feat: 内容数据迁移至数据库，合规审查全链路打通
- 文章 HTML 存储从文件系统迁移至 articles 表，删除 releases 目录 - 合规审查从 DB 读取 HTML，审查结果写回 DB，通过后自动推进至待发布 - 新增 todayCount 筛选按钮，与系统概览统计数据一致 - 全屏预览修复：提升 z-index 超过侧边栏，添加退出全屏/关闭按钮 - 统一 '优化' → '审查' 命名，消除前后端术语不一致 - 调度器创作完成后自动触发审查（生成 → 审查 → 待发布） - 清理旧备份/调试文件、过期大纲和研究笔记
This commit is contained in:
Yuzhiran Dev
+11
-45
@@ -1,73 +1,39 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
调整选题优先级（数据库 + JSON 备份）
|
||||
"""
|
||||
|
||||
import sys
|
||||
from pathlib import Path
|
||||
from datetime import datetime
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
sys.path.insert(0, str(PROJECT_ROOT / 'platform' / 'backend'))
|
||||
|
||||
try:
|
||||
from db_helper import get_topic_by_id, update_topic_status
|
||||
from app.database import SessionLocal
|
||||
from app.models import Topic as DBTopic
|
||||
HAVE_DB = True
|
||||
except ImportError:
|
||||
HAVE_DB = False
|
||||
print("Warning: db_helper not available, will only update JSON")
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TOPICS_FILE = DATA_DIR / "sustainability_topics.json"
|
||||
|
||||
def adjust_json_priority(adjustments):
|
||||
try:
|
||||
with open(TOPICS_FILE, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
except:
|
||||
topics = []
|
||||
updated = []
|
||||
for t in topics:
|
||||
if t['id'] in adjustments:
|
||||
old = t.get('priority_score', 0)
|
||||
t['priority_score'] = adjustments[t['id']]
|
||||
updated.append(f"{t['id']}: {old} -> {adjustments[t['id']]}")
|
||||
with open(TOPICS_FILE, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
return updated
|
||||
from db_helper import get_topic_by_id
|
||||
from app.database import SessionLocal
|
||||
from app.models import Topic as DBTopic
|
||||
|
||||
def adjust_db_priority(adjustments):
|
||||
if not HAVE_DB:
|
||||
return []
|
||||
updated = []
|
||||
db = SessionLocal()
|
||||
try:
|
||||
for tid, new_score in adjustments.items():
|
||||
topic = db.query(DBTopic).filter(DBTopic.id == tid).first()
|
||||
if topic:
|
||||
old = topic.priority_score
|
||||
topic.priority_score = new_score
|
||||
topic.updated_at = datetime.now()
|
||||
updated.append(f"{tid}: {topic.priority_score} -> {new_score}")
|
||||
updated.append(f"{tid}: {old} -> {new_score}")
|
||||
db.commit()
|
||||
finally:
|
||||
db.close()
|
||||
return updated
|
||||
|
||||
def main():
|
||||
# 定义需要调整的优先级：ID -> 新分数
|
||||
adjustments = {
|
||||
'D01': 11,
|
||||
'B05': 10
|
||||
}
|
||||
adjustments = {'D01': 11, 'B05': 10}
|
||||
print("调整优先级...")
|
||||
db_updated = adjust_db_priority(adjustments) if HAVE_DB else []
|
||||
if db_updated:
|
||||
print("[DB] updated:", ', '.join(db_updated))
|
||||
json_updated = adjust_json_priority(adjustments)
|
||||
print("[JSON] updated:", ', '.join(json_updated))
|
||||
updated = adjust_db_priority(adjustments)
|
||||
if updated:
|
||||
print("[DB] updated:", ', '.join(updated))
|
||||
print("完成")
|
||||
|
||||
if __name__ == "__main__":
|
||||
import json, datetime
|
||||
main()
|
||||
|
||||
@@ -4,7 +4,7 @@
|
||||
遍历指定日期所有发布版本，执行合规检查，生成汇总报告
|
||||
"""
|
||||
|
||||
import json, re, datetime
|
||||
import re, datetime
|
||||
from pathlib import Path
|
||||
import sys
|
||||
|
||||
@@ -12,29 +12,11 @@ PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
|
||||
from scripts.compliance_checker import check_article
|
||||
|
||||
# 尝试导入数据库
|
||||
try:
|
||||
from db_helper import export_topics_to_json
|
||||
HAVE_DB = True
|
||||
except ImportError:
|
||||
HAVE_DB = False
|
||||
from db_helper import get_topic_by_id
|
||||
|
||||
# 配置
|
||||
RELEASE_DIR = PROJECT_ROOT / "automation" / "data" / "releases"
|
||||
TODAY = datetime.date.today().isoformat() # 默认今天，可修改
|
||||
|
||||
def load_topics_from_db():
|
||||
if not HAVE_DB:
|
||||
raise RuntimeError("Database not available")
|
||||
topics = export_topics_to_json()
|
||||
return {t['id']: t for t in topics}
|
||||
|
||||
def load_topics_from_json():
|
||||
json_path = PROJECT_ROOT / "automation" / "data" / "sustainability_topics.json"
|
||||
with open(json_path, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
return {t['id']: t for t in topics}
|
||||
TODAY = datetime.date.today().isoformat()
|
||||
|
||||
def extract_topic_id(filename: Path) -> str:
|
||||
stem = filename.stem
|
||||
@@ -48,14 +30,6 @@ def main(target_date: str = None):
|
||||
target_date = TODAY
|
||||
print(f"批量合规审查: {target_date}")
|
||||
|
||||
# 加载选题数据（优先DB，失败则备援JSON）
|
||||
try:
|
||||
topics_by_id = load_topics_from_db()
|
||||
print("[数据源] 数据库")
|
||||
except Exception as e:
|
||||
print(f"[数据源] 数据库失败: {e}, 改用 JSON")
|
||||
topics_by_id = load_topics_from_json()
|
||||
|
||||
release_path = RELEASE_DIR / target_date
|
||||
if not release_path.exists():
|
||||
print(f"错误：发布日期目录不存在 {release_path}")
|
||||
@@ -68,7 +42,7 @@ def main(target_date: str = None):
|
||||
for html_file in html_files:
|
||||
platform = html_file.parent.name
|
||||
topic_id = extract_topic_id(html_file)
|
||||
topic_data = topics_by_id.get(topic_id) if topic_id else None
|
||||
topic_data = get_topic_by_id(topic_id) if topic_id else None
|
||||
|
||||
with open(html_file, 'r', encoding='utf-8') as f:
|
||||
html_content = f.read()
|
||||
@@ -80,7 +54,6 @@ def main(target_date: str = None):
|
||||
result['topic_title'] = topic_data.get('title') if topic_data else "未知"
|
||||
results.append(result)
|
||||
|
||||
# 输出摘要
|
||||
passed = sum(1 for r in results if r['passed'])
|
||||
failed = len(results) - passed
|
||||
print(f"\n✅ 通过: {passed}, ⚠️ 需人工: {failed}")
|
||||
|
||||
+19
-35
@@ -1,46 +1,30 @@
|
||||
#!/usr/bin/env python3
|
||||
import sys
|
||||
from pathlib import Path
|
||||
from datetime import datetime
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
try:
|
||||
from db_helper import update_topic_status
|
||||
from app.database import SessionLocal
|
||||
from app.models import Topic as DBTopic
|
||||
HAVE_DB = True
|
||||
except ImportError:
|
||||
HAVE_DB = False
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TOPICS_FILE = DATA_DIR / "sustainability_topics.json"
|
||||
sys.path.insert(0, str(PROJECT_ROOT / 'platform' / 'backend'))
|
||||
|
||||
from app.database import SessionLocal
|
||||
from app.models import Topic as DBTopic
|
||||
|
||||
def adjust(adjustments):
|
||||
db_ok = False
|
||||
if HAVE_DB:
|
||||
db = SessionLocal()
|
||||
try:
|
||||
for tid, new_score in adjustments.items():
|
||||
topic = db.query(DBTopic).filter(DBTopic.id == tid).first()
|
||||
if topic:
|
||||
topic.priority_score = new_score
|
||||
topic.updated_at = datetime.datetime.now()
|
||||
db.commit()
|
||||
db_ok = True
|
||||
finally:
|
||||
db.close()
|
||||
db = SessionLocal()
|
||||
try:
|
||||
with open(TOPICS_FILE, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
for t in topics:
|
||||
if t['id'] in adjustments:
|
||||
t['priority_score'] = adjustments[t['id']]
|
||||
with open(TOPICS_FILE, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
except:
|
||||
pass
|
||||
return db_ok
|
||||
for tid, new_score in adjustments.items():
|
||||
topic = db.query(DBTopic).filter(DBTopic.id == tid).first()
|
||||
if topic:
|
||||
topic.priority_score = new_score
|
||||
topic.updated_at = datetime.now()
|
||||
db.commit()
|
||||
finally:
|
||||
db.close()
|
||||
|
||||
def main():
|
||||
adjustments = {'B05': 15}
|
||||
adjust(adjustments)
|
||||
adjust({'B05': 15})
|
||||
print("B05 priority_score set to 15")
|
||||
|
||||
if __name__ == "__main__":
|
||||
import json, datetime
|
||||
main()
|
||||
|
||||
+19
-35
@@ -1,46 +1,30 @@
|
||||
#!/usr/bin/env python3
|
||||
import sys
|
||||
from pathlib import Path
|
||||
from datetime import datetime
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
try:
|
||||
from db_helper import update_topic_status
|
||||
from app.database import SessionLocal
|
||||
from app.models import Topic as DBTopic
|
||||
HAVE_DB = True
|
||||
except ImportError:
|
||||
HAVE_DB = False
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TOPICS_FILE = DATA_DIR / "sustainability_topics.json"
|
||||
sys.path.insert(0, str(PROJECT_ROOT / 'platform' / 'backend'))
|
||||
|
||||
from app.database import SessionLocal
|
||||
from app.models import Topic as DBTopic
|
||||
|
||||
def adjust(adjustments):
|
||||
db_ok = False
|
||||
if HAVE_DB:
|
||||
db = SessionLocal()
|
||||
try:
|
||||
for tid, new_score in adjustments.items():
|
||||
topic = db.query(DBTopic).filter(DBTopic.id == tid).first()
|
||||
if topic:
|
||||
topic.priority_score = new_score
|
||||
topic.updated_at = datetime.datetime.now()
|
||||
db.commit()
|
||||
db_ok = True
|
||||
finally:
|
||||
db.close()
|
||||
db = SessionLocal()
|
||||
try:
|
||||
with open(TOPICS_FILE, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
for t in topics:
|
||||
if t['id'] in adjustments:
|
||||
t['priority_score'] = adjustments[t['id']]
|
||||
with open(TOPICS_FILE, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
except:
|
||||
pass
|
||||
return db_ok
|
||||
for tid, new_score in adjustments.items():
|
||||
topic = db.query(DBTopic).filter(DBTopic.id == tid).first()
|
||||
if topic:
|
||||
topic.priority_score = new_score
|
||||
topic.updated_at = datetime.now()
|
||||
db.commit()
|
||||
finally:
|
||||
db.close()
|
||||
|
||||
def main():
|
||||
adjustments = {'D01': 12}
|
||||
adjust(adjustments)
|
||||
adjust({'D01': 12})
|
||||
print("D01 priority_score set to 12")
|
||||
|
||||
if __name__ == "__main__":
|
||||
import json, datetime
|
||||
main()
|
||||
|
||||
@@ -1,7 +1,10 @@
|
||||
#!/usr/bin/env python3
|
||||
import json
|
||||
data = json.load(open('automation/data/sustainability_topics.json', 'r', encoding='utf-8'))
|
||||
for t in data:
|
||||
if t['id'] == 'D01':
|
||||
print(f"D01 title: {t['title']} (length: {len(t['title'])})")
|
||||
print("Field:", t.get('field'))
|
||||
import sys
|
||||
from pathlib import Path
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent))
|
||||
from db_helper import get_topic_by_id
|
||||
|
||||
t = get_topic_by_id('D01')
|
||||
if t:
|
||||
print(f"D01 title: {t['title']} (length: {len(t['title'])})")
|
||||
print("Field:", t.get('field'))
|
||||
|
||||
@@ -1,6 +1,10 @@
|
||||
#!/usr/bin/env python3
|
||||
import json
|
||||
data = json.load(open('automation/data/sustainability_topics.json', 'r', encoding='utf-8'))
|
||||
for t in data:
|
||||
if t['id'] in ['D01','B05']:
|
||||
print(f"{t['id']}: priority_score = {t['priority_score']}, status = {t.get('status')}")
|
||||
import sys
|
||||
from pathlib import Path
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent))
|
||||
from db_helper import get_topic_by_id
|
||||
|
||||
for tid in ['D01', 'B05']:
|
||||
t = get_topic_by_id(tid)
|
||||
if t:
|
||||
print(f"{tid}: priority_score = {t['priority_score']}, status = {t.get('status')}")
|
||||
|
||||
+1
-15
@@ -544,21 +544,7 @@ class SustainabilityCollector:
|
||||
except Exception as e:
|
||||
logger.error(f"选题数据库保存失败: {e}")
|
||||
|
||||
# 3. 可选: 更新 JSON 备份 (仅新增,避免覆盖锁信息)
|
||||
main_topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
try:
|
||||
if main_topics_file.exists():
|
||||
with open(main_topics_file, 'r', encoding='utf-8') as f:
|
||||
existing_topics = json.load(f)
|
||||
else:
|
||||
existing_topics = []
|
||||
existing_ids = {t['id'] for t in existing_topics}
|
||||
new_additions = [asdict(topic) for topic in self.new_topics if topic.id not in existing_ids]
|
||||
existing_topics.extend(new_additions)
|
||||
with open(main_topics_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(existing_topics, f, ensure_ascii=False, indent=2)
|
||||
except Exception as e:
|
||||
logger.error(f"JSON备份失败: {e}")
|
||||
# DB 是唯一数据源，不再写 JSON
|
||||
|
||||
def send_wecom_notification(self):
|
||||
"""发送企业微信通知"""
|
||||
|
||||
@@ -1,438 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
可持续性内容收集脚本
|
||||
每天凌晨5:00运行，收集全球可持续性趋势信息，提炼选题和案例
|
||||
"""
|
||||
|
||||
import os
|
||||
import sys
|
||||
import yaml
|
||||
import json
|
||||
import datetime
|
||||
import logging
|
||||
from pathlib import Path
|
||||
import feedparser
|
||||
import requests
|
||||
import re
|
||||
from typing import Dict, List, Optional, Tuple
|
||||
import hashlib
|
||||
from dataclasses import dataclass, asdict
