配置全面迁移数据库：PromptConfig、TaskConfig动态调度、敏感词/清洗规则/趋势映射/平台标签/痛点模板全部可编辑
- 新增 PromptConfig 模型 + API，支持提示词在线编辑（16条默认） - 调度器动态读取 TaskConfig.schedule，admin 可调执行时间 - 新增 KeywordDomainMap、SensitiveWord、ContentCleanRule、TrendFieldMapping 表 - DOMAINS、TREND_DOMAIN_MAP、PLATFORM_TAGS、china_pains、RSS关键词、priority_weights 全部迁移到 DB - tasks.html 重构：卡片网格+配置/产出/历史/提示词四个Tab，折叠显示 - 清理冗余代码：DEFAULT_PROMPTS死代码、collector.py unreachable代码、compliance_checker bug - strip_thinking_html 改用 DB 规则优先
This commit is contained in:
Yuzhiran Dev
@@ -0,0 +1,108 @@
|
||||
import os
|
||||
from typing import Optional, Dict, Any
|
||||
|
||||
USE_POSTGRES = os.getenv('USE_POSTGRES', 'true').lower() == 'true'
|
||||
if USE_POSTGRES:
|
||||
from sqlalchemy import create_engine
|
||||
from sqlalchemy.orm import sessionmaker
|
||||
POSTGRES_CONFIG = {
|
||||
'host': os.getenv('PG_HOST', '127.0.0.1'),
|
||||
'port': os.getenv('PG_PORT', '5432'),
|
||||
'database': os.getenv('PG_DATABASE', 'yzr_nr'),
|
||||
'user': os.getenv('PG_USER', 'yzr_nr'),
|
||||
'password': os.getenv('PG_PASSWORD', 'aTX3WKKnPfRnM5PC')
|
||||
}
|
||||
SQLALCHEMY_DATABASE_URL = (
|
||||
f"postgresql://{POSTGRES_CONFIG['user']}:{POSTGRES_CONFIG['password']}"
|
||||
f"@{POSTGRES_CONFIG['host']}:{POSTGRES_CONFIG['port']}/{POSTGRES_CONFIG['database']}"
|
||||
)
|
||||
_engine = create_engine(SQLALCHEMY_DATABASE_URL, pool_pre_ping=True)
|
||||
else:
|
||||
from pathlib import Path
|
||||
PROJECT_ROOT = Path(__file__).resolve().parent.parent.parent.parent
|
||||
DATA_DIR = os.getenv('DATA_DIR', str(PROJECT_ROOT / 'data'))
|
||||
os.makedirs(DATA_DIR, exist_ok=True)
|
||||
DB_PATH = os.path.join(DATA_DIR, 'yzr.db')
|
||||
from sqlalchemy import create_engine
|
||||
_engine = create_engine(f"sqlite:///{DB_PATH}", connect_args={"check_same_thread": False})
|
||||
|
||||
_SessionLocal = sessionmaker(autocommit=False, autoflush=False, bind=_engine)
|
||||
|
||||
|
||||
def _get_session():
|
||||
session = _SessionLocal()
|
||||
try:
|
||||
return session
|
||||
except:
|
||||
session.close()
|
||||
raise
|
||||
|
||||
|
||||
_PROMPT_CACHE: Dict[str, Dict[str, Any]] = {}
|
||||
_CACHE_VERSION = "v2"
|
||||
|
||||
|
||||
def load_prompt_config(key: str, module_id: Optional[str] = None) -> Optional[Dict[str, Any]]:
|
||||
if key in _PROMPT_CACHE and _PROMPT_CACHE[key].get("_v") == _CACHE_VERSION:
|
||||
return _PROMPT_CACHE[key]
|
||||
try:
|
||||
session = _get_session()
|
||||
try:
|
||||
from .models import PromptConfig
|
||||
prompt = session.query(PromptConfig).filter(
|
||||
PromptConfig.key == key,
|
||||
PromptConfig.enabled == True
|
||||
).first()
|
||||
if not prompt and module_id:
|
||||
prompt = session.query(PromptConfig).filter(
|
||||
PromptConfig.key == key,
|
||||
PromptConfig.module_id.in_([module_id, "all"]),
|
||||
PromptConfig.enabled == True
|
||||
).first()
|
||||
if prompt:
|
||||
result = {
|
||||
"_v": _CACHE_VERSION,
|
||||
"content": prompt.content,
|
||||
"temperature": prompt.temperature,
|
||||
"max_tokens": prompt.max_tokens,
|
||||
"variables": prompt.variables or [],
|
||||
"description": prompt.description,
|
||||
"category": prompt.category,
|
||||
}
|
||||
_PROMPT_CACHE[key] = result
|
||||
return result
|
||||
finally:
|
||||
session.close()
|
||||
except Exception as e:
|
||||
import warnings
|
||||
warnings.warn(f"load_prompt_config({key}) failed: {e}")
|
||||
return None
|
||||
|
||||
|
||||
def clear_prompt_cache():
|
||||
_PROMPT_CACHE.clear()
|
||||
|
||||
|
||||
def get_prompt(key: str, module_id: Optional[str] = None, **kwargs) -> str:
