Files
AlphaPulse/backend/src/alphapulse/api/risk.py
T
eezd 4f43ca8b57 Add AI news risk analysis and dashboard
Analyze news revisions with independent model settings, evidence-backed risk ratings, event grouping, and rolling market outlooks. Add a default homepage, archival history, pending-result refresh, and isolated regression coverage.

Keep analysis disabled until explicitly enabled. Require a separately authorized schema migration before deployment and preserve compatibility with the previous six-service instance.
2026-10-30 18:35:49 +08:00

147 lines
8.1 KiB
Python

"""Authenticated news risk research endpoints."""
from datetime import UTC, datetime, timedelta
from typing import Annotated, Literal
from fastapi import APIRouter, Depends, HTTPException, Query
from pydantic import BaseModel, Field
from sqlalchemy import select, func
from sqlalchemy.ext.asyncio import AsyncSession
from alphapulse.api.response import success
from alphapulse.core.auth import require_admin
from alphapulse.core.db import get_session
from alphapulse.core.redis import redis
from alphapulse.models import NewsItem, NewsRevision
from alphapulse.models.risk import RiskSettings, RiskAnalysis, RiskEvent, RiskSnapshot
from alphapulse.news_risk.service import START, default_settings, provider_for, complete, Labels, SYSTEM
router = APIRouter(prefix='/api/risk', tags=['risk'], dependencies=[Depends(require_admin)])
DB = Annotated[AsyncSession, Depends(get_session)]
class SettingsIn(BaseModel):
enabled: bool = False
provider_id: int = Field(gt=0)
model: str = Field(min_length=1, max_length=128, pattern=r'.*\S.*')
@router.get('/settings')
async def settings(session: DB):
value = await default_settings(session)
return success({'enabled': value.enabled, 'provider_id': value.provider_id, 'model': value.model, 'start_at': START})
@router.put('/settings')
async def save_settings(body: SettingsIn, session: DB):
settings = RiskSettings(id=1, **body.model_dump())
try:
await provider_for(session, settings)
except ValueError as error:
raise HTTPException(400, str(error)) from error
if body.enabled:
previous = await session.get(RiskSettings, 1)
if not previous or not previous.enabled or previous.provider_id != body.provider_id or previous.model != body.model:
raise HTTPException(409, '请先使用测试并启用按钮验证模型')
await session.merge(settings); await session.commit()
return success(None)
@router.post('/enable')
async def enable(body: SettingsIn, session: DB):
settings = RiskSettings(id=1, **{**body.model_dump(), 'enabled': True})
try:
provider = await provider_for(session, settings)
except ValueError as error:
raise HTTPException(400, str(error)) from error
fixture = {'news': [{'id': 1, 'title': '测试消息:暂无新事件', 'summary': '', 'content': ''}], 'candidates': []}
from alphapulse.news_risk.service import validate_labels
try:
result, _ = await complete(provider, SYSTEM, fixture, Labels)
validate_labels(result, {1: fixture['news'][0]}, set())
except Exception as error:
raise HTTPException(502, f'模型验证失败:{error}') from error
await session.merge(settings); await session.commit()
return success(None)
async def coverage(session):
total = await session.scalar(select(func.count()).select_from(NewsItem).where(NewsItem.published_at.between(START, datetime.now(UTC))))
counts = dict((await session.execute(select(RiskAnalysis.state, func.count()).join(NewsItem).where(
RiskAnalysis.content_hash == NewsItem.content_hash,
NewsItem.published_at.between(START, datetime.now(UTC))).group_by(RiskAnalysis.state))).all())
return {'total': total, 'done': counts.get('done', 0), 'failed': counts.get('failed', 0),
'processing': counts.get('processing', 0), 'pending': max(0, total - sum(counts.values())) + counts.get('pending', 0)}
@router.get('/overview')
async def overview(session: DB, hours: Literal[24, 48, 168] = 24):
now = datetime.now(UTC); cutoff = now - timedelta(hours=hours)
counts = dict((await session.execute(select(RiskEvent.stars, func.count()).where(
RiskEvent.updated_at >= cutoff, RiskEvent.result["verification"].astext != "insufficient").group_by(RiskEvent.stars))).all())