|
||||
import subprocess
|
||||
|
||||
# 项目根目录
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
|
||||
# 配置路径
|
||||
CONFIG_DIR = PROJECT_ROOT / "config"
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
# 日志配置
|
||||
logging.basicConfig(
|
||||
level=logging.INFO,
|
||||
format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
|
||||
handlers=[
|
||||
logging.FileHandler(LOGS_DIR / f"collector_{TODAY}.log"),
|
||||
logging.StreamHandler()
|
||||
]
|
||||
)
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
@dataclass
|
||||
class SustainabilitySource:
|
||||
"""可持续性信息源"""
|
||||
name: str
|
||||
type: str # rss, web, api, report
|
||||
url: str
|
||||
update_frequency: str
|
||||
credibility: str # high, medium, low
|
||||
focus: str
|
||||
|
||||
@dataclass
|
||||
class SustainabilityCase:
|
||||
"""可持续性案例"""
|
||||
id: str
|
||||
country: str
|
||||
category: str # 子领域：城市农业、零浪费生活等
|
||||
title: str
|
||||
core_idea: str
|
||||
data_facts: str
|
||||
global_advantage: str
|
||||
china_pain_point: str
|
||||
localization_suggestion: str
|
||||
mvp_action: str
|
||||
source_url: str
|
||||
credibility_rating: str # ⭐⭐ ⭐⭐⭐
|
||||
china_applicability: str # ⭐ ⭐⭐ ⭐⭐⭐
|
||||
collection_date: str
|
||||
status: str = "待验证"
|
||||
|
||||
@dataclass
|
||||
class SustainabilityTopic:
|
||||
"""可持续性选题"""
|
||||
id: str
|
||||
title: str
|
||||
cases: List[str] # 关联的案例ID列表
|
||||
audience: str # 目标受众
|
||||
china_pain_points: str
|
||||
localization_solution: str
|
||||
mvp_actions: str
|
||||
estimated_length: int
|
||||
priority_score: float
|
||||
status: str = "待创作"
|
||||
|
||||
class SustainabilityCollector:
|
||||
"""可持续性内容收集器"""
|
||||
|
||||
def __init__(self):
|
||||
self.load_config()
|
||||
self.today_dir = DATA_DIR / "sustainability_raw" / TODAY
|
||||
self.today_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# 结果存储
|
||||
self.new_cases: List[SustainabilityCase] = []
|
||||
self.new_topics: List[SustainabilityTopic] = []
|
||||
|
||||
def load_config(self):
|
||||
"""加载配置文件"""
|
||||
with open(CONFIG_DIR / "sources.yaml", "r", encoding='utf-8') as f:
|
||||
self.config = yaml.safe_load(f)
|
||||
|
||||
with open(CONFIG_DIR / "wecom_config.yaml", "r", encoding='utf-8') as f:
|
||||
self.wecom_config = yaml.safe_load(f)
|
||||
|
||||
self.sources = []
|
||||
for source_group in self.config["sustainability_sources"].values():
|
||||
for source_info in source_group:
|
||||
# Handle both 'url' and 'base_url' in config
|
||||
source_info = source_info.copy()
|
||||
if 'base_url' in source_info and 'url' not in source_info:
|
||||
source_info['url'] = source_info.pop('base_url')
|
||||
# Provide defaults for missing optional fields
|
||||
source_info.setdefault('update_frequency', 'daily')
|
||||
source_info.setdefault('focus', '可持续性')
|
||||
# Filter to only fields accepted by SustainabilitySource
|
||||
allowed_keys = {'name', 'type', 'url', 'update_frequency', 'credibility', 'focus'}
|
||||
filtered_info = {k: v for k, v in source_info.items() if k in allowed_keys}
|
||||
self.sources.append(SustainabilitySource(**filtered_info))
|
||||
|
||||
logger.info(f"加载了 {len(self.sources)} 个信息源")
|
||||
|
||||
def fetch_rss_feed(self, source: SustainabilitySource) -> List[Dict]:
|
||||
"""获取RSS订阅内容"""
|
||||
try:
|
||||
feed = feedparser.parse(source.url)
|
||||
articles = []
|
||||
|
||||
for entry in feed.entries[:10]: # 限制数量
|
||||
# 检查是否包含可持续性关键词
|
||||
content = entry.get('summary', entry.get('description', ''))
|
||||
title = entry.get('title', '')
|
||||
|
||||
# 可持续性关键词匹配
|
||||
sustainability_keywords = [
|
||||
'sustainable', 'green', 'eco', 'circular', 'climate',
|
||||
'carbon', 'zero waste', 'renewable', 'recycle',
|
||||
'环保', '可持续', '碳中和', '循环经济', '零浪费'
|
||||
]
|
||||
|
||||
if any(keyword.lower() in (title + content).lower() for keyword in sustainability_keywords):
|
||||
articles.append({
|
||||
'title': title,
|
||||
'url': entry.get('link', ''),
|
||||
'content': content,
|
||||
'published': entry.get('published', ''),
|
||||
'source_name': source.name
|
||||
})
|
||||
|
||||
logger.info(f"从 {source.name} 获取到 {len(articles)} 篇可持续性文章")
|
||||
return articles
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"获取RSS失败 {source.name}: {e}")
|
||||
return []
|
||||
|
||||
def fetch_web_content(self, source: SustainabilitySource) -> List[Dict]:
|
||||
"""获取网页内容（简化版，实际需要更复杂的抓取）"""
|
||||
# 简化实现：只记录，不实际抓取
|
||||
logger.info(f"网页信息源 {source.name} 需要手动处理")
|
||||
return []
|
||||
|
||||
def analyze_article(self, article: Dict) -> Optional[SustainabilityCase]:
|
||||
"""分析文章内容，提炼案例"""
|
||||
try:
|
||||
content = article['content']
|
||||
title = article['title']
|
||||
|
||||
# 提取关键数据（简化版，实际可用NLP）
|
||||
data_patterns = [
|
||||
r'(\d+\.?\d*)\s*(?:percent|%|百分比)',
|
||||
r'(\d+\.?\d*)\s*(?:million|billion|万|亿)',
|
||||
r'(\d+\.?\d*)\s*(?:tons|tonnes|吨)',
|
||||
r'(\d+\.?\d*)\s*(?:reduction|increase|减少|增加)'
|
||||
]
|
||||
|
||||
data_points = []
|
||||
for pattern in data_patterns:
|
||||
matches = re.findall(pattern, content, re.IGNORECASE)
|
||||
if matches:
|
||||
data_points.extend(matches[:3]) # 限制数量
|
||||
|
||||
if len(data_points) < 2:
|
||||
logger.info(f"文章数据不足: {title}")
|
||||
return None
|
||||
|
||||
# 确定国家（简化判断）
|
||||
countries = ['China', 'Japan', 'Germany', 'US', 'UK', 'Sweden', 'Netherlands']
|
||||
country = 'Global' # 默认
|
||||
for c in countries:
|
||||
if c.lower() in content.lower():
|
||||
country = c
|
||||
break
|
||||
|
||||
# 确定子领域
|
||||
categories = self.config["sustainability_categories"]
|
||||
category = categories[0] # 默认第一个
|
||||
for cat in categories:
|
||||
if any(keyword in content.lower() for keyword in [cat.lower(), cat[:4].lower()]):
|
||||
category = cat
|
||||
break
|
||||
|
||||
# 生成案例ID
|
||||
case_id = f"SUS-{hashlib.md5(title.encode()).hexdigest()[:8].upper()}"
|
||||
|
||||
# 提取核心观点（简化版）
|
||||
# 实际应用中可用AI提取，这里用前100字符
|
||||
core_idea = content[:200] if len(content) > 200 else content
|
||||
|
||||
# 生成中国痛点（基于类别模板）
|
||||
china_pains = {
|
||||
"城市农业": "中国城市空间小、光照不足、怕邻居投诉",
|
||||
"零浪费生活": "中国垃圾分类执行难、环保产品溢价高",
|
||||
"低碳出行": "中国电动车充电难、城市规划不支持",
|
||||
"循环消费": "中国二手文化不成熟、维修成本高",
|
||||
"能源效率": "中国能源价格波动、设备更换成本高",
|
||||
"可持续饮食": "中国预制菜泛滥、有机食品价格高",
|
||||
"环保科技产品": "中国消费者关注价格多于环保"
|
||||
}
|
||||
china_pain = china_pains.get(category, "中国相关数据不足，需本土化验证")
|
||||
|
||||
# 生成案例
|
||||
case = SustainabilityCase(
|
||||
id=case_id,
|
||||
country=country,
|
||||
category=category,
|
||||
title=title,
|
||||
core_idea=core_idea,
|
||||
data_facts=f"数据点: {', '.join(data_points[:3])}",
|
||||
global_advantage="需进一步分析全球优势",
|
||||
china_pain_point=china_pain,
|
||||
localization_suggestion="需基于中国现实调整实施",
|
||||
mvp_action="建议先小规模试点验证",
|
||||
source_url=article['url'],
|
||||
credibility_rating="⭐⭐" if article['source_name'] in ['GreenBiz', 'Sustainable Brands'] else "⭐",
|
||||
china_applicability="⭐⭐",
|
||||
collection_date=TODAY
|
||||
)
|
||||
|
||||
return case
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"分析文章失败: {e}")
|
||||
return None
|
||||
|
||||
def generate_topic_from_cases(self, cases: List[SustainabilityCase]) -> Optional[SustainabilityTopic]:
|
||||
"""从案例组合生成选题"""
|
||||
if len(cases) < 2:
|
||||
return None
|
||||
|
||||
# 按类别分组
|
||||
category_cases = {}
|
||||
for case in cases:
|
||||
if case.category not in category_cases:
|
||||
category_cases[case.category] = []
|
||||
category_cases[case.category].append(case)
|
||||
|
||||
# 选择案例数最多的类别
|
||||
main_category = max(category_cases, key=lambda k: len(category_cases[k]))
|
||||
main_cases = category_cases[main_category]
|
||||
|
||||
if len(main_cases) < 2:
|
||||
return None
|
||||
|
||||
# 生成选题ID
|
||||
topic_id = f"TOPIC-.{hashlib.md5((main_category + TODAY).encode()).hexdigest()[:6].upper()}"
|
||||
|
||||
# 组合标题
|
||||
case_titles = [case.title[:30] for case in main_cases[:2]]
|
||||
topic_title = f"{main_category}新趋势: {case_titles[0]}与{case_titles[1]}的中国落地路径"
|
||||
|
||||
# 计算优先级分数
|
||||
priority_weights = self.config["topic_priority"]
|
||||
priority_score = (
|
||||
priority_weights["audience_match"] * 0.8 + # 受众匹配度预估
|
||||
priority_weights["data_availability"] * 0.9 + # 数据可得性
|
||||
priority_weights["uniqueness"] * 0.7 + # 独特性
|
||||
priority_weights["executability"] * 0.6 + # 可执行性
|
||||
priority_weights["brand_fit"] * 0.9 # 品牌契合度
|
||||
)
|
||||
|
||||
topic = SustainabilityTopic(
|
||||
id=topic_id,
|
||||
title=topic_title,
|
||||
cases=[case.id for case in main_cases],
|
||||
audience="城市焦虑青年（26-35岁）",
|
||||
china_pain_points=f"{main_category}在中国面临的主要问题",
|
||||
localization_solution="国际案例中国化适配方案",
|
||||
mvp_actions="读者可立即尝试的3个行动",
|
||||
estimated_length=2500,
|
||||
priority_score=round(priority_score, 2)
|
||||
)
|
||||
|
||||
return topic
|
||||
|
||||
def save_results(self):
|
||||
"""保存收集结果"""
|
||||
# 保存案例
|
||||
cases_file = self.today_dir / "new_cases.json"
|
||||
with open(cases_file, 'w', encoding='utf-8') as f:
|
||||
json.dump([asdict(case) for case in self.new_cases], f, ensure_ascii=False, indent=2)
|
||||
|
||||
# 保存选题
|
||||
topics_file = self.today_dir / "new_topics.json"
|
||||
with open(topics_file, 'w', encoding='utf-8') as f:
|
||||
json.dump([asdict(topic) for topic in self.new_topics], f, ensure_ascii=False, indent=2)
|
||||
|
||||
# 更新主数据库
|
||||
self.update_main_database()
|
||||
|
||||
logger.info(f"保存了 {len(self.new_cases)} 个案例和 {len(self.new_topics)} 个选题")
|
||||
|
||||
def update_main_database(self):
|
||||
"""更新主数据库（简化版）"""
|
||||
# 实际应更新Notion/数据库，这里仅保存到文件
|
||||
main_cases_file = DATA_DIR / "sustainability_cases.json"
|
||||
main_topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
|
||||
# 读取现有数据
|
||||
existing_cases = []
|
||||
existing_topics = []
|
||||
|
||||
if main_cases_file.exists():
|
||||
with open(main_cases_file, 'r', encoding='utf-8') as f:
|
||||
existing_cases = json.load(f)
|
||||
|
||||
if main_topics_file.exists():
|
||||
with open(main_topics_file, 'r', encoding='utf-8') as f:
|
||||
existing_topics = json.load(f)
|
||||
|
||||
# 合并新数据
|
||||
all_cases = existing_cases + [asdict(case) for case in self.new_cases]
|
||||
all_topics = existing_topics + [asdict(topic) for topic in self.new_topics]
|
||||
|
||||
# 保存（限制总数）
|
||||
with open(main_cases_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(all_cases[:100], f, ensure_ascii=False, indent=2)
|
||||
|
||||
with open(main_topics_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(all_topics[:50], f, ensure_ascii=False, indent=2)
|
||||
|
||||
def send_wecom_notification(self):
|
||||
"""发送企业微信通知"""
|
||||
try:
|
||||
# 调用通知脚本
|
||||
notification_script = PROJECT_ROOT / "scripts" / "wecom_notifier.py"
|
||||
if not notification_script.exists():
|
||||
logger.warning("企业微信通知脚本不存在")
|
||||
return
|
||||
|
||||
# 准备通知数据
|
||||
notification_data = {
|
||||
"task": "sustainability_collection",
|
||||
"time": datetime.datetime.now().strftime("%Y-%m-%d %H:%M"),
|
||||
"topic_count": len(self.new_topics),
|
||||
"case_count": len(self.new_cases),
|
||||
"source_count": len(self.sources),
|
||||
"details_link": str(self.today_dir.relative_to(PROJECT_ROOT))
|
||||
}
|
||||
|
||||
data_file = self.today_dir / "notification_data.json"
|
||||
with open(data_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(notification_data, f, ensure_ascii=False)
|
||||
|
||||
# 运行通知脚本
|
||||
result = subprocess.run(
|
||||
[sys.executable, str(notification_script), str(data_file)],
|
||||
capture_output=True,
|
||||
text=True,
|
||||
cwd=PROJECT_ROOT
|
||||
)
|
||||
|
||||
if result.returncode == 0:
|
||||
logger.info("企业微信通知发送成功")
|
||||
else:
|
||||
logger.error(f"通知发送失败: {result.stderr}")
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"发送通知失败: {e}")
|
||||
|
||||
def run(self):
|
||||
"""主运行流程"""
|
||||
logger.info("开始可持续性内容收集")
|
||||
|
||||
# 1. 从所有信息源收集
|
||||
all_articles = []
|
||||
for source in self.sources:
|
||||
if source.type == 'rss':
|
||||
articles = self.fetch_rss_feed(source)
|
||||
all_articles.extend(articles)
|
||||
elif source.type == 'web':
|
||||
articles = self.fetch_web_content(source)
|
||||
all_articles.extend(articles)
|
||||
|
||||
logger.info(f"总共收集到 {len(all_articles)} 篇可持续性文章")