|
||||
cfg = load_prompt_config(key, module_id)
|
||||
if cfg:
|
||||
content = cfg["content"]
|
||||
for var in cfg.get("variables", []):
|
||||
name = var.get("name")
|
||||
if name and name in kwargs:
|
||||
content = content.replace("{" + name + "}", str(kwargs[name]))
|
||||
elif name:
|
||||
default = var.get("default_value", "")
|
||||
content = content.replace("{" + name + "}", str(default))
|
||||
return content
|
||||
return ""
|
||||
|
||||
|
||||
def get_llm_params(key: str, module_id: Optional[str] = None) -> Dict[str, Any]:
|
||||
cfg = load_prompt_config(key, module_id)
|
||||
if cfg:
|
||||
return {
|
||||
"temperature": cfg.get("temperature"),
|
||||
"max_tokens": cfg.get("max_tokens"),
|
||||
}
|
||||
return {}
|
||||
@@ -2,11 +2,15 @@
|
||||
定时任务调度器
|
||||
基于 APScheduler，支持在 FastAPI 生命周期内运行定时任务
|
||||
"""
|
||||
import os
|
||||
import sys
|
||||
import logging
|
||||
from datetime import datetime
|
||||
import os, sys, logging, json
|
||||
from pathlib import Path
|
||||
from datetime import datetime, timezone
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent.parent.parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT / 'scripts'))
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
from prompt_loader import get_prompt, get_prompt_params
|
||||
|
||||
from apscheduler.schedulers.background import BackgroundScheduler
|
||||
from apscheduler.triggers.cron import CronTrigger
|
||||
from .generator import run_creator_blocking
|
||||
@@ -15,6 +19,69 @@ from .collector import run_collector_blocking
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
MODULES = {
|
||||
"scheduled_refresh_search_cache": {"name": "🔍 搜索缓存", "cron": "01:00"},
|
||||
"scheduled_fetch_trends": {"name": "🔥 热点趋势", "cron": "01:10"},
|
||||
"scheduled_collect": {"name": "📡 内容采集", "cron": "01:30"},
|
||||
"scheduled_generate": {"name": "🤖 内容创作", "cron": "02:00"},
|
||||
"scheduled_optimize": {"name": "🔍 合规审查", "cron": "03:00"},
|
||||
"scheduled_optimize_sources": {"name": "📡 信息源优化", "cron": "05:00"},
|
||||
"scheduled_metrics_sync": {"name": "📊 指标同步", "cron": "06:00"},
|
||||
}
|
||||
|
||||
def _log_task(module_id: str, status: str, message: str = None,
|
||||
error_trace: str = None, result_data: dict = None,
|
||||
started_at: datetime = None, finished_at: datetime = None,
|
||||
triggered_by: str = "scheduler", next_run_time: datetime = None):
|
||||
try:
|
||||
from ..database import SessionLocal
|
||||
from ..models import TaskLog
|
||||
db = SessionLocal()
|
||||
try:
|
||||
duration = None
|
||||
if started_at and finished_at:
|
||||
duration = int((finished_at - started_at).total_seconds())
|
||||
log = TaskLog(
|
||||
module_id=module_id,
|
||||
task_name=MODULES.get(module_id, {}).get("name", module_id),
|
||||
status=status,
|
||||
message=message,
|
||||
error_trace=error_trace,
|
||||
triggered_by=triggered_by,
|
||||
result_data=result_data or {},
|
||||
started_at=started_at or datetime.now(timezone.utc),
|
||||
finished_at=finished_at,
|
||||
duration=duration,
|
||||
next_run_time=next_run_time,
|
||||
)
|
||||
db.add(log)
|
||||
db.commit()
|
||||
finally:
|
||||
db.close()
|
||||
except Exception:
|
||||
pass
|
||||
|
||||
def _wrap_task(module_id: str, target_fn, *args, **kwargs):
|
||||
started = datetime.now(timezone.utc)
|
||||
status = "running"
|
||||
error_trace = None
|
||||
result_data = None
|
||||
try:
|
||||
result = target_fn(*args, **kwargs)
|
||||
status = "success"
|
||||
if isinstance(result, dict):
|
||||
