok": True, "source": "created", "data": ContentMetricsResponse.model_validate(metric)}

    except HTTPException:
        raise
    except Exception as e:
        raise HTTPException(status_code=502, detail=f"获取知乎数据失败: {str(e)}")