|
||||
|
||||
# 2. 分析文章，提炼案例
|
||||
for article in all_articles[:20]: # 限制分析数量
|
||||
case = self.analyze_article(article)
|
||||
if case:
|
||||
self.new_cases.append(case)
|
||||
|
||||
# 3. 生成选题
|
||||
if self.new_cases:
|
||||
topic = self.generate_topic_from_cases(self.new_cases)
|
||||
if topic:
|
||||
self.new_topics.append(topic)
|
||||
|
||||
# 4. 保存结果
|
||||
self.save_results()
|
||||
|
||||
# 5. 发送通知
|
||||
self.send_wecom_notification()
|
||||
|
||||
logger.info(f"收集完成: {len(self.new_cases)} 案例, {len(self.new_topics)} 选题")
|
||||
return len(self.new_cases), len(self.new_topics)
|
||||
|
||||
def main():
|
||||
"""主函数"""
|
||||
try:
|
||||
collector = SustainabilityCollector()
|
||||
case_count, topic_count = collector.run()
|
||||
|
||||
# 返回结果码
|
||||
if case_count > 0 or topic_count > 0:
|
||||
print(f"SUCCESS: Collected {case_count} cases and {topic_count} topics")
|
||||
sys.exit(0)
|
||||
else:
|
||||
print("WARNING: No new content found")
|
||||
sys.exit(1)
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"收集任务失败: {e}")
|
||||
print(f"ERROR: {e}")
|
||||
sys.exit(1)
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,265 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
内容合规审查模块
|
||||
检查文章是否符合法律法规、平台规则、品牌规范
|
||||
"""
|
||||
|
||||
import re
|
||||
from typing import Dict, List, Tuple
|
||||
|
||||
# 敏感词库（示例，需要持续更新）
|
||||
SENSITIVE_WORDS = {
|
||||
"政治敏感": ["国家主席", "政治局", "常委", "军委", "统战部", "颠覆国家", "分裂主义", "台独", "疆独", "藏独"],
|
||||
"违禁内容": ["赌博", "毒品", "迷药", "枪支", "炸药", "色情", "低俗", "反动", "邪教"],
|
||||
"不实信息": [" guaranteed 赚钱", "一夜暴富", "100%有效", "包治百病", "绝对正确"],
|
||||
"领导人相关": ["主席", "总理", "总书记", "国家领导人"] # 需上下文判断
|
||||
}
|
||||
|
||||
# 平台规则限制
|
||||
PLATFORM_RULES = {
|
||||
"zhihu": {
|
||||
"max_title_len": 100,
|
||||
"min_word_count": 1000,
|
||||
"allowed_tags": ["科技", "生活", "职场", "教育", "可持续", "AI", "远程工作", "个人成长"],
|
||||
"forbidden_patterns": [r"加微信", r"私聊", r"付费咨询", r"点击领取"]
|
||||
},
|
||||
"wechat": {
|
||||
"max_title_len": 32,
|
||||
"min_word_count": 800,
|
||||
"allowed_tags": ["科技", "生活", "职场", "教育", "可持续", "AI", "远程工作", "个人成长"],
|
||||
"forbidden_patterns": [r"诱导分享", r"朋友圈", r"转发群"]
|
||||
},
|
||||
"xiaohongshu": {
|
||||
"max_title_len": 50,
|
||||
"min_word_count": 400,
|
||||
"allowed_tags": ["生活方式", "可持续", "AI", "个人成长", "极简", "环保"],
|
||||
"forbidden_patterns": [r"私信", r"加群", r"导流"]
|
||||
}
|
||||
}
|
||||
|
||||
class ComplianceChecker:
|
||||
"""合规审查器"""
|
||||
|
||||
def __init__(self):
|
||||
self.issues = []
|
||||
|
||||
def check_text(self, text: str, platform: str, topic_data: Dict = None) -> Dict:
|
||||
"""执行全面合规检查"""
|
||||
self.issues = []
|
||||
|
||||
# 1. 敏感词检查
|
||||
self._check_sensitive_words(text)
|
||||
|
||||
# 2. 平台规则检查
|
||||
self._check_platform_rules(text, platform)
|
||||
|
||||
# 3. 法律法规检查
|
||||
self._check_legal_compliance(text)
|
||||
|
||||
# 4. 品牌调性检查
|
||||
self._check_brand_guidelines(text)
|
||||
|
||||
# 5. 内容事实性检查（如有主题数据）
|
||||
if topic_data:
|
||||
self._check_factual_consistency(text, topic_data)
|
||||
|
||||
# 6. 最小字数检查
|
||||
self._check_min_length(text, platform)
|
||||
|
||||
# 7. 结构完整性检查（必须包含关键章节）
|
||||
self._check_required_sections(text)
|
||||
|
||||
return {
|
||||
"passed": len(self.issues) == 0,
|
||||
"issues": self.issues,
|
||||
"score": max(0, 100 - len(self.issues) * 10)
|
||||
}
|
||||
|
||||
def _check_sensitive_words(self, text: str):
|
||||
"""检查敏感词"""
|
||||
for category, words in SENSITIVE_WORDS.items():
|
||||
for word in words:
|
||||
if word in text:
|
||||
self.issues.append({
|
||||
"type": "敏感词",
|
||||
"category": category,
|
||||
"word": word,
|
||||
"suggestion": f"删除或替换'{word}'"
|
||||
})
|
||||
|
||||
def _check_platform_rules(self, text: str, platform: str):
|
||||
"""检查平台特定规则"""
|
||||
rules = PLATFORM_RULES.get(platform, {})
|
||||
|
||||
# 标题长度（从HTML中提取）
|
||||
title_match = re.search(r'<title>([^<]+)</title>', text) or re.search(r'<h1[^>]*>([^<]+)</h1>', text)
|
||||
if title_match and rules.get("max_title_len"):
|
||||
title_len = len(title_match.group(1))
|
||||
if title_len > rules["max_title_len"]:
|
||||
self.issues.append({
|
||||
"type": "平台规则",
|
||||
"category": "标题长度",
|
||||
"detail": f"标题{title_len}字，超过{platform}限制{rules['max_title_len']}字",
|
||||
"suggestion": "缩短标题"
|
||||
})
|
||||
|
||||
# 禁止的模式匹配
|
||||
for pattern in rules.get("forbidden_patterns", []):
|
||||
if re.search(pattern, text):
|
||||
self.issues.append({
|
||||
"type": "平台规则",
|
||||
"category": "禁止内容",
|
||||
"pattern": pattern,
|
||||
"suggestion": "移除违规内容或联系方式"
|
||||
})
|
||||
|
||||
# 标签检查（只匹配 #话题 格式，排除颜色码如 #1a1a1a）
|
||||
# 标签模式：#开头，后跟字母数字，长度2-10，不全是十六进制字符
|
||||
tags = re.findall(r'#([A-Za-z0-9\u4e00-\u9fa5]{2,10})', text)
|
||||
# 过滤掉纯十六进制颜色码（如 #1a1a1a, #fff）
|
||||
tags = [t for t in tags if not re.fullmatch(r'[0-9a-fA-F]{3,6}', t)]
|
||||
allowed = rules.get("allowed_tags", [])
|
||||
if allowed:
|
||||
for tag in tags:
|
||||
if tag not in allowed:
|
||||
self.issues.append({
|
||||
"type": "平台规则",
|
||||
"category": "标签合规",
|
||||
"tag": tag,
|
||||
"suggestion": f"使用平台允许的标签，如{', '.join(allowed[:3])}"
|
||||
})
|
||||
|
||||
def _check_legal_compliance(self, text: str):
|
||||
"""检查法律法规合规性"""
|
||||
# 检查是否涉及国家秘密、国家安全
|
||||
if re.search(r'国家机密|军事秘密|绝密|机密', text):
|
||||
self.issues.append({
|
||||
"type": "法律法规",
|
||||
"category": "国家秘密",
|
||||
"suggestion": "立即删除涉密内容"
|
||||
})
|
||||
|
||||
# 检查是否宣传迷信、邪教
|
||||
if re.search(r'算命|看相|测八字|跳大神|法轮功', text):
|
||||
self.issues.append({
|
||||
"type": "法律法规",
|
||||
"category": "封建迷信",
|
||||
"suggestion": "删除迷信内容"
|
||||
})
|
||||
|
||||
# 检查是否赌博相关
|
||||
if re.search(r'赌|博彩|下注|时时彩|六合彩', text):
|
||||
self.issues.append({
|
||||
"type": "法律法规",
|
||||
"category": "赌博违法",
|
||||
"suggestion": "删除赌博相关内容"
|
||||
})
|
||||
|
||||
# 检查版权问题（是否使用未授权素材）
|
||||
if re.search(r'版权声明.*?未经授权|转载请联系|盗用', text, re.IGNORECASE):
|
||||
self.issues.append({
|
||||
"type": "法律法规",
|
||||
"category": "版权风险",
|
||||
"suggestion": "确保所有引用已标注来源或获得授权"
|
||||
})
|
||||
|
||||
def _check_brand_guidelines(self, text: str):
|
||||
"""检查品牌调性（宇之然）"""
|
||||
# 检查是否使用第一人称"我"
|
||||
first_person_count = len(re.findall(r'^(我|本人|笔者)\b', text, re.MULTILINE))
|
||||
if first_person_count > 2: # 允许少量情感连接
|
||||
self.issues.append({
|
||||
"type": "品牌规范",
|
||||
"category": "人称使用",
|
||||
"detail": f"发现{first_person_count}处第一人称，建议使用客观叙事",
|
||||
"suggestion": "改为'实践者'、'本专栏'等客观表述"
|
||||
})
|
||||
|
||||
# 检查是否有商业推广倾向
|
||||
if re.search(r'强烈推荐|必买|最好的|最赚钱|独家', text):
|
||||
self.issues.append({
|
||||
"type": "品牌规范",
|
||||
"category": "过度推广",
|
||||
"suggestion": "使用更中立的表达，避免绝对化用语"
|
||||
})
|
||||
|
||||
# 检查是否提及具体品牌（需模糊化）
|
||||
known_brands = ["米家", "花帮主", "园艺助手", "Aerogarden"]
|
||||
for brand in known_brands:
|
||||
if brand in text:
|
||||
self.issues.append({
|
||||
"type": "品牌规范",
|
||||
"category": "品牌露出",
|
||||
"brand": brand,
|
||||
"suggestion": f"将'{brand}'改为'一些第三方工具'或'智能设备'"
|
||||
})
|
||||
|
||||
def _check_factual_consistency(self, text: str, topic_data: Dict):
|
||||
"""检查内容与选题的一致性"""
|
||||
topic = topic_data.get("topic", {})
|
||||
expected_title = topic.get("title", "")
|
||||
expected_field = topic.get("field", "")
|
||||
|
||||
# 检查标题是否出现在文章中
|
||||
if expected_title and expected_title[:5] not in text:
|
||||
self.issues.append({
|
||||
"type": "内容质量",
|
||||
"category": "主题一致性",
|
||||
"detail": f"文章可能偏离选题'{expected_title}'",
|
||||
"suggestion": "确认内容围绕选题展开"
|
||||
})
|
||||
|
||||
# 检查是否有核心观点
|
||||
core_concept = topic.get("core_concept", "")
|
||||
if core_concept and len(core_concept) > 10:
|
||||
# 核心概念应出现在前1/3内容
|
||||
first_third = text[:len(text)//3]
|
||||
if core_concept[:10] not in first_third:
|
||||
self.issues.append({
|
||||
"type": "内容质量",
|
||||
"category": "核心观点",
|
||||
"suggestion": "在文章前1/3部分明确阐述核心观点"
|
||||
})
|
||||
|
||||
def _check_min_length(self, text: str, platform: str):
|
||||
"""检查文章最小字数（去除HTML标签）"""
|
||||
# 简单去除HTML标签
|
||||
plain = re.sub(r'<[^>]+>', '', text)
|
||||
word_count = len(plain.strip())
|
||||
min_words = PLATFORM_RULES.get(platform, {}).get("min_word_count", 1000)
|
||||
if word_count < min_words:
|
||||
self.issues.append({
|
||||
"type": "内容完整度",
|
||||
"category": "字数不足",
|
||||
"detail": f"当前{word_count}字，低于平台要求{min_words}字",
|
||||
"suggestion": "扩写内容至最低要求"
|
||||
})
|
||||
|
||||
def _check_required_sections(self, text: str):
|
||||
"""检查是否包含必要章节（如引言、核心观点、总结等）"""
|
||||
required_headings = [
|
||||
"引言", "核心观点", "受众痛点", "总结", "行动指南"
|
||||
]
|
||||
missing = []
|
||||
for heading in required_headings:
|
||||
# 检查 h2 或 h3 中是否出现 heading
|
||||
if not re.search(r'<h[23][^>]*>.*' + re.escape(heading) + r'.*</h[23]>', text, re.IGNORECASE):
|
||||
missing.append(heading)
|
||||
if missing:
|
||||
self.issues.append({
|
||||
"type": "结构完整",
|
||||
"category": "章节缺失",
|
||||
"detail": f"缺少必要章节：{', '.join(missing)}",
|
||||
"suggestion": "补充缺失章节"
|
||||
})
|
||||
|
||||
def check_article(html_content: str, platform: str, topic_data: Dict = None) -> Dict:
|
||||
"""便捷函数：执行完整合规检查"""
|
||||
checker = ComplianceChecker()
|
||||
return checker.check_text(html_content, platform, topic_data)
|
||||
|
||||
if __name__ == "__main__":
|
||||
# 测试
|
||||
test_html = "<html><body><h1>测试</h1>内容涉及赌博网站</body></html>"
|
||||
result = check_article(test_html, "zhihu")
|
||||
print(json.dumps(result, ensure_ascii=False, indent=2))
|
||||
+138
-131
@@ -1,36 +1,23 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
合规审查与优化任务（数据库版）
|
||||
每天 05:45 运行，处理当天所有 draft 文章：
|
||||
1. 执行合规检查（compliance_checker）
|
||||
2. 自动修复已知问题（标题、标签）
|
||||
3. 重写合规版本
|
||||
4. 更新选题状态为「待发布」
|
||||
5. 生成优化报告通知
|
||||
"""
|
||||
|
||||
import json, datetime, logging, sys, re
|
||||
from pathlib import Path
|
||||
from typing import Dict, List
|
||||
from typing import Dict, List, Optional, Tuple
|
||||
from dataclasses import dataclass, asdict
|
||||
|
||||
PROJECT_ROOT = Path('/root/openclaw-workspace/projects/yu-zhi-ran')
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
from scripts.compliance_checker import check_article
|
||||
|
||||
# 导入 LLM 客户端（合规优化使用 NVIDIA）
|
||||
sys.path.insert(0, str(PROJECT_ROOT / "platform" / "backend"))
|
||||
|
||||
from scripts.compliance_checker import check_article
|
||||
try:
|
||||
from app.core.modelscope_client import call_llm
|
||||
from app.core.nvidia_client import call_llm
|
||||
HAVE_LLM = True
|
||||
except ImportError:
|
||||
HAVE_LLM = False
|
||||
|
||||
# 导入数据库辅助模块
|
||||
from db_helper import get_topic_by_id, update_topic_status
|
||||
from db_helper import get_topic_by_id, update_topic_status, get_active_llm_config, get_articles_by_topic, save_article
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
RELEASES_DIR = DATA_DIR / "releases"
|
||||
DRAFTS_DIR = DATA_DIR / "drafts"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
@@ -39,12 +26,19 @@ logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(
|
||||
handlers=[logging.FileHandler(LOGS_DIR / f"optimizer_{TODAY}.log"), logging.StreamHandler()])
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
# 平台白名单标签
|
||||
PLATFORM_TAGS = {
|
||||
"zhihu": ["科技", "职场"],
|
||||
"xiaohongshu": ["AI", "可持续", "生活方式"]
|
||||
}
|
||||
|
||||
_llm_config_cache = None
|
||||
|
||||
def get_llm_config():
|
||||
global _llm_config_cache
|
||||
if _llm_config_cache is None:
|
||||
_llm_config_cache = get_active_llm_config()
|
||||
return _llm_config_cache
|
||||
|
||||
@dataclass
|
||||
class OptimizationResult:
|
||||
file: str
|
||||
@@ -57,22 +51,40 @@ class OptimizationResult:
|
||||
status: str
|
||||
|
||||
def load_topic_map():
|
||||
"""从数据库加载所有选题数据"""
|
||||
from db_helper import export_topics_to_json
|
||||
topics = export_topics_to_json()
|
||||
return {t['id']: t for t in topics}
|
||||
|
||||
def update_topic_status_db_only(topic_id: str, status: str):
|
||||
"""仅更新数据库状态（不更新JSON）"""
|
||||
from db_helper import update_topic_status
|
||||
update_topic_status(topic_id, status)
|
||||
def get_articles_from_db(topic_ids: Optional[List[str]] = None) -> List[Tuple[str, str, str]]:
|
||||
"""从 articles 表读取 HTML 内容
|
||||
|
||||
Returns: [(html_content, platform, topic_id), ...]