result_data = {k: v for k, v in result.items() if isinstance(v, (str, int, float, bool, list, dict)) and k not in ("stdout", "stderr")}
|
||||
return result
|
||||
except Exception as e:
|
||||
status = "failed"
|
||||
import traceback
|
||||
error_trace = traceback.format_exc()
|
||||
raise
|
||||
finally:
|
||||
_log_task(module_id, status=status, message=None, error_trace=error_trace,
|
||||
result_data=result_data, started_at=started,
|
||||
finished_at=datetime.now(timezone.utc))
|
||||
|
||||
class TaskScheduler:
|
||||
def __init__(self):
|
||||
self.scheduler = BackgroundScheduler()
|
||||
@@ -24,64 +91,47 @@ class TaskScheduler:
|
||||
if self._started:
|
||||
logger.warning("Scheduler already started")
|
||||
return
|
||||
# 使用 CronTrigger 设置每日固定时间点
|
||||
# 顺序: 搜索缓存(01:00)→趋势(01:10)→采集(01:30)→创作(02:00)→审查(03:00)→源优化(05:00)→指标(06:00)
|
||||
self.scheduler.add_job(
|
||||
self._run_refresh_search_cache,
|
||||
CronTrigger(hour=1, minute=0),
|
||||
id='scheduled_refresh_search_cache',
|
||||
replace_existing=True,
|
||||
max_instances=1,
|
||||
coalesce=True
|
||||
)
|
||||
self.scheduler.add_job(
|
||||
self._run_fetch_trends,
|
||||
CronTrigger(hour=1, minute=10),
|
||||
id='scheduled_fetch_trends',
|
||||
replace_existing=True,
|
||||
max_instances=1,
|
||||
coalesce=True
|
||||
)
|
||||
self.scheduler.add_job(
|
||||
self._run_collect,
|
||||
CronTrigger(hour=1, minute=30),
|
||||
id='scheduled_collect',
|
||||
)
|
||||
self.scheduler.add_job(
|
||||
self._run_generate,
|
||||
CronTrigger(hour=2, minute=0),
|
||||
id='scheduled_generate',
|
||||
replace_existing=True,
|
||||
max_instances=1,
|
||||
coalesce=True
|
||||
)
|
||||
self.scheduler.add_job(
|
||||
self._run_optimize,
|
||||
CronTrigger(hour=3, minute=0),
|
||||
id='scheduled_optimize',
|
||||
replace_existing=True,
|
||||
max_instances=1,
|
||||
coalesce=True
|
||||
)
|
||||
self.scheduler.add_job(
|
||||
self._run_optimize_sources,
|
||||
CronTrigger(hour=5, minute=0),
|
||||
id='scheduled_optimize_sources',
|
||||
replace_existing=True,
|
||||
max_instances=1,
|
||||
coalesce=True
|
||||
)
|
||||
self.scheduler.add_job(
|
||||
self._run_metrics_sync,
|
||||
CronTrigger(hour=6, minute=0),
|
||||
id='scheduled_metrics_sync',
|
||||
replace_existing=True,
|
||||
max_instances=1,
|
||||
coalesce=True
|
||||
)
|
||||
from ..database import SessionLocal
|
||||
from ..models import TaskConfig
|
||||
db = SessionLocal()
|
||||
try:
|
||||
configs = {c.module_id: c for c in db.query(TaskConfig).all()}
|
||||
finally:
|
||||
db.close()
|
||||
|
||||
MODULE_JOBS = [
|
||||
("scheduled_refresh_search_cache", self._run_refresh_search_cache, "搜索缓存"),
|
||||
("scheduled_fetch_trends", self._run_fetch_trends, "热点趋势"),
|
||||
("scheduled_collect", self._run_collect, "内容采集"),
|
||||
("scheduled_generate", self._run_generate, "内容创作"),
|
||||
("scheduled_optimize", self._run_optimize, "合规审查"),
|
||||
("scheduled_optimize_sources", self._run_optimize_sources, "信息源优化"),
|
||||
("scheduled_metrics_sync", self._run_metrics_sync, "指标同步"),
|
||||
]
|
||||
|
||||
for module_id, fn, name in MODULE_JOBS:
|
||||
cfg = configs.get(module_id)
|
||||
if cfg and not cfg.enabled:
|
||||
logger.info(f"跳过禁用任务: {module_id}")
|
||||
continue
|
||||
schedule = (cfg.schedule if cfg else None) or MODULES.get(module_id, {}).get("cron", "01:00")
|
||||
try:
|
||||
hour, minute = map(int, schedule.split(":"))
|
||||
except (ValueError, AttributeError):
|
||||
hour, minute = 1, 0
|
||||
self.scheduler.add_job(
|
||||
fn,
|
||||
CronTrigger(hour=hour, minute=minute),
|
||||
id=module_id,
|
||||
replace_existing=True,