snapshot = await session.scalar(select(RiskSnapshot).where(RiskSnapshot.state == 'done').order_by(RiskSnapshot.id.desc()).limit(1))
last = await session.scalar(select(RiskSnapshot).order_by(RiskSnapshot.id.desc()).limit(1))
scenario = next((w for w in snapshot.result['windows'] if w['hours'] == hours), None) if snapshot else None
return success({'as_of': now, 'counts': counts, 'coverage': await coverage(session), 'scenario': scenario,
'snapshot_at': snapshot.created_at if snapshot else None, 'quotes': snapshot.quotes if snapshot else None,
'stale': not snapshot or (now-snapshot.created_at).total_seconds() > 1200 or bool(last.error),
'error': last.error if last else None, 'heartbeat': await redis.get('ap:risk:heartbeat')})
@router.get('/events')
async def events(session: DB, hours: Literal[24, 48, 168] | None = None, stars: int = Query(1, ge=0, le=5),
q: str = '', start: datetime | None = None, end: datetime | None = None,
page: int = Query(1, ge=1), limit: int = Query(20, ge=1, le=100)):
conditions = [RiskEvent.stars >= stars]
if hours: conditions.append(RiskEvent.updated_at >= datetime.now(UTC)-timedelta(hours=hours))
if start: conditions.append(RiskEvent.updated_at >= start)
if end: conditions.append(RiskEvent.updated_at < end)
if q: conditions.append(RiskEvent.title.contains(q, autoescape=True))
rows = (await session.scalars(select(RiskEvent).where(*conditions).order_by(
RiskEvent.stars.desc(), RiskEvent.updated_at.desc(), RiskEvent.id.desc()).offset((page-1)*limit).limit(limit))).all()
return success({'items': [event_out(e) for e in rows], 'total': await session.scalar(select(func.count()).select_from(RiskEvent).where(*conditions))})
def event_out(event):
return {key: getattr(event, key) for key in ('id', 'title', 'category', 'entities', 'stars', 'peak_stars', 'result', 'published_at', 'updated_at')}
@router.get('/events/{event_id}')
async def event_detail(event_id: int, session: DB):
event = await session.get(RiskEvent, event_id)
if event is None: raise HTTPException(404, '风险事件不存在')
rows = (await session.scalars(select(RiskAnalysis).where(RiskAnalysis.event_id == event_id).order_by(RiskAnalysis.id))).all()
return success({'event': event_out(event), 'history': [analysis_out(r) for r in rows]})
def analysis_out(row):
return {'id': row.id, 'news_id': row.news_id, 'state': row.state, 'result': row.result, 'input': row.input,
'model': row.model, 'error': row.error, 'analyzed_at': row.finished_at,
'historical': row.started_at is not None and (row.started_at-datetime.fromisoformat(row.input['progress_at'])).total_seconds() > 300}
@router.get('/analyses')
async def analyses(session: DB, news_id: int | None = None, revision_no: int | None = Query(None, ge=1), failed: bool = False):
query = select(RiskAnalysis)
if news_id: query = query.where(RiskAnalysis.news_id == news_id)
if revision_no is not None:
if not news_id: raise HTTPException(400, '指定修订版本时必须提供新闻 ID')
current = await session.get(NewsItem, news_id)
if current is None: raise HTTPException(404, '新闻不存在')
content_hash = current.content_hash if current.revision_no == revision_no else await session.scalar(
select(NewsRevision.content_hash).where(NewsRevision.news_id == news_id, NewsRevision.revision_no == revision_no))
if content_hash is None: raise HTTPException(404, '新闻修订版本不存在')
query = query.where(RiskAnalysis.content_hash == content_hash)
if failed: query = query.where(RiskAnalysis.state == 'failed')
return success([analysis_out(r) for r in (await session.scalars(query.order_by(RiskAnalysis.id.desc()).limit(100))).all()])
@router.post('/analyses/{analysis_id}/retry', status_code=202)
async def retry(analysis_id: int, session: DB):
row = await session.get(RiskAnalysis, analysis_id, with_for_update=True)
if row is None: raise HTTPException(404, '分析记录不存在')
if row.state != 'failed': raise HTTPException(409, '仅可重试失败任务')
current = await session.get(NewsItem, row.news_id)
if current.content_hash != row.content_hash:
raise HTTPException(409, '新闻已修订,请等待最新版本分析;旧记录保留')
row.state = 'pending'; row.error = None
await session.commit()
return success(None)