|
||||
"""
|
||||
from db_helper import get_articles_by_topic
|
||||
results = []
|
||||
seen_topics = set()
|
||||
if topic_ids:
|
||||
for tid in topic_ids:
|
||||
articles = get_articles_by_topic(tid)
|
||||
for a in articles:
|
||||
if a.get("html_content"):
|
||||
results.append((a["html_content"], a["platform"], a["topic_id"]))
|
||||
seen_topics.add(a["topic_id"])
|
||||
else:
|
||||
from app.database import SessionLocal
|
||||
from app.models import Article
|
||||
db = SessionLocal()
|
||||
try:
|
||||
all_articles = db.query(Article).filter(Article.html_content.isnot(None)).all()
|
||||
for a in all_articles:
|
||||
results.append((a.html_content, a.platform, a.topic_id))
|
||||
finally:
|
||||
db.close()
|
||||
return results
|
||||
|
||||
def fix_wechat_title(html: str, title: str) -> str:
|
||||
"""微信标题优化：<title>和<h1>都控制长度（考虑后缀）"""
|
||||
suffix = f" - {TODAY} - 微信公众号"
|
||||
max_base_len = 32 - len(suffix) # <title> 中 base 部分允许的最大长度
|
||||
|
||||
# 处理 <title>...</title>
|
||||
max_base_len = 32 - len(suffix)
|
||||
title_tag = re.search(r'<title>([^<]+)</title>', html)
|
||||
if title_tag:
|
||||
full_title = title_tag.group(1)
|
||||
@@ -84,8 +96,6 @@ def fix_wechat_title(html: str, title: str) -> str:
|
||||
base = base[:max_base_len-3] + "..."
|
||||
new_full = base + suffix
|
||||
html = html.replace(full_title, new_full)
|
||||
|
||||
# 处理 <h1>...</h1>
|
||||
h1_match = re.search(r'<h1[^>]*>([^<]+)</h1>', html)
|
||||
if h1_match:
|
||||
current_h1 = h1_match.group(1)
|
||||
@@ -93,11 +103,9 @@ def fix_wechat_title(html: str, title: str) -> str:
|
||||
if len(base_h1) > 32:
|
||||
base_h1 = base_h1[:29] + "..."
|
||||
html = html.replace(current_h1, base_h1)
|
||||
|
||||
return html
|
||||
|
||||
def fix_tags(html: str, platform: str) -> str:
|
||||
"""强制替换标签为平台白名单"""
|
||||
if platform == "zhihu":
|
||||
tags_str = " ".join(f"#{t}" for t in PLATFORM_TAGS["zhihu"])
|
||||
if '<div class="tags">' in html:
|
||||
@@ -110,7 +118,31 @@ def fix_tags(html: str, platform: str) -> str:
|
||||
html = html.replace(f'<div class="hashtags">{old}</div>', f'<div class="hashtags">{tags_str}</div>')
|
||||
return html
|
||||
|
||||
def optimize_article(html: str, platform: str, topic_data: Dict) -> (str, List[str]):
|
||||
def polish_with_llm(html: str, platform: str) -> Tuple[str, Optional[str]]:
|
||||
"""用 LLM 优化文章内容，返回 (html, log_message_or_None)"""
|
||||
if not HAVE_LLM:
|
||||
return html, None
|
||||
llm_cfg = get_llm_config()
|
||||
try:
|
||||
model = llm_cfg.get('model') if llm_cfg else None
|
||||
temperature = llm_cfg.get('temperature', 0.5) if llm_cfg else 0.5
|
||||
max_tokens = llm_cfg.get('max_tokens', 4000) if llm_cfg else 4000
|
||||
system_prompt = llm_cfg.get('system_prompt') if llm_cfg else "你是一个专业的内容创作助手。"
|
||||
|
||||
polish_prompt = f"""你是一个专业的内容润色助手。请优化以下文章内容，提升表达的专业性和可读性，保持原文事实、数据、章节结构不变，输出相同的HTML格式（保留<h2>, <h3>, <p>标签）。
|
||||
|
||||
原文：
|
||||
{html}
|
||||
|
||||
优化后："""
|
||||
polished = call_llm(polish_prompt, model=model, temperature=temperature, max_tokens=max_tokens, system_prompt=system_prompt)
|
||||
if '<h2' in polished or '<p>' in polished:
|
||||
return polished, "LLM 内容优化"
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM 优化失败: {e}")
|
||||
return html, None
|
||||
|
||||
def optimize_article(html: str, platform: str, topic_data: Dict) -> Tuple[str, List[str]]:
|
||||
logs = []
|
||||
if platform == "wechat":
|
||||
html = fix_wechat_title(html, topic_data.get("title", ""))
|
||||
@@ -127,124 +159,99 @@ def optimize_article(html: str, platform: str, topic_data: Dict) -> (str, List[s
|
||||
src = m.group(1)
|
||||
if not src.startswith('data:image/'):
|
||||
logs.append(f"图片未内联: {src[:50]}... 需手动修复")
|
||||
|
||||
if HAVE_LLM:
|
||||
try:
|
||||
polish_prompt = f"""你是一个专业的内容润色助手。请优化以下文章内容，提升表达的专业性和可读性，保持原文事实、数据、章节结构不变，输出相同的HTML格式（保留<h2>, <h3>, <p>标签）。
|
||||
|
||||
原文：
|
||||
{html}
|
||||
|
||||
优化后："""
|
||||
polished = call_llm(polish_prompt, temperature=0.5, max_tokens=4000)
|
||||
if '<h2' in polished or '<p>' in polished:
|
||||
html = polished
|
||||
logs.append("LLM 内容优化（NVIDIA）")
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM 优化失败: {e}")
|
||||
polished, pol_log = polish_with_llm(html, platform)
|
||||
if pol_log:
|
||||
html = polished
|
||||
logs.append(pol_log)
|
||||
return html, logs
|
||||
|
||||
def main(topic_ids: List[str] = None):
|
||||
logger.info("=== 合规审查与优化开始 ===")
|
||||
release_dir = RELEASES_DIR / TODAY
|
||||
if not release_dir.exists():
|
||||
logger.warning(f"今日发布目录不存在: {release_dir}")
|
||||
return
|
||||
llm_cfg = get_llm_config()
|
||||
if llm_cfg:
|
||||
logger.info(f"LLM 配置: {llm_cfg['name']} (model={llm_cfg['model']})")
|
||||
else:
|
||||
logger.info("LLM 配置: 使用环境变量默认值")
|
||||
|
||||
articles = get_articles_from_db(topic_ids)
|
||||
if not articles:
|
||||
logger.warning("未找到任何文章（可能尚未创作或同步到 DB）")
|
||||
report_file = DRAFTS_DIR / TODAY / "optimization_report.json"
|
||||
report_file.parent.mkdir(parents=True, exist_ok=True)
|
||||
with open(report_file, 'w', encoding='utf-8') as f:
|
||||
json.dump({
|
||||
"date": TODAY,
|
||||
"summary": {"total_articles": 0, "passed_auto": 0, "need_manual": 0, "average_score": 0},
|
||||
"details": [], "all_passed": True
|
||||
}, f, ensure_ascii=False, indent=2)
|
||||
print("OPTIMIZATION_COMPLETE: 0 articles found")
|
||||
sys.exit(0)
|
||||
|
||||
topic_map = load_topic_map()
|
||||
results = []
|
||||
all_passed = True
|
||||
|
||||
for platform_dir in ["zhihu", "wechat", "xiaohongshu"]:
|
||||
platform_path = release_dir / platform_dir
|
||||
if not platform_path.exists():
|
||||
for html, platform_dir, topic_id in articles:
|
||||
topic_data = topic_map.get(topic_id)
|
||||
if not topic_data:
|
||||
logger.warning(f"未找到选题: {topic_id}")
|
||||
continue
|
||||
for html_file in platform_path.glob("*.html"):
|
||||
stem = html_file.stem
|
||||
parts = stem.split('_')
|
||||
if len(parts) < 2:
|
||||
continue
|
||||
topic_id = parts[1]
|
||||
if topic_ids is not None and topic_id not in topic_ids:
|
||||
continue
|
||||
topic_data = topic_map.get(topic_id)
|
||||
if not topic_data:
|
||||
logger.warning(f"未找到选题: {topic_id}")
|
||||
continue
|
||||
|
||||
html = html_file.read_text(encoding='utf-8')
|
||||
check_result = check_article(html, platform_dir, topic_data)
|
||||
issues = check_result['issues']
|
||||
score = check_result['score']
|
||||
check_result = check_article(html, platform_dir, topic_data)
|
||||
issues = check_result['issues']
|
||||
score = check_result['score']
|
||||
label = f"{platform_dir}/{topic_id}"
|
||||
|
||||
if any(issue['type'] == '平台规则' for issue in issues):
|
||||
optimized_html, opt_logs = optimize_article(html, platform_dir, topic_data)
|
||||
recheck = check_article(optimized_html, platform_dir, topic_data)
|
||||
if recheck['passed']:
|
||||
html_file.write_text(optimized_html, encoding='utf-8')
|
||||
logger.info(f"✅ {html_file.name} 已优化并通过合规检查")
|
||||
results.append(OptimizationResult(
|
||||
file=str(html_file.relative_to(PROJECT_ROOT)),
|
||||
platform=platform_dir,
|
||||
topic_id=topic_id,
|
||||
title=topic_data.get('title',''),
|
||||
original_issues=len(issues),
|
||||
fixed_issues=len(issues) - len(recheck['issues']),
|
||||
final_score=recheck['score'],
|
||||
status="passed"
|
||||
))
|
||||
update_topic_status_db_only(topic_id, 'pending')
|
||||
else:
|
||||
logger.warning(f"⚠️ {html_file.name} 优化后仍有问题，需人工审核")
|
||||
results.append(OptimizationResult(
|
||||
file=str(html_file.relative_to(PROJECT_ROOT)),
|
||||
platform=platform_dir,
|
||||
topic_id=topic_id,
|
||||
title=topic_data.get('title',''),
|
||||
original_issues=len(issues),
|
||||
fixed_issues=len(issues) - len(recheck['issues']),
|
||||
final_score=recheck['score'],
|
||||
status="manual_review"
|
||||
))
|
||||
all_passed = False
|
||||
else:
|
||||
if issues:
|
||||
optimized_html, opt_logs = optimize_article(html, platform_dir, topic_data)
|
||||
recheck = check_article(optimized_html, platform_dir, topic_data)
|
||||
if recheck['passed']:
|
||||
save_article(topic_id, platform_dir, optimized_html)
|
||||
logger.info(f"✅ {label} 已修复并通过审查 ({len(issues)} issues fixed)")
|
||||
results.append(OptimizationResult(
|
||||
file=str(html_file.relative_to(PROJECT_ROOT)),
|
||||
file=f"db:{platform_dir}_{topic_id}",
|
||||
platform=platform_dir,
|
||||
topic_id=topic_id,
|
||||
title=topic_data.get('title',''),
|
||||
original_issues=0,
|
||||
fixed_issues=0,
|
||||
final_score=score,
|
||||
status="passed" if check_result['passed'] else "manual_review"
|
||||
original_issues=len(issues),
|
||||
fixed_issues=len(issues) - len(recheck['issues']),
|
||||
final_score=recheck['score'],
|
||||
status="passed"
|
||||
))
|
||||
if not check_result['passed']:
|
||||
all_passed = False
|
||||
else:
|
||||
logger.warning(f"⚠️ {label} 仍有 {len(recheck['issues'])} 个问题需人工处理")
|
||||
results.append(OptimizationResult(
|
||||
file=f"db:{platform_dir}_{topic_id}",
|
||||
platform=platform_dir,
|
||||
topic_id=topic_id,
|
||||
title=topic_data.get('title',''),
|
||||
original_issues=len(issues),
|
||||
fixed_issues=len(issues) - len(recheck['issues']),
|
||||
final_score=recheck['score'],
|
||||
status="manual_review"
|
||||
))
|
||||
all_passed = False
|
||||
else:
|
||||
results.append(OptimizationResult(
|
||||
file=f"db:{platform_dir}_{topic_id}",
|
||||
platform=platform_dir,
|
||||
topic_id=topic_id,
|
||||
title=topic_data.get('title',''),
|
||||
original_issues=0,
|
||||
fixed_issues=0,
|
||||
final_score=score,
|
||||
status="passed"
|
||||
))
|
||||
logger.info(f"✅ {label} 合规检查通过 ({score}分)")
|
||||
|
||||
# 更新选题状态 (JSON + 数据库)
|
||||
passed_ids = set()
|
||||
for res in results:
|
||||
if res.status == "passed":
|
||||
tid = res.topic_id
|
||||
# 更新数据库状态为 'pending'（待发布）
|
||||
update_topic_status(tid, 'ready')
|
||||
# 可选：同时更新 JSON 以保持兼容
|
||||
# (已废弃，但保留更新，避免其他组件出错)
|
||||
try:
|
||||
json_path = DATA_DIR / "sustainability_topics.json"
|
||||
with open(json_path, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
for t in topics:
|
||||
if t.get('id') == tid:
|
||||
t['status'] = 'ready'
|
||||
t['ready_at'] = TODAY
|
||||
t['compliance_score'] = res.final_score
|
||||
break
|
||||
with open(json_path, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
except Exception as e:
|
||||
logger.warning(f"更新 JSON 失败: {e}")
|
||||
passed_ids.add(res.topic_id)
|
||||
for tid in passed_ids:
|
||||
update_topic_status(tid, 'ready')
|
||||
logger.info(f"选题 {tid} 状态 → ready（待发布）")
|
||||
|
||||
# 生成报告
|
||||
report_file = DRAFTS_DIR / TODAY / "optimization_report.json"
|
||||
report_file.parent.mkdir(parents=True, exist_ok=True)
|
||||
report = {
|
||||
@@ -261,7 +268,7 @@ def main(topic_ids: List[str] = None):
|
||||
with open(report_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(report, f, ensure_ascii=False, indent=2)
|
||||
|
||||
logger.info(f"✅ 合规优化完成: {len(results)} 篇文章, {sum(1 for r in results if r.status=='passed')} 篇自动通过")
|
||||
logger.info(f"✅ 合规审查完成: {len(results)} 篇文章, {sum(1 for r in results if r.status=='passed')} 篇通过")
|
||||
print(f"OPTIMIZATION_COMPLETE: {len(results)} articles, {sum(1 for r in results if r.status=='passed')} passed, {sum(1 for r in results if r.status=='manual_review')} need manual review")
|
||||
sys.exit(0)
|
||||
|
||||
|
||||
+12
-9
@@ -14,7 +14,6 @@ sys.path.insert(0, str(PROJECT_ROOT))
|
||||
from db_helper import get_topic_by_id, get_next_topic, update_topic_status
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TOPICS_FILE = DATA_DIR / "sustainability_topics.json" # 保留用于备份
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
@@ -29,26 +28,30 @@ logging.basicConfig(
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
def select_next_topic(topic_id: str = None) -> Dict:
|
||||
"""选择并锁定要创作的选题（从数据库）"""
|
||||
"""选择并锁定要创作的选题（趋势引擎匹配 → 回退优先级）"""
|
||||
if topic_id:
|
||||
# 指定ID，查询数据库
|
||||
topic = get_topic_by_id(topic_id)
|
||||
if not topic:
|
||||
raise ValueError(f"Topic {topic_id} not found")
|
||||
# 检查状态：禁止已发布状态重新创作
|
||||
current_status = topic.get('status')
|
||||
if current_status in ['已发布', 'published']:
|
||||
raise ValueError(f"Topic {topic_id} is already published, cannot recreate")
|
||||
# 更新状态为「审查中」表示已经开始处理
|
||||
update_topic_status(topic_id, 'review')
|
||||
return topic
|
||||
|
||||
# 自动选择：下一个待处理的选题
|
||||
|
||||
try:
|
||||
from topic_selector import select_best_topic as engine_select