|
||||
max_instances=1,
|
||||
coalesce=True
|
||||
)
|
||||
logger.info(f"调度任务: {module_id} -> {schedule}")
|
||||
|
||||
self.scheduler.start()
|
||||
self._started = True
|
||||
logger.info("Scheduler started: 01:00 search 01:10 trends 01:30 collect 02:00 create 03:00 review 05:00 sources 06:00 metrics")
|
||||
logger.info("Scheduler started with dynamic schedule from TaskConfig")
|
||||
def shutdown(self):
|
||||
if self.scheduler.running:
|
||||
self.scheduler.shutdown()
|
||||
@@ -89,6 +139,8 @@ class TaskScheduler:
|
||||
|
||||
def _run_fetch_trends(self):
|
||||
"""定时刷新热点趋势（百度/微博/知乎实时热搜 + LLM补充）"""
|
||||
started = datetime.now(timezone.utc)
|
||||
_log_task("scheduled_fetch_trends", "running", started_at=started)
|
||||
try:
|
||||
logger.info("[Scheduled] Fetching hot trends...")
|
||||
import subprocess
|
||||
@@ -100,14 +152,26 @@ class TaskScheduler:
|
||||
for line in result.stdout.strip().split("\n"):
|
||||
if line.strip():
|
||||
logger.info("[Trends] %s", line.strip())
|
||||
logger.info("[Scheduled] Trends refreshed successfully")
|
||||
_log_task("scheduled_fetch_trends", "success",
|
||||
message="趋势刷新成功",
|
||||
result_data={"output_lines": len(result.stdout.splitlines())},
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
else:
|
||||
logger.warning("[Scheduled] Trends refresh failed: %s", result.stderr[-500:])
|
||||
_log_task("scheduled_fetch_trends", "failed",
|
||||
message=f"返回码 {result.returncode}",
|
||||
error_trace=result.stderr[-500:],
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
except Exception as e:
|
||||
_log_task("scheduled_fetch_trends", "failed",
|
||||
message=str(e),
|
||||
error_trace=traceback.format_exc(),
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.exception("[Scheduled] Trends refresh error: %s", e)
|
||||
|
||||
def _run_refresh_search_cache(self):
|
||||
"""定时刷新搜索缓存（通过 opencode webfetch）"""
|
||||
started = datetime.now(timezone.utc)
|
||||
_log_task("scheduled_refresh_search_cache", "running", started_at=started)
|
||||
try:
|
||||
logger.info("[Scheduled] Refreshing search cache via opencode...")
|
||||
import subprocess
|
||||
@@ -122,48 +186,95 @@ class TaskScheduler:
|
||||
if line.strip():
|
||||
logger.warning("[SearchCache] %s", line.strip())
|
||||
if result.returncode == 0:
|
||||
_log_task("scheduled_refresh_search_cache", "success",
|
||||
message="搜索缓存刷新成功",
|
||||
result_data={"output_lines": len(result.stdout.splitlines())},
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.info("[Scheduled] Search cache refreshed")
|
||||
else:
|
||||
_log_task("scheduled_refresh_search_cache", "failed",
|
||||
message="部分失败",
|
||||
error_trace=result.stderr[-500:],
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.warning("[Scheduled] Search cache refresh may have partial failures")
|
||||
except subprocess.TimeoutExpired:
|
||||
_log_task("scheduled_refresh_search_cache", "failed",
|
||||
message="超时",
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.warning("[Scheduled] Search cache refresh timed out")
|
||||
except Exception as e:
|
||||
_log_task("scheduled_refresh_search_cache", "failed",
|
||||
message=str(e),
|
||||
error_trace=traceback.format_exc(),
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.exception("[Scheduled] Search cache refresh error: %s", e)
|
||||
|
||||
def _run_generate(self):
|
||||
started = datetime.now(timezone.utc)
|
||||
_log_task("scheduled_generate", "running", started_at=started)
|
||||
try:
|
||||
logger.info("[Scheduled] Starting content generation...")