|
||||
topic = engine_select()
|
||||
if topic:
|
||||
logger.info(f"选题引擎推荐: {topic['id']} {topic['title']}")
|
||||
update_topic_status(topic['id'], 'review')
|
||||
return topic
|
||||
except Exception as e:
|
||||
logger.warning(f"选题引擎失效，回退简单策略: {e}")
|
||||
|
||||
topic = get_next_topic(priority='高') or get_next_topic()
|
||||
if not topic:
|
||||
raise ValueError("No available topics to create (all locked or wrong status)")
|
||||
|
||||
# 更新状态为「审查中」表示已锁定
|
||||
update_topic_status(topic['id'], 'review')
|
||||
return topic
|
||||
|
||||
|
||||
@@ -1,208 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
内容创作脚本（修复版）
|
||||
支持：标题长度限制、标签合规、状态流程
|
||||
"""
|
||||
|
||||
import os, sys, yaml, json, datetime, logging, random
|
||||
from pathlib import Path
|
||||
from typing import Dict, List
|
||||
import subprocess
|
||||
from dataclasses import dataclass, asdict
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
from scripts.image_generator import ImageGenerator
|
||||
|
||||
CONFIG_DIR = PROJECT_ROOT / "config"
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TEMPLATES_DIR = PROJECT_ROOT / "automation" / "templates"
|
||||
IMAGES_DIR = PROJECT_ROOT / "automation" / "images"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s',
|
||||
handlers=[logging.FileHandler(LOGS_DIR / f"creator_{TODAY}.log"), logging.StreamHandler()])
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
@dataclass
|
||||
class ContentArticle:
|
||||
id: str
|
||||
topic_id: str
|
||||
title: str
|
||||
platform: str
|
||||
content: str
|
||||
image_paths: List[str]
|
||||
metadata: dict
|
||||
created_date: str
|
||||
output_dir: str
|
||||
status: str = "draft" # draft, pending_review, ready_for_publish, published
|
||||
|
||||
class ContentCreator:
|
||||
def __init__(self):
|
||||
self.articles = []
|
||||
self.release_dir = DATA_DIR / "releases" / TODAY
|
||||
self.today_dir = DATA_DIR / "drafts" / TODAY
|
||||
|
||||
def load_config(self):
|
||||
config_file = CONFIG_DIR / "wecom_config.yaml"
|
||||
if not config_file.exists():
|
||||
logger.warning("配置文件不存在，使用默认")
|
||||
self.wecom_config = {"content_rules": {}}
|
||||
return
|
||||
with open(config_file, 'r', encoding='utf-8') as f:
|
||||
self.wecom_config = yaml.safe_load(f)
|
||||
logger.info("配置加载完成")
|
||||
|
||||
def select_topic_for_today(self):
|
||||
topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
if not topics_file.exists():
|
||||
logger.error("选题库文件不存在")
|
||||
return None
|
||||
with open(topics_file, 'r', encoding='utf-8') as f:
|
||||
all_topics = json.load(f)
|
||||
available = [t for t in all_topics if t.get("status") != "已发布" and t.get("status") != "待发布"]
|
||||
if not available:
|
||||
logger.warning("没有可选选题")
|
||||
return None
|
||||
selected = max(available, key=lambda t: t.get("priority_score", 0))
|
||||
logger.info(f"选择了选题: {selected.get('title')} (优先级: {selected.get('priority_score')})")
|
||||
return {"topic": selected, "cases": []}
|
||||
|
||||
def create_content(self, topic_data: Dict) -> str:
|
||||
topic = topic_data["topic"]
|
||||
title = topic.get("title", "")
|
||||
sections = [
|
||||
f"<h2>{title}</h2>",
|
||||
f"<p>今天是{TODAY}，我们探讨「{title}」。根据全球案例与本土分析，给出以下建议：</p>",
|
||||
"<h3>核心观点</h3><p>" + topic.get("core_concept", "待补充") + "</p>",
|
||||
"<h3>目标受众痛点</h3><p>" + topic.get("audience_pain", "待补充") + "</p>",
|
||||
"<h3>本土化方案</h3><ul><li>" + topic.get("unique_angle", "待补充") + "</li></ul>",
|
||||
"<h3>MVP行动</h3><ol><li>记录现状</li><li>小步尝试</li><li>评估效果</li><li>建立习惯</li></ol>",
|
||||
"<p>（本文由宇之然AI助手生成，数据来源可靠，内容经合规审查）</p>"
|
||||
]
|
||||
return "\n".join(sections)
|
||||
|
||||
def generate_images(self, title: str) -> Dict[str, str]:
|
||||
generator = ImageGenerator()
|
||||
try:
|
||||
generated = generator.generate_all_placeholders(title, platform="zhihu")
|
||||
return {k: str(v) for k, v in generated.items()}
|
||||
except Exception as e:
|
||||
logger.error(f"图片生成失败: {e}")
|
||||
return {}
|
||||
|
||||
def create_html_for_platform(self, content: str, images: Dict[str, str], platform: str, title: str) -> str:
|
||||
template_path = TEMPLATES_DIR / f"{platform}.html"
|
||||
if template_path.exists():
|
||||
with open(template_path, 'r', encoding='utf-8') as f:
|
||||
template = f.read()
|
||||
else:
|
||||
template = "<!DOCTYPE html><html><body>{{TITLE}}<hr><!-- CONTENT --></body></html>"
|
||||
|
||||
# 插入图片标记
|
||||
for marker, path in images.items():
|
||||
if path:
|
||||
img_tag = f'<img src="{path}" alt="{marker}">'
|
||||
content = content.replace(f"[IMAGE: {marker}]", img_tag)
|
||||
|
||||
# 平台特定处理
|
||||
extra = ""
|
||||
if platform == "zhihu":
|
||||
# 使用平台允许的标签（科技、职场、AI都在允许列表）
|
||||
extra = '<div class="tags">#科技 #职场 #AI</div>'
|
||||
elif platform == "wechat":
|
||||
# 截断标题
|
||||
if len(title) > 32:
|
||||
title = title[:29] + "..."
|
||||
abstract = content[:100] + "..."
|
||||
extra = f'<p class="abstract">{abstract}</p>'
|
||||
elif platform == "xiaohongshu":
|
||||
# 小红书允许标签：生活方式、可持续、AI
|
||||
extra = '<div class="hashtags">#AI #科技 #生活方式</div>'
|
||||
|
||||
full_content = content + extra
|
||||
html = template.replace("<!-- CONTENT -->", full_content)
|
||||
html = html.replace("{{DATE}}", TODAY)
|
||||
html = html.replace("{{TITLE}}", title)
|
||||
return html
|
||||
|
||||
def mark_topic_ready(self, topic_id: str):
|
||||
"""标记选题为「待发布」(审查通过)"""
|
||||
topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
with open(topics_file, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
for t in topics:
|
||||
if t.get("id") == topic_id:
|
||||
t["status"] = "待发布"
|
||||
t["ready_at"] = TODAY
|
||||
break
|
||||
with open(topics_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
logger.info(f"选题 {topic_id} 已标记为「待发布」")
|
||||
|
||||
def run(self):
|
||||
logger.info("开始内容创作")
|
||||
self.load_config()
|
||||
topic_data = self.select_topic_for_today()
|
||||
if not topic_data:
|
||||
logger.error("未能选择选题，任务结束")
|
||||
return False
|
||||
|
||||
content = self.create_content(topic_data)
|
||||
images = self.generate_images(topic_data["topic"].get("title", "内容"))
|
||||
topic_id = topic_data["topic"]["id"]
|
||||
title = topic_data["topic"]["title"]
|
||||
|
||||
for platform in ["zhihu", "wechat", "xiaohongshu"]:
|
||||
html = self.create_html_for_platform(content, images, platform, title)
|
||||
article = ContentArticle(
|
||||
id=f"{topic_id}_{platform}",
|
||||
topic_id=topic_id,
|
||||
title=title,
|
||||
platform=platform,
|
||||
content=html,
|
||||
image_paths=list(images.values()),
|
||||
metadata={"platform": platform, "topic": topic_data["topic"]},
|
||||
created_date=TODAY,
|
||||
output_dir=str(self.release_dir / platform),
|
||||
status="draft"
|
||||
)
|
||||
self.save_article(article)
|
||||
self.articles.append(article)
|
||||
|
||||
# 标记为待发布（而不是已发布）
|
||||
self.mark_topic_ready(topic_id)
|
||||
|
||||
# 发送通知（可选）
|
||||
logger.info(f"创作完成: {len(self.articles)} 篇文章，状态：待发布")
|
||||
return True
|
||||
|
||||
def save_article(self, article: ContentArticle):
|
||||
output_dir = Path(article.output_dir)
|
||||
output_dir.mkdir(parents=True, exist_ok=True)
|
||||
html_file = output_dir / f"{article.platform}_{article.id}.html"
|
||||
with open(html_file, 'w', encoding='utf-8') as f:
|
||||
f.write(article.content)
|
||||
meta_file = output_dir / f"{article.platform}_{article.id}.json"
|
||||
with open(meta_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(asdict(article), f, ensure_ascii=False, indent=2)
|
||||
logger.info(f"保存了 {article.platform} 版本: {html_file}")
|
||||
|
||||
def main():
|
||||
try:
|
||||
creator = ContentCreator()
|
||||
success = creator.run()
|
||||
if success:
|
||||
print(f"SUCCESS: Created {len(creator.articles)} articles for {TODAY} (status: 待发布)")
|
||||
sys.exit(0)
|
||||
else:
|
||||
print("WARNING: Content creation failed")
|
||||
sys.exit(1)
|
||||
except Exception as e:
|
||||
logger.error(f"创作任务失败: {e}")
|
||||
print(f"ERROR: {e}")
|
||||
sys.exit(1)
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,231 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
内容创作脚本（最终修复版）
|
||||
- 标题长度：考虑模板后缀，整体限制在32字内（微信）
|
||||
- 标签按领域动态映射（知乎、小红书）
|
||||
- 状态：生成后为「待发布」，人工发布后手动改为「已发布」
|
||||
"""
|
||||
|
||||
import os, sys, yaml, json, datetime, logging
|
||||
from pathlib import Path
|
||||
from typing import Dict, List
|
||||
from dataclasses import dataclass, asdict
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
from scripts.image_generator import ImageGenerator
|
||||
|
||||
CONFIG_DIR = PROJECT_ROOT / "config"
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TEMPLATES_DIR = PROJECT_ROOT / "automation" / "templates"
|
||||
IMAGES_DIR = PROJECT_ROOT / "automation" / "images"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s',
|
||||
handlers=[logging.FileHandler(LOGS_DIR / f"creator_{TODAY}.log"), logging.StreamHandler()])
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
@dataclass
|
||||
class ContentArticle:
|
||||
id: str
|
||||
topic_id: str
|
||||
title: str
|
||||
platform: str
|
||||
content: str
|
||||
image_paths: List[str]
|
||||
metadata: dict
|
||||
created_date: str
|
||||
output_dir: str
|
||||
status: str = "draft" # draft → pending_review → ready_for_publish → published
|
||||
|
||||
class ContentCreator:
|
||||
def __init__(self):
|
||||
self.articles = []
|
||||
self.release_dir = DATA_DIR / "releases" / TODAY
|
||||
self.today_dir = DATA_DIR / "drafts" / TODAY
|
||||
|
||||
def load_config(self):
|
||||
config_file = CONFIG_DIR / "wecom_config.yaml"
|
||||
if config_file.exists():
|
||||
with open(config_file, 'r', encoding='utf-8') as f:
|
||||
self.wecom_config = yaml.safe_load(f)
|
||||
else:
|
||||
self.wecom_config = {"content_rules": {}}
|
||||
logger.info("配置加载完成")
|
||||
|
||||
def select_topic_for_today(self):
|
||||
topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
with open(topics_file, 'r', encoding='utf-8') as f:
|
||||
all_topics = json.load(f)
|
||||
available = [t for t in all_topics if t.get("status") not in ["已发布", "待发布"]]
|
||||
if not available:
|
||||
logger.warning("没有可选选题")
|
||||
return None
|
||||
selected = max(available, key=lambda t: t.get("priority_score", 0))
|
||||
logger.info(f"选择了选题: {selected.get('title')} (优先级: {selected.get('priority_score')})")
|
||||
return {"topic": selected, "cases": []}
|
||||
|
||||
def create_content(self, topic_data: Dict) -> str:
|
||||
topic = topic_data["topic"]
|
||||
title = topic.get("title", "")
|
||||
# 使用真实字段构建内容
|
||||
sections = [
|
||||
f"<h2>{title}</h2>",
|
||||
f"<p>今天是{TODAY}，我们探讨「{title}」。基于全球案例与本土实践，提供可执行的建议。</p>",
|
||||
"<h3>核心观点</h3><p>" + topic.get("core_concept", "待补充") + "</p>",
|
||||
"<h3>目标受众痛点</h3><p>" + topic.get("audience_pain", "待补充") + "</p>",
|
||||
"<h3>独特视角</h3><p>" + topic.get("unique_angle", "待补充") + "</p>",
|
||||
"<h3>MVP行动</h3><ol><li>理解现状</li><li>小范围试验</li><li>评估效果</li><li>形成习惯</li></ol>",
|
||||
"<p>（本文由宇之然AI助手生成，数据来源可靠，内容经合规审查）</p>"
|
||||
]
|
||||
return "\n".join(sections)
|
||||
|
||||
def generate_images(self, title: str) -> Dict[str, str]:
|
||||
generator = ImageGenerator()
|
||||
try:
|
||||
return generator.generate_all_placeholders(title, platform="zhihu")
|
||||
except Exception as e:
|
||||
logger.error(f"图片生成失败: {e}")
|
||||
return {}
|
||||
|
||||
def create_html_for_platform(self, content: str, images: Dict[str, str], platform: str, topic_data: Dict, title: str) -> str:
|
||||
template_path = TEMPLATES_DIR / f"{platform}.html"
|
||||
if template_path.exists():
|
||||
with open(template_path, 'r', encoding='utf-8') as f:
|
||||
template = f.read()
|
||||
else:
|
||||
template = "<!DOCTYPE html><html><body>{{TITLE}}<hr><!-- CONTENT --></body></html>"
|
||||
|
||||
# 插入图片（将 Path 转换为字符串）
|
||||
for marker, path in images.items():
|
||||
if path:
|
||||
img_src = str(path)
|
||||
content = content.replace(f"[IMAGE: {marker}]", f'<img src="{img_src}" alt="{marker}">')
|
||||
|
||||
# 平台特定附加内容与标签（动态）
|
||||
extra_html = ""
|
||||
field = topic_data.get("topic", {}).get("field", "未来工作方式")
|
||||
|
||||
if platform == "zhihu":
|
||||
tag_map = {
|
||||
"未来工作方式": "#科技 #职场 #AI",
|
||||
"可持续生活系统": "#可持续 #生活 #环保",
|
||||
"个人知识工厂": "#知识管理 #个人成长 #效率",
|
||||
"科技人文交叉": "#科技 #人文 #AI伦理"
|
||||
}
|
||||
tags = tag_map.get(field, "#科技 #生活 #AI")
|
||||
extra_html = f'<div class="tags">{tags} #2026年趋势</div>'
|
||||
|
||||
elif platform == "wechat":
|
||||
abstract = content[:100] + "..."
|
||||
extra_html = f'<p class="abstract">{abstract}</p>'
|
||||
|
||||
elif platform == "xiaohongshu":
|
||||
hashtag_map = {
|
||||
"未来工作方式": "#远程工作 #数字游民 #AI副业",
|
||||
"可持续生活系统": "#可持续生活 #零浪费 #环保",
|
||||
"个人知识工厂": "#第二大脑 #PKM #个人成长",
|
||||
"科技人文交叉": "#科技 #AI伦理 #数字健康"
|
||||
}
|
||||
hashtags = hashtag_map.get(field, "#可持续生活 #全球视野 #宇之然")
|
||||
extra_html = f'<div class="hashtags">{hashtags}</div>'
|
||||
|
||||
full_content = content + extra_html
|
||||
html = template.replace("<!-- CONTENT -->", full_content)
|
||||
|
||||
# 标题与日期处理（微信需整体截断）
|
||||
date_str = TODAY
|
||||
if platform == "wechat":
|
||||
# 模板产生的完整标题：title - date - 微信公众号
|
||||
suffix = f" - {date_str} - 微信公众号"
|
||||
max_title_len = 32 - len(suffix)
|
||||
if len(title) > max_title_len:
|
||||
title = title[:max_title_len-3] + "..."
|
||||
full_title = title + suffix
|
||||
else:
|
||||
full_title = title
|
||||
|
||||
html = html.replace("{{DATE}}", date_str)
|
||||
html = html.replace("{{TITLE}}", full_title)
|
||||
return html
|
||||
|
||||
def mark_topic_ready(self, topic_id: str):
|
||||
"""标记选题为「待发布」(审查通过)"""
|
||||
topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
with open(topics_file, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
for t in topics:
|
||||
if t.get("id") == topic_id:
|
||||
t["status"] = "待发布"
|
||||
t["ready_at"] = TODAY
|
||||
break
|
||||
with open(topics_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
logger.info(f"选题 {topic_id} 已标记为「待发布」")
|
||||
|
||||
def run(self):
|
||||
logger.info("开始内容创作")
|
||||
self.load_config()
|
||||
topic_data = self.select_topic_for_today()
|
||||
if not topic_data:
|
||||
logger.error("未能选择选题，任务结束")
|
||||
return False
|
||||
|
||||
content = self.create_content(topic_data)
|
||||
images = self.generate_images(topic_data["topic"].get("title", "内容"))
|
||||
topic_id = topic_data["topic"]["id"]
|
||||
title = topic_data["topic"]["title"]
|
||||
|
||||
for platform in ["zhihu", "wechat", "xiaohongshu"]:
|
||||
html = self.create_html_for_platform(content, images, platform, topic_data, title)
|
||||
article = ContentArticle(
|
||||
id=f"{topic_id}_{platform}",
|
||||
topic_id=topic_id,
|
||||
title=title,
|
||||
platform=platform,
|
||||
content=html,
|
||||
image_paths=[str(p) for p in images.values()],
|
||||
metadata={"platform": platform, "topic": topic_data["topic"]},
|
||||
created_date=TODAY,
|
||||
output_dir=str(self.release_dir / platform),
|
||||
status="pending_review" # 生成后待审查
|
||||
)
|
||||
self.save_article(article)
|
||||
self.articles.append(article)
|
||||
|
||||
# 标记选题为待发布（审查通过）
|
||||
self.mark_topic_ready(topic_id)
|
||||
logger.info(f"创作完成: {len(self.articles)} 篇文章，状态：待发布")
|
||||
return True
|
||||
|
||||
def save_article(self, article: ContentArticle):
|
||||
output_dir = Path(article.output_dir)
|
||||
output_dir.mkdir(parents=True, exist_ok=True)
|
||||
html_file = output_dir / f"{article.platform}_{article.id}.html"
|
||||
with open(html_file, 'w', encoding='utf-8') as f:
|
||||
f.write(article.content)
|
||||
meta_file = output_dir / f"{article.platform}_{article.id}.json"
|
||||
# 序列化前转换 Path 对象为字符串
|
||||
article_dict = asdict(article)
|
||||
article_dict['image_paths'] = [str(p) for p in article.image_paths]
|
||||
json.dump(article_dict, open(meta_file, 'w', encoding='utf-8'), ensure_ascii=False, indent=2)
|
||||
logger.info(f"保存了 {article.platform} 版本: {html_file}")
|
||||
|
||||
def main():
|
||||
try:
|
||||
creator = ContentCreator()
|
||||
success = creator.run()
|
||||
if success:
|
||||
print(f"SUCCESS: Created {len(creator.articles)} articles for {TODAY} (status: 待发布)")
|
||||
sys.exit(0)
|
||||
else:
|
||||
print("WARNING: Content creation failed")
|
||||
sys.exit(1)
|
||||
except Exception as e:
|
||||
logger.error(f"创作任务失败: {e}")
|
||||
print(f"ERROR: {e}")
|
||||
sys.exit(1)
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,208 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
内容创作脚本（最终修复版 v3）
|
||||
- 知乎标签：严格使用 #科技 #职场
|
||||
- 小红书标签：严格使用 #AI #可持续 #生活方式
|
||||
- 微信标题截断：整体长度≤32字
|
||||
"""
|
||||
|
||||
import os, sys, yaml, json, datetime, logging
|
||||
from pathlib import Path
|
||||
from typing import Dict, List
|
||||
from dataclasses import dataclass, asdict
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
from scripts.image_generator import ImageGenerator
|
||||
|
||||
CONFIG_DIR = PROJECT_ROOT / "config"
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TEMPLATES_DIR = PROJECT_ROOT / "automation" / "templates"
|
||||
IMAGES_DIR = PROJECT_ROOT / "automation" / "images"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s',
|
||||
handlers=[logging.FileHandler(LOGS_DIR / f"creator_{TODAY}.log"), logging.StreamHandler()])
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
@dataclass
|
||||
class ContentArticle:
|
||||
id: str
|
||||
topic_id: str
|
||||
title: str
|
||||
platform: str
|
||||
content: str
|
||||
image_paths: List[str]
|
||||
metadata: dict
|
||||
created_date: str
|
||||
output_dir: str
|
||||
status: str = "pending_review"
|
||||
|
||||
class ContentCreator:
|
||||
def __init__(self):
|
||||
self.articles = []
|
||||
self.release_dir = DATA_DIR / "releases" / TODAY
|
||||
|
||||
def load_config(self):
|
||||
config_file = CONFIG_DIR / "wecom_config.yaml"
|
||||
if config_file.exists():
|
||||
with open(config_file, 'r', encoding='utf-8') as f:
|
||||
self.wecom_config = yaml.safe_load(f)
|
||||
else:
|
||||
self.wecom_config = {"content_rules": {}}
|
||||
logger.info("配置加载完成")
|
||||
|
||||
def select_topic_for_today(self):
|
||||
topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
with open(topics_file, 'r', encoding='utf-8') as f:
|
||||
all_topics = json.load(f)
|
||||
available = [t for t in all_topics if t.get("status") not in ["已发布", "待发布"]]
|
||||
if not available:
|
||||
logger.warning("没有可选选题")
|
||||
return None
|
||||
selected = max(available, key=lambda t: t.get("priority_score", 0))
|
||||
logger.info(f"选择了选题: {selected.get('title')} (优先级: {selected.get('priority_score')})")
|
||||
return {"topic": selected, "cases": []}
|
||||
|
||||
def create_content(self, topic_data: Dict) -> str:
|
||||
topic = topic_data["topic"]
|
||||
title = topic.get("title", "")
|
||||
sections = [
|
||||
f"<h2>{title}</h2>",
|
||||
f"<p>今天是{TODAY}，我们探讨「{title}」。基于全球案例与本土实践，提供可执行的建议。</p>",
|
||||
"<h3>核心观点</h3><p>" + topic.get("core_concept", "待补充") + "</p>",
|
||||
"<h3>目标受众痛点</h3><p>" + topic.get("audience_pain", "待补充") + "</p>",
|
||||
"<h3>独特视角</h3><p>" + topic.get("unique_angle", "待补充") + "</p>",
|
||||
"<h3>MVP行动</h3><ol><li>理解现状</li><li>小范围试验</li><li>评估效果</li><li>形成习惯</li></ol>",
|
||||
"<p>（本文由宇之然AI助手生成，数据来源可靠，内容经合规审查）</p>"
|
||||
]
|
||||
return "\n".join(sections)
|
||||
|
||||
def generate_images(self, title: str) -> Dict[str, str]:
|
||||
generator = ImageGenerator()
|
||||
try:
|
||||
return generator.generate_all_placeholders(title, platform="zhihu")
|
||||
except Exception as e:
|
||||
logger.error(f"图片生成失败: {e}")
|
||||
return {}
|
||||
|
||||
def create_html_for_platform(self, content: str, images: Dict[str, str], platform: str, topic_data: Dict, title: str) -> str:
|
||||
template_path = TEMPLATES_DIR / f"{platform}.html"
|
||||
if template_path.exists():
|
||||
with open(template_path, 'r', encoding='utf-8') as f:
|
||||
template = f.read()
|
||||
else:
|
||||
template = "<!DOCTYPE html><html><body>{{TITLE}}<hr><!-- CONTENT --></body></html>"
|
||||
|
||||
# 插入图片（Path 转字符串）
|
||||
for marker, path in images.items():
|
||||
if path:
|
||||
content = content.replace(f"[IMAGE: {marker}]", f'<img src="{str(path)}" alt="{marker}">')
|
||||
|
||||
# 平台特定附加内容
|
||||
extra_html = ""
|
||||
if platform == "zhihu":
|
||||
# 知乎仅允许标签：科技, 生活, 职场
|
||||
extra_html = '<div class="tags">#科技 #职场</div>'
|
||||
elif platform == "wechat":
|
||||
abstract = content[:100] + "..."
|
||||
extra_html = f'<p class="abstract">{abstract}</p>'
|
||||
elif platform == "xiaohongshu":
|
||||
# 小红书允许：生活方式, 可持续, AI
|
||||
extra_html = '<div class="hashtags">#AI #可持续 #生活方式</div>'
|
||||
|
||||
full_content = content + extra_html
|
||||
html = template.replace("<!-- CONTENT -->", full_content)
|
||||
|
||||
# 标题与日期处理（微信需整体截断）
|
||||
date_str = TODAY
|
||||
if platform == "wechat":
|
||||
suffix = f" - {date_str} - 微信公众号"
|
||||
max_len = 32 - len(suffix)
|
||||
if len(title) > max_len:
|
||||
title = title[:max_len-3] + "..."
|
||||
full_title = title + suffix
|
||||
else:
|
||||
full_title = title
|
||||
|
||||
html = html.replace("{{DATE}}", date_str)
|
||||
html = html.replace("{{TITLE}}", full_title)
|
||||
return html
|
||||
|
||||
def mark_topic_ready(self, topic_id: str):
|
||||
topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
with open(topics_file, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
for t in topics:
|
||||
if t.get("id") == topic_id:
|
||||
t["status"] = "待发布"
|
||||
t["ready_at"] = TODAY
|
||||
break
|
||||
with open(topics_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
logger.info(f"选题 {topic_id} 已标记为「待发布」")
|
||||
|
||||
def run(self):
|
||||
logger.info("开始内容创作")
|
||||
self.load_config()
|
||||
topic_data = self.select_topic_for_today()
|
||||
if not topic_data:
|
||||
logger.error("未能选择选题，任务结束")
|
||||
return False
|
||||
|
||||
content = self.create_content(topic_data)
|
||||
images = self.generate_images(topic_data["topic"].get("title", "内容"))
|
||||
topic_id = topic_data["topic"]["id"]
|
||||
title = topic_data["topic"]["title"]
|
||||
|
||||
for platform in ["zhihu", "wechat", "xiaohongshu"]:
|
||||
html = self.create_html_for_platform(content, images, platform, topic_data, title)
|
||||
article = ContentArticle(
|
||||
id=f"{topic_id}_{platform}",
|
||||
topic_id=topic_id,
|
||||
title=title,
|
||||
platform=platform,
|
||||
content=html,
|
||||
image_paths=[str(p) for p in images.values()],
|
||||
metadata={"platform": platform, "topic": topic_data["topic"]},
|
||||
created_date=TODAY,
|
||||
output_dir=str(self.release_dir / platform),
|
||||
status="pending_review"
|
||||
)
|
||||
self.save_article(article)
|
||||
self.articles.append(article)
|
||||
|
||||
self.mark_topic_ready(topic_id)
|
||||
logger.info(f"创作完成: {len(self.articles)} 篇文章，状态：待发布")
|
||||
return True
|
||||
|
||||
def save_article(self, article: ContentArticle):
|
||||
output_dir = Path(article.output_dir)
|
||||
output_dir.mkdir(parents=True, exist_ok=True)
|
||||
html_file = output_dir / f"{article.platform}_{article.id}.html"
|
||||
with open(html_file, 'w', encoding='utf-8') as f:
|
||||
f.write(article.content)
|
||||
meta_file = output_dir / f"{article.platform}_{article.id}.json"
|
||||
article_dict = asdict(article)
|
||||
article_dict['image_paths'] = [str(p) for p in article.image_paths]
|
||||
json.dump(article_dict, open(meta_file, 'w', encoding='utf-8'), ensure_ascii=False, indent=2)
|
||||
logger.info(f"保存了 {article.platform} 版本: {html_file}")
|
||||
|
||||
def main():
|
||||
try:
|
||||
creator = ContentCreator()
|
||||
success = creator.run()
|
||||
if success:
|
||||
print(f"SUCCESS: Created {len(creator.articles)} articles for {TODAY} (status: 待发布)")
|
||||
sys.exit(0)
|
||||
else:
|
||||
print("WARNING: Content creation failed")
|
||||
sys.exit(1)
|
||||
except Exception as e:
|
||||
logger.error(f"创作任务失败: {e}")
|
||||
print(f"ERROR: {e}")
|
||||
sys.exit(1)
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,651 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
内容创作脚本
|
||||
每天凌晨5:30运行，从选题库选出最适合当天发布的题目，创作内容，生成图片内联HTML
|
||||
"""
|
||||
|
||||
import os
|
||||
import sys
|
||||
import yaml
|
||||
import json
|
||||
import datetime
|
||||
import logging
|
||||
import random
|
||||
from pathlib import Path
|
||||
from typing import Dict, List, Optional, Tuple
|
||||
import hashlib
|
||||
from dataclasses import dataclass, asdict
|
||||
import subprocess
|
||||
import re
|
||||
|
||||
# 项目根目录
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
|
||||
# 导入图片生成器
|
||||
from scripts.image_generator import ImageGenerator
|
||||
|
||||
# 配置路径
|
||||
CONFIG_DIR = PROJECT_ROOT / "config"
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TEMPLATES_DIR = PROJECT_ROOT / "automation" / "templates"
|
||||