|
||||
result = run_creator_blocking()
|
||||
logger.info("[Scheduled] Generation completed: %s", result)
|
||||
created_id = result.get("topic_id") if isinstance(result, dict) else None
|
||||
review_result = None
|
||||
if created_id:
|
||||
logger.info("[Scheduled] Running compliance review on %s...", created_id)
|
||||
review_result = run_optimizer_blocking([created_id])
|
||||
if review_result.get("ok"):
|
||||
logger.info("[Scheduled] Review completed for %s", created_id)
|
||||
else:
|
||||
logger.warning("[Scheduled] Review failed: %s", review_result.get("error"))
|
||||
_log_task("scheduled_generate", "success",
|
||||
message=f"创作完成" + (f", 选题 {created_id}" if created_id else ""),
|
||||
result_data={"topic_id": created_id, "review_ok": review_result.get("ok") if review_result else None},
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
except Exception as e:
|
||||
_log_task("scheduled_generate", "failed",
|
||||
message=str(e),
|
||||
error_trace=traceback.format_exc(),
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.exception("[Scheduled] Generation pipeline failed: %s", e)
|
||||
|
||||
def _run_optimize(self):
|
||||
started = datetime.now(timezone.utc)
|
||||
_log_task("scheduled_optimize", "running", started_at=started)
|
||||
try:
|
||||
logger.info("[Scheduled] Starting compliance review...")
|
||||
result = run_optimizer_blocking()
|
||||
_log_task("scheduled_optimize", "success",
|
||||
message="合规审查完成",
|
||||
result_data={"processed": result.get("processed", 0), "passed": result.get("passed", 0)},
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.info("[Scheduled] Review completed: %s", result)
|
||||
except Exception as e:
|
||||
_log_task("scheduled_optimize", "failed",
|
||||
message=str(e),
|
||||
error_trace=traceback.format_exc(),
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.exception("[Scheduled] Review failed: %s", e)
|
||||
|
||||
def _run_collect(self):
|
||||
started = datetime.now(timezone.utc)
|
||||
_log_task("scheduled_collect", "running", started_at=started)
|
||||
try:
|
||||
logger.info("[Scheduled] Starting topic collection...")
|
||||
result = run_collector_blocking()
|
||||
topics_count = result.get("topics_count", 0)
|
||||
_log_task("scheduled_collect", "success",
|
||||
message=f"采集完成，找到 {topics_count} 个选题",
|
||||
result_data={"topics_count": topics_count, "output": str(result.get("output", ""))[:200]},
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.info("[Scheduled] Collection completed: %s", result.get("output", "")[-200:])
|
||||
except Exception as e:
|
||||
_log_task("scheduled_collect", "failed",
|
||||
message=str(e),
|
||||
error_trace=traceback.format_exc(),
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.exception("[Scheduled] Collection failed: %s", e)
|
||||
|
||||
def _run_optimize_sources(self):
|
||||
"""AI自动优化采集类别与信息源：对比市场热点和当前配置，给出调整建议"""
|
||||
started = datetime.now(timezone.utc)
|
||||
_log_task("scheduled_optimize_sources", "running", started_at=started)
|
||||
try:
|
||||
logger.info("[Scheduled] Starting source optimization with AI...")