IMAGES_DIR = PROJECT_ROOT / "automation" / "images"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
# 日志配置
|
||||
logging.basicConfig(
|
||||
level=logging.INFO,
|
||||
format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
|
||||
handlers=[
|
||||
logging.FileHandler(LOGS_DIR / f"creator_{TODAY}.log"),
|
||||
logging.StreamHandler()
|
||||
]
|
||||
)
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
@dataclass
|
||||
class ContentArticle:
|
||||
"""内容文章"""
|
||||
id: str
|
||||
topic_id: str
|
||||
title: str
|
||||
platform: str # zhihu, wechat, xiaohongshu
|
||||
content: str # HTML内容
|
||||
image_paths: List[str]
|
||||
metadata: Dict
|
||||
created_date: str
|
||||
output_dir: str
|
||||
|
||||
class ContentCreator:
|
||||
"""内容创作器"""
|
||||
|
||||
def __init__(self):
|
||||
self.load_config()
|
||||
self.today_dir = DATA_DIR / "drafts" / TODAY
|
||||
self.today_dir.mkdir(parents=True, exist_ok=True)
|
||||
self.release_dir = DATA_DIR / "releases" / TODAY
|
||||
self.release_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# 结果存储
|
||||
self.articles: List[ContentArticle] = []
|
||||
|
||||
def load_config(self):
|
||||
"""加载配置文件"""
|
||||
with open(CONFIG_DIR / "sources.yaml", "r", encoding='utf-8') as f:
|
||||
self.config = yaml.safe_load(f)
|
||||
|
||||
with open(CONFIG_DIR / "wecom_config.yaml", "r", encoding='utf-8') as f:
|
||||
self.wecom_config = yaml.safe_load(f)
|
||||
|
||||
# 平台规则
|
||||
self.platform_rules = self.wecom_config["content_rules"]
|
||||
|
||||
logger.info("配置加载完成")
|
||||
|
||||
def select_topic_for_today(self) -> Optional[Dict]:
|
||||
"""选择最适合当天发布的选题"""
|
||||
topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
if not topics_file.exists():
|
||||
logger.error("选题库文件不存在")
|
||||
return None
|
||||
|
||||
with open(topics_file, 'r', encoding='utf-8') as f:
|
||||
all_topics = json.load(f)
|
||||
|
||||
if not all_topics:
|
||||
logger.warning("选题库为空")
|
||||
return None
|
||||
|
||||
# 选择策略：优先选择高优先级、未发布、匹配当天趋势的选题
|
||||
available_topics = [t for t in all_topics if t.get("status") != "已发布"]
|
||||
|
||||
if not available_topics:
|
||||
logger.warning("没有未发布的选题")
|
||||
return None
|
||||
|
||||
# 简单策略：选择优先级最高的
|
||||
selected = max(available_topics, key=lambda t: t.get("priority_score", 0))
|
||||
|
||||
# 获取相关案例
|
||||
case_ids = selected.get("cases", [])
|
||||
cases_file = DATA_DIR / "sustainability_cases.json"
|
||||
related_cases = []
|
||||
|
||||
if cases_file.exists():
|
||||
with open(cases_file, 'r', encoding='utf-8') as f:
|
||||
all_cases = json.load(f)
|
||||
related_cases = [c for c in all_cases if c.get("id") in case_ids]
|
||||
|
||||
logger.info(f"选择了选题: {selected.get('title')} (优先级: {selected.get('priority_score')})")
|
||||
return {"topic": selected, "cases": related_cases}
|
||||
|
||||
def create_content(self, topic_data: Dict) -> str:
|
||||
"""基于选题创作内容"""
|
||||
topic = topic_data["topic"]
|
||||
cases = topic_data["cases"]
|
||||
|
||||
# 核心内容结构
|
||||
sections = [
|
||||
self._create_introduction(topic, cases),
|
||||
self._create_global_cases_section(cases),
|
||||
self._create_china_pain_analysis(topic),
|
||||
self._create_localization_solution(topic),
|
||||
self._create_mvp_actions(topic),
|
||||
self._create_conclusion(topic)
|
||||
]
|
||||
|
||||
content = "\n\n".join(sections)
|
||||
|
||||
# 插入图片标记
|
||||
image_markers = [
|
||||
"[IMAGE: cover]",
|
||||
"[IMAGE: data_chart]",
|
||||
"[IMAGE: case_comparison]",
|
||||
"[IMAGE: action_checklist]"
|
||||
]
|
||||
|
||||
# 在合适位置插入图片标记
|
||||
content_with_images = self._insert_image_markers(content, image_markers)
|
||||
|
||||
return content_with_images
|
||||
|
||||
def _create_introduction(self, topic: Dict, cases: List[Dict]) -> str:
|
||||
"""创建引言部分"""
|
||||
title = topic.get("title", "")
|
||||
category = cases[0].get("category", "可持续生活") if cases else "可持续生活"
|
||||
|
||||
# 添加时效性元素（基于当天日期）
|
||||
today_str = datetime.datetime.now().strftime("%Y年%m月%d日")
|
||||
|
||||
introduction = f"""
|
||||
<h2>{title}</h2>
|
||||
|
||||
<p>今天是{today_str}，全球可持续性领域又有新的进展。根据最新收集的数据和案例，我们发现{category}领域出现了一些值得关注的新趋势。</p>
|
||||
|
||||
<p>这些国际案例对中国读者有什么启示？它们能否在中国落地？本文将通过全球案例对比和中国痛点分析，给出具体的本土化建议和可执行行动清单。</p>
|
||||
|
||||
<p><strong>核心观点</strong>：国际先进经验不能盲目照搬，必须结合中国现实进行调整。关键在于找到"最小可行行动"，让可持续生活从理念变为日常实践。</p>
|
||||
"""
|
||||
return introduction.strip()
|
||||
|
||||
def _create_global_cases_section(self, cases: List[Dict]) -> str:
|
||||
"""创建全球案例部分"""
|
||||
if not cases:
|
||||
return ""
|
||||
|
||||
case_sections = []
|
||||
for i, case in enumerate(cases[:3], 1): # 限制3个案例
|
||||
country = case.get("country", "全球")
|
||||
title = case.get("title", "")
|
||||
core_idea = case.get("core_idea", "")
|
||||
data_facts = case.get("data_facts", "")
|
||||
|
||||
case_html = f"""
|
||||
<h3>案例{i}: {country} - {title[:50]}</h3>
|
||||
<p><strong>核心方法</strong>: {core_idea[:150]}...</p>
|
||||
<p><strong>数据支撑</strong>: {data_facts}</p>
|
||||
<p><strong>全球优势</strong>: {case.get('global_advantage', '需进一步分析')}</p>
|
||||
"""
|
||||
case_sections.append(case_html.strip())
|
||||
|
||||
return "\n".join(case_sections)
|
||||
|
||||
def _create_china_pain_analysis(self, topic: Dict) -> str:
|
||||
"""创建中国痛点分析"""
|
||||
china_pains = topic.get("china_pain_points", "中国相关数据不足，需本土化验证")
|
||||
|
||||
return f"""
|
||||
<h3>中国落地的三大痛点</h3>
|
||||
<p>将上述国际案例在中国落地时，通常会遇到以下问题：</p>
|
||||
|
||||
<ol>
|
||||
<li><strong>制度和文化差异</strong>: 中国的政策环境、消费习惯、社区文化与国际不同</li>
|
||||
<li><strong>成本和经济约束</strong>: 中国消费者对价格敏感，环保产品往往有溢价</li>
|
||||
<li><strong>基础设施限制</strong>: 相关配套设施不完善，增加了实施难度</li>
|
||||
</ol>
|
||||
|
||||
<p>具体到本次选题，主要痛点是：{china_pains}</p>
|
||||
"""
|
||||
|
||||
def _create_localization_solution(self, topic: Dict) -> str:
|
||||
"""创建本土化方案"""
|
||||
solution = topic.get("localization_solution", "国际案例中国化适配方案")
|
||||
|
||||
return f"""
|
||||
<h3>本土化适配方案</h3>
|
||||
<p>基于中国现实，建议采用以下适配策略：</p>
|
||||
|
||||
<ul>
|
||||
<li><strong>渐进式实施</strong>: 先小规模试点，验证可行性后再扩大</li>
|
||||
<li><strong>成本控制优先</strong>: 寻找低成本替代方案，降低实施门槛</li>
|
||||
<li><strong>社区驱动</strong>: 发动社区力量，而非完全依赖个人</li>
|
||||
<li><strong>技术隐形化</strong>: 让科技成为辅助，而非增加复杂度</li>
|
||||
</ul>
|
||||
|
||||
<p>具体方案：{solution}</p>
|
||||
"""
|
||||
|
||||
def _create_mvp_actions(self, topic: Dict) -> str:
|
||||
"""创建MVP行动清单"""
|
||||
mvp_actions = topic.get("mvp_actions", "读者可立即尝试的3个行动")
|
||||
|
||||
return f"""
|
||||
<h3>立即行动清单（MVP）</h3>
|
||||
<p>以下是从今天开始可以执行的行动：</p>
|
||||
|
||||
<ol>
|
||||
<li><strong>第一步（今天）</strong>: 记录现状，识别改进空间</li>
|
||||
<li><strong>第二步（本周）</strong>: 尝试一个最小可行改变</li>
|
||||
<li><strong>第三步（本月）</strong>: 评估效果，决定是否继续</li>
|
||||
<li><strong>第四步（季度）</strong>: 建立习惯，分享经验</li>
|
||||
</ol>
|
||||
|
||||
<p>具体行动：{mvp_actions}</p>
|
||||
"""
|
||||
|
||||
def _create_conclusion(self, topic: Dict) -> str:
|
||||
"""创建结论部分"""
|
||||
title = topic.get("title", "")
|
||||
|
||||
return f"""
|
||||
<h3>总结与展望</h3>
|
||||
<p>{title}的核心在于<strong>行动而非理论</strong>。国际案例提供参考，但最终的成功取决于在中国环境下的创造性适配。</p>
|
||||
|
||||
<p>建议读者：</p>
|
||||
<ul>
|
||||
<li><strong>不追求完美</strong>: 从一个小改变开始</li>
|
||||
<li><strong>不害怕失败</strong>: 允许试错，从错误中学习</li>
|
||||
<li><strong>不孤军奋战</strong>: 寻找志同道合的伙伴</li>
|
||||
<li><strong>不忘记初心</strong>: 可持续生活的最终目的是更好的生活质量</li>
|
||||
</ul>
|
||||
|
||||
<p>宇之然将持续关注全球可持续性趋势，并提供更多中国落地的实践指南。</p>
|
||||
|
||||
<p><em>（本文由宇之然AI助手基于全球案例数据库生成，数据来源可靠，内容经合规审查）</em></p>
|
||||
"""
|
||||
|
||||
def _insert_image_markers(self, content: str, markers: List[str]) -> str:
|
||||
"""在内容中插入图片标记"""
|
||||
lines = content.split('\n')
|
||||
result_lines = []
|
||||
image_index = 0
|
||||
|
||||
for line in lines:
|
||||
result_lines.append(line)
|
||||
# 在合适位置插入图片标记（如段落之后）
|
||||
if line.startswith('<h3>') and image_index < len(markers):
|
||||
result_lines.append(markers[image_index])
|
||||
image_index += 1
|
||||
|
||||
# 开头添加封面图
|
||||
result_lines.insert(2, markers[0]) if markers else None
|
||||
|
||||
return '\n'.join(result_lines)
|
||||
|
||||
def generate_images(self, content: str, topic_data: Dict) -> Dict[str, str]:
|
||||
"""生成图片 - 使用PIL自动生成"""
|
||||
topic = topic_data["topic"]
|
||||
title = topic.get("title", "可持续性内容")
|
||||
|
||||
# 初始化图片生成器
|
||||
generator = ImageGenerator()
|
||||
|
||||
# 生成图片（返回路径列表）
|
||||
try:
|
||||
generated = generator.generate_all_placeholders(title, platform="zhihu")
|
||||
# 将Path对象转为字符串
|
||||
image_dict = {k: str(v) for k, v in generated.items()}
|
||||
logger.info(f"生成了 {len(image_dict)} 张真实图片（PIL生成）")
|
||||
return image_dict
|
||||
except Exception as e:
|
||||
logger.error(f"图片生成失败，回退到占位符: {e}")
|
||||
# 回退：生成占位符文本文件
|
||||
return self._generate_placeholder_images(title)
|
||||
|
||||
def _generate_placeholder_images(self, title: str) -> Dict[str, str]:
|
||||
"""生成占位符图片（文本文件）"""
|
||||
images_dir = IMAGES_DIR / "generated" / TODAY
|
||||
images_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
image_dict = {}
|
||||
# 1. 封面图
|
||||
cover_path = images_dir / "cover.txt"
|
||||
with open(cover_path, 'w', encoding='utf-8') as f:
|
||||
f.write(f"封面图: {title}\n日期: {TODAY}\n作者: 宇之然")
|
||||
image_dict["cover"] = str(cover_path)
|
||||
|
||||
# 2. 数据图表
|
||||
chart_path = images_dir / "data_chart.txt"
|
||||
with open(chart_path, 'w', encoding='utf-8') as f:
|
||||
f.write("数据图表（示例）\n")
|
||||
f.write("可持续性效果对比\n")
|
||||
f.write("国际案例 vs 中国实践")
|
||||
image_dict["data_chart"] = str(chart_path)
|
||||
|
||||
# 3. 案例对比
|
||||
comparison_path = images_dir / "case_comparison.txt"
|
||||
with open(comparison_path, 'w', encoding='utf-8') as f:
|
||||
f.write("案例对比表格\n")
|
||||
f.write("全球最佳实践 → 中国适配建议")
|
||||
image_dict["case_comparison"] = str(comparison_path)
|
||||
|
||||
# 4. 行动清单
|
||||
checklist_path = images_dir / "action_checklist.txt"
|
||||
with open(checklist_path, 'w', encoding='utf-8') as f:
|
||||
f.write("立即行动清单\n")
|
||||
f.write("1. 记录现状\n2. 小步尝试\n3. 评估效果\n4. 建立习惯")
|
||||
image_dict["action_checklist"] = str(checklist_path)
|
||||
|
||||
logger.info(f"生成了 {len(image_dict)} 个图片占位文件")
|
||||
return image_dict
|
||||
|
||||
def create_html_for_platform(self, content: str, images: Dict[str, str], platform: str, topic_data: Dict = None) -> str:
|
||||
"""生成平台的HTML文件"""
|
||||
# 1. 在 content 中替换图片标记
|
||||
image_markers = {
|
||||
"[IMAGE: cover]": images.get("cover", ""),
|
||||
"[IMAGE: data_chart]": images.get("data_chart", ""),
|
||||
"[IMAGE: case_comparison]": images.get("case_comparison", ""),
|
||||
"[IMAGE: action_checklist]": images.get("action_checklist", ""),
|
||||
"[IMAGE: equipment]": images.get("equipment", ""),
|
||||
}
|
||||
for marker, image_path in image_markers.items():
|
||||
if image_path:
|
||||
img_tag = f'<img src="{image_path}" alt="{marker}">'
|
||||
content = content.replace(marker, img_tag)
|
||||
|
||||
# 2. 替换索引标记
|
||||
ordered_keys = ["cover", "data_chart", "case_comparison", "action_checklist", "equipment"]
|
||||
for i, key in enumerate(ordered_keys, 1):
|
||||
if key in images:
|
||||
marker = f"[IMAGE: image_{i}]"
|
||||
img_tag = f'<img src="{images[key]}" alt="图片{i}" class="platform-{platform}">'
|
||||
content = content.replace(marker, img_tag)
|
||||
|
||||
# 3. 填充模板（将内容放入）
|
||||
template = self._get_base_template()
|
||||
html = template.replace("<!-- CONTENT -->", content)
|
||||
|
||||
# 4. 平台特定占位符替换（在 html 上进行）
|
||||
if platform == "zhihu":
|
||||
# 确定选题领域
|
||||
field = "可持续生活"
|
||||
if topic_data and isinstance(topic_data, dict):
|
||||
topic_field = topic_data.get("topic", {}).get("field", "")
|
||||
if topic_field:
|
||||
field = topic_field
|
||||
# 领域映射到标签
|
||||
tag_map = {
|
||||
"未来工作方式": ["科技", "职场", "AI"],
|
||||
"可持续生活系统": ["可持续", "生活", "环保"],
|
||||
"个人知识工厂": ["知识管理", "个人成长", "效率"],
|
||||
"科技人文交叉": ["科技", "人文", "AI伦理"],
|
||||
}
|
||||
tags = tag_map.get(field, ["科技", "生活", "可持续"])[:4]
|
||||
tags.append(TODAY[:4]+"年趋势")
|
||||
tags_section = '<div class="tags">' + " ".join(f'#{tag}' for tag in tags) + '</div>'
|
||||
logger.info(f"[DEBUG] Replacing TAGS with: {tags_section}")
|
||||
if "<!-- TAGS -->" in html:
|
||||
html = html.replace("<!-- TAGS -->", tags_section)
|
||||
logger.info(f"[DEBUG] After TAGS replace, length: {len(html)}")
|
||||
else:
|
||||
logger.warning("[DEBUG] TAGS placeholder not found in HTML! Template may be missing.")