|
||||
from .nvidia_client import call_llm
|
||||
@@ -183,32 +294,16 @@ class TaskScheduler:
|
||||
cat_names = [c.name for c in cats]
|
||||
src_summary = "\n".join(f"- [{s.source_type}] {s.name}: {s.query or s.url or ''}" for s in sources)
|
||||
|
||||
prompt = f"""你是一个内容策略分析师。分析当前中文互联网可持续生活领域的真实热点，与以下配置进行对比。
|
||||
prompt = get_prompt("sources_optimization",
|
||||
n=len(cat_names),
|
||||
cat_names="\n".join(f"- {n}" for n in cat_names),
|
||||
n2=len(sources),
|
||||
src_summary=src_summary,
|
||||
year=datetime.now().year,
|
||||
)
|
||||
|
||||
当前配置的类别（{len(cat_names)}个）：
|
||||
{chr(10).join(f'- {n}' for n in cat_names)}
|
||||
|
||||
当前配置的信息源（{len(sources)}个）：
|
||||
{src_summary}
|
||||
|
||||
请完成以下任务：
|
||||
1. 评估每个类别是否仍符合2026年中国市场真实热点（基于你的知识）
|
||||
2. 评估每个信息源是否可能在中国正常访问
|
||||
3. 建议新增或删除的类别（最多2条）
|
||||
4. 建议新增的信息源搜索词（最多3条，包含具体搜索词）
|
||||
|
||||
输出 JSON 格式：
|
||||
{{
|
||||
"category_assessment": [{{"name": "类别名", "status": "保留/淘汰/合并", "reason": "原因"}}],
|
||||
"source_assessment": [{{"name": "源名", "status": "保留/淘汰/替换", "reason": "原因"}}],
|
||||
"suggested_new_categories": [{{"name": "类别名", "search_query": "搜索词", "reason": "推荐原因"}}],
|
||||
"suggested_new_sources": [{{"name": "源名", "type": "web_search", "query": "搜索词", "focus": "聚焦领域"}}],
|
||||
"summary": "一句话总结本次优化建议"
|
||||
}}
|
||||
|
||||
只输出JSON，不要其他文字。"""
|
||||
|
||||
resp = call_llm(prompt, temperature=0.5, max_tokens=2000)
|
||||
params = get_prompt_params("sources_optimization")
|
||||
resp = call_llm(prompt, temperature=params.get("temperature", 0.5), max_tokens=params.get("max_tokens", 3000))
|
||||
if resp.startswith("```"):
|
||||
resp = resp.split("\n", 1)[1].rsplit("\n", 1)[0]
|
||||
result = json.loads(resp)
|
||||
@@ -222,12 +317,25 @@ class TaskScheduler:
|
||||
db.add(SystemConfig(key="collector_ai_advice", value=json.dumps(result, ensure_ascii=False), description="AI每日采集优化建议"))
|
||||
db.commit()
|
||||
logger.info("[Scheduled] Source AI optimization completed: %s", result.get("summary", ""))
|
||||
_log_task("scheduled_optimize_sources", "success",
|
||||
message=result.get("summary", "优化完成"),
|
||||
result_data={"categories_assessed": len(result.get("category_assessment", [])),
|
||||
"sources_assessed": len(result.get("source_assessment", [])),
|
||||
"suggested_cats": len(result.get("suggested_new_categories", [])),
|
||||
"suggested_srcs": len(result.get("suggested_new_sources", []))},
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
db.close()
|
||||
except Exception as e:
|
||||
_log_task("scheduled_optimize_sources", "failed",
|
||||
message=str(e),
|
||||
error_trace=traceback.format_exc(),
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.exception("[Scheduled] Source AI optimization failed: %s", e)
|
||||
|
||||
def _run_metrics_sync(self):
|
||||
"""定时从各平台公开API获取发布文章的效果数据（当前仅支持知乎）"""
|
||||
started = datetime.now(timezone.utc)
|
||||
_log_task("scheduled_metrics_sync", "running", started_at=started)
|
||||
try:
|
||||
logger.info("[Scheduled] Starting metrics sync (zhihu auto-fetch)...")
|
||||
from ..database import SessionLocal
|
||||
@@ -286,6 +394,10 @@ class TaskScheduler:
|
||||
if count:
|
||||
db.commit()
|
||||
logger.info("[Scheduled] Metrics sync completed: synced %d zhihu articles", count)
|
||||
_log_task("scheduled_metrics_sync", "success",
|
||||
message=f"同步完成，{count} 篇知乎文章",
|
||||
result_data={"articles_synced": count},
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
# 生成指标反馈：按 field 聚合表现，写入 metrics_feedback.json 供 collector 读取
|
||||
try:
|
||||
import json as json_mod
|
||||
@@ -317,9 +429,16 @@ class TaskScheduler:
|
||||
except Exception as e_fb:
|
||||
logger.warning("[Scheduled] Metrics feedback generation failed: %s", e_fb)
|
||||
else:
|
||||
_log_task("scheduled_metrics_sync", "success",
|
||||
message="无已发布的知乎文章",
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.info("[Scheduled] Metrics sync: no zhihu articles to sync")
|
||||
db.close()
|
||||
except Exception as e:
|
||||
_log_task("scheduled_metrics_sync", "failed",
|
||||
message=str(e),
|
||||
error_trace=traceback.format_exc(),
|
||||
started_at=started, finished_at=datetime.now(timezone.utc))
|
||||
logger.exception("[Scheduled] Metrics sync failed: %s", e)
|
||||
|
||||
def get_jobs(self):
|
||||
|
||||
Reference in New Issue
Block a user