|
||||
# Fallback: append tags at end of content
|
||||
html = html.replace("</body>", tags_section + "\n</body>")
|
||||
|
||||
elif platform == "wechat":
|
||||
abstract = content[:100] + "..."
|
||||
html = html.replace("<!-- ABSTRACT -->", f'<p class="abstract">{abstract}</p>')
|
||||
|
||||
elif platform == "xiaohongshu":
|
||||
field = "可持续生活"
|
||||
if topic_data and isinstance(topic_data, dict):
|
||||
topic_field = topic_data.get("topic", {}).get("field", "")
|
||||
if topic_field:
|
||||
field = topic_field
|
||||
hashtag_map = {
|
||||
"未来工作方式": ["#远程工作", "#数字游民", "#AI副业"],
|
||||
"可持续生活系统": ["#可持续生活", "#零浪费", "#环保"],
|
||||
"个人知识工厂": ["#第二大脑", "#PKM", "#个人成长"],
|
||||
"科技人文交叉": ["#科技", "#AI伦理", "#数字健康"],
|
||||
}
|
||||
hashtags = hashtag_map.get(field, ["#可持续生活", "#全球视野", "#宇之然"])[:5]
|
||||
hashtags_section = '<div class="hashtags">' + " ".join(hashtags) + '</div>'
|
||||
html = html.replace("<!-- HASHTAGS -->", hashtags_section)
|
||||
|
||||
# 5. 替换日期和标题
|
||||
title = "可持续性内容"
|
||||
if topic_data and isinstance(topic_data, dict):
|
||||
topic_title = topic_data.get("topic", {}).get("title")
|
||||
if topic_title:
|
||||
title = topic_title
|
||||
html = html.replace("{{DATE}}", TODAY)
|
||||
html = html.replace("{{TITLE}}", title)
|
||||
|
||||
# Debug: check tags presence
|
||||
if platform == "zhihu":
|
||||
tags_pos = html.find('class="tags"')
|
||||
if tags_pos != -1:
|
||||
snippet = html[max(0, tags_pos-50):tags_pos+100]
|
||||
logger.info(f"[DEBUG] Tags found: ...{snippet}...")
|
||||
else:
|
||||
logger.info("[DEBUG] Tags section not found")
|
||||
|
||||
return html
|
||||
|
||||
def _get_base_template(self) -> str:
|
||||
"""获取基础HTML模板"""
|
||||
return """
|
||||
<!DOCTYPE html>
|
||||
<html lang="zh-CN">
|
||||
<head>
|
||||
<meta charset="UTF-8">
|
||||
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
||||
<title>{{TITLE}} - {{DATE}}</title>
|
||||
<style>
|
||||
body { font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, sans-serif; line-height: 1.6; max-width: 800px; margin: 0 auto; padding:66px; }
|
||||
h1, h2, h3 { color: #333; margin-top: 1.5em; }
|
||||
p { margin: 1em 0; color: #555; }
|
||||
ul, ol { padding-left: 1.5em; }
|
||||
img { max-width: 100%; height: auto; margin: 1em 0; border: 1px solid #eee; }
|
||||
.tags, .hashtags { margin: 1em 0; color: #666; }
|
||||
.abstract { background: #f9f9f9; padding: 1em; border-left: 4px solid #4CAF50; }
|
||||
footer { margin-top: 2em; padding-top: 1em; border-top: 1px solid #eee; color: #888; font-size: 0.9em; }
|
||||
</style>
|
||||
</head>
|
||||
<body>
|
||||
<!-- CONTENT -->
|
||||
<footer>
|
||||
<p>本文由宇之然AI助手生成 | 数据来源：全球可持续性信息源 | 生成日期：{{DATE}}</p>
|
||||
</footer>
|
||||
</body>
|
||||
</html>
|
||||
"""
|
||||
|
||||
def save_article(self, article: ContentArticle):
|
||||
"""保存文章"""
|
||||
output_dir = Path(article.output_dir)
|
||||
output_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# HTML文件
|
||||
html_file = output_dir / f"{article.platform}_{article.id}.html"
|
||||
with open(html_file, 'w', encoding='utf-8') as f:
|
||||
f.write(article.content)
|
||||
|
||||
# 元数据
|
||||
meta_file = output_dir / f"{article.platform}_{article.id}.json"
|
||||
with open(meta_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(asdict(article), f, ensure_ascii=False, indent=2)
|
||||
|
||||
logger.info(f"保存了 {article.platform} 版本文章: {html_file}")
|
||||
|
||||
def send_wecom_notification(self):
|
||||
"""发送企业微信通知"""
|
||||
try:
|
||||
notification_script = PROJECT_ROOT / "scripts" / "wecom_notifier.py"
|
||||
if not notification_script.exists():
|
||||
logger.warning("企业微信通知脚本不存在")
|
||||
return
|
||||
|
||||
# 准备通知数据
|
||||
notification_data = {
|
||||
"task": "content_creation",
|
||||
"time": datetime.datetime.now().strftime("%Y-%m-%d %H:%M"),
|
||||
"topic_title": self.articles[0].title if self.articles else "无",
|
||||
"image_count": len(self.articles[0].image_paths) if self.articles else 0,
|
||||
"output_dir": str(self.release_dir.relative_to(PROJECT_ROOT)),
|
||||
"status": "完成" if self.articles else "失败"
|
||||
}
|
||||
|
||||
data_file = self.today_dir / "creator_notification.json"
|
||||
with open(data_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(notification_data, f, ensure_ascii=False)
|
||||
|
||||
# 运行通知脚本
|
||||
result = subprocess.run(
|
||||
[sys.executable, str(notification_script), str(data_file)],
|
||||
capture_output=True,
|
||||
text=True,
|
||||
cwd=PROJECT_ROOT
|
||||
)
|
||||
|
||||
if result.returncode == 0:
|
||||
logger.info("企业微信通知发送成功")
|
||||
else:
|
||||
logger.error(f"通知发送失败: {result.stderr}")
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"发送通知失败: {e}")
|
||||
|
||||
def run(self):
|
||||
"""主运行流程"""
|
||||
logger.info("开始内容创作")
|
||||
|
||||
# 1. 选择选题
|
||||
topic_data = self.select_topic_for_today()
|
||||
if not topic_data:
|
||||
logger.error("未能选择选题，任务结束")
|
||||
return False
|
||||
|
||||
# 2. 创作内容
|
||||
content = self.create_content(topic_data)
|
||||
|
||||
# 3. 生成图片 (返回字典)
|
||||
images_dict = self.generate_images(content, topic_data)
|
||||
|
||||
# 4. 为每个平台生成HTML
|
||||
platforms = ["zhihu", "wechat", "xiaohongshu"]
|
||||
topic_id = topic_data["topic"].get("id", "unknown")
|
||||
|
||||
# 确定图片顺序
|
||||
ordered_keys = ["cover", "data_chart", "case_comparison", "action_checklist", "equipment"]
|
||||
images_list = [images_dict[k] for k in ordered_keys if k in images_dict]
|
||||
|
||||
for platform in platforms:
|
||||
html = self.create_html_for_platform(content, images_dict, platform, topic_data)
|
||||
|
||||
article = ContentArticle(
|
||||
id=f"{topic_id}_{platform}",
|
||||
topic_id=topic_id,
|
||||
title=topic_data["topic"].get("title", ""),
|
||||
platform=platform,
|
||||
content=html,
|
||||
image_paths=images_list, # 有序列表
|
||||
metadata={
|
||||
"platform": platform,
|
||||
"topic": topic_data["topic"],
|
||||
"cases": topic_data["cases"],
|
||||
"word_count": len(content)
|
||||
},
|
||||
created_date=TODAY,
|
||||
output_dir=str(self.release_dir / platform)
|
||||
)
|
||||
|
||||
self.save_article(article)
|
||||
self.articles.append(article)
|
||||
|
||||
# 5. 更新选题状态
|
||||
self.update_topic_status(topic_id)
|
||||
|
||||
# 6. 发送通知
|
||||
self.send_wecom_notification()
|
||||
|
||||
# 7. 合规审查（如有生成文章）
|
||||
if self.articles:
|
||||
self.run_compliance_check(self.articles[0])
|
||||
|
||||
logger.info(f"创作完成: {len(self.articles)} 篇文章")
|
||||
return True
|
||||
|
||||
def update_topic_status(self, topic_id: str):
|
||||
"""更新选题状态为已发布"""
|
||||
topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
if not topics_file.exists():
|
||||
return
|
||||
|
||||
with open(topics_file, 'r', encoding='utf-8') as f:
|
||||
all_topics = json.load(f)
|
||||
|
||||
for topic in all_topics:
|
||||
if topic.get("id") == topic_id:
|
||||
topic["status"] = "已发布"
|
||||
topic["published_date"] = TODAY
|
||||
break
|
||||
|
||||
with open(topics_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(all_topics, f, ensure_ascii=False, indent=2)
|
||||
|
||||
logger.info(f"更新选题 {topic_id} 状态为已发布")
|
||||
|
||||
def run_compliance_check(self, article):
|
||||
"""运行合规审查"""
|
||||
try:
|
||||
from scripts.compliance_checker import check_article
|
||||
|
||||
result = check_article(
|
||||
article.content,
|
||||
article.platform,
|
||||
article.metadata.get("topic") if hasattr(article, 'metadata') else None
|
||||
)
|
||||
|
||||
# 记录审查结果
|
||||
log_msg = f"合规审查: {article.platform} - {article.title[:30]} - 得分: {result['score']} - 问题数: {len(result['issues'])}"
|
||||
if result['passed']:
|
||||
logger.info(log_msg)
|
||||
else:
|
||||
logger.warning(log_msg)
|
||||
for issue in result['issues']:
|
||||
logger.warning(f" [合规问题] {issue['type']}/{issue.get('category','')}: {issue.get('suggestion','')}")
|
||||
|
||||
# 保存审查报告
|
||||
compliance_file = self.today_dir / f"compliance_{article.id}.json"
|
||||
with open(compliance_file, 'w', encoding='utf-8') as f:
|
||||
json.dump(result, f, ensure_ascii=False, indent=2)
|
||||
|
||||
logger.info(f"合规报告已保存: {compliance_file}")
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"合规审查失败: {e}")
|
||||
|
||||
|
||||
def main():
|
||||
"""主函数"""
|
||||
try:
|
||||
creator = ContentCreator()
|
||||
success = creator.run()
|
||||
|
||||
if success:
|
||||
print(f"SUCCESS: Created {len(creator.articles)} articles for {TODAY}")
|
||||
sys.exit(0)
|
||||
else:
|
||||
print("WARNING: Content creation failed")
|
||||
sys.exit(1)
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"创作任务失败: {e}")
|
||||
print(f"ERROR: {e}")
|
||||
sys.exit(1)
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
+136
-3
@@ -80,11 +80,50 @@ def update_topic_status(topic_id: str, status: str, db: Optional[Session] = None
|
||||
if close_db:
|
||||
db.close()
|
||||
|
||||
def get_active_llm_config(db: Optional[Session] = None) -> Optional[Dict]:
|
||||
"""获取启用的 LLM 配置，优先读取 system_configs 中的 review_llm_id
|
||||
|
||||
返回格式: { "model", "temperature", "max_tokens", "system_prompt", "name" }
|
||||
无配置时返回 None（调用方应使用环境变量默认值）
|
||||
"""
|
||||
close_db = False
|
||||
if db is None:
|
||||
db = SessionLocal()
|
||||
close_db = True
|
||||
try:
|
||||
from app.models import LLMConfig, SystemConfig
|
||||
# 先查系统配置中指定的 review_llm_id
|
||||
sc = db.query(SystemConfig).filter(SystemConfig.key == 'review_llm_id').first()
|
||||
target_id = None
|
||||
if sc and sc.value and sc.value.isdigit():
|
||||
target_id = int(sc.value)
|
||||
query = db.query(LLMConfig)
|
||||
if target_id:
|
||||
query = query.filter(LLMConfig.id == target_id)
|
||||
query = query.filter(LLMConfig.is_active == True)
|
||||
cfg = query.first()
|
||||
if not cfg:
|
||||
cfg = db.query(LLMConfig).filter(LLMConfig.is_active == True).first()
|
||||
if cfg:
|
||||
return {
|
||||
"id": cfg.id,
|
||||
"name": cfg.name,
|
||||
"model": cfg.model,
|
||||
"temperature": cfg.temperature,
|
||||
"max_tokens": cfg.max_tokens,
|
||||
"system_prompt": cfg.system_prompt,
|
||||
"user_prompt_template": cfg.user_prompt_template,
|
||||
}
|
||||
return None
|
||||
finally:
|
||||
if close_db:
|
||||
db.close()
|
||||
|
||||
def topic_to_dict(topic: Topic) -> Dict:
|
||||
return {
|
||||
'id': topic.id,
|
||||
'title': topic.title,
|
||||
'field': topic.field.name if topic.field else None,
|
||||
'field': topic.field_name,
|
||||
'format': topic.format,
|
||||
'core_concept': topic.core_concept,
|
||||
'audience_pain': topic.audience_pain,
|
||||
@@ -131,7 +170,7 @@ def save_topics_to_db(topics_data: List[Dict]):
|
||||
existing = db.query(Topic).filter(Topic.id == t['id']).first()
|
||||
if existing:
|
||||
# 更新字段
|
||||
for field in ['title', 'field', 'format', 'core_concept', 'audience_pain', 'unique_angle', 'priority', 'priority_score', 'total_score', 'status', 'cases', 'source_file', 'compliance_score', 'platform_urls']:
|
||||
for field in ['title', 'field_name', 'format', 'core_concept', 'audience_pain', 'unique_angle', 'priority', 'priority_score', 'total_score', 'status', 'cases', 'source_file', 'compliance_score', 'platform_urls']:
|
||||
setattr(existing, field, t.get(field, getattr(existing, field)))
|
||||
if t.get('ready_at'):
|
||||
try:
|
||||
@@ -148,7 +187,7 @@ def save_topics_to_db(topics_data: List[Dict]):
|
||||
new_topic = Topic(
|
||||
id=t['id'],
|
||||
title=t['title'],
|
||||
field=t.get('field', '可持续生活系统'),
|
||||
field_name=t.get('field_name') or t.get('field', '可持续生活系统'),
|
||||
format=t.get('format'),
|
||||
core_concept=t.get('core_concept'),
|
||||
audience_pain=t.get('audience_pain'),
|
||||
@@ -173,3 +212,97 @@ def save_topics_to_db(topics_data: List[Dict]):
|
||||
raise
|
||||
finally:
|
||||
db.close()
|
||||
|
||||
def save_article(topic_id: str, platform: str, html_content: str, db: Optional[Session] = None) -> Dict:
|
||||
"""保存/更新文章到 articles 表"""
|
||||
close_db = False
|
||||
if db is None:
|
||||
db = SessionLocal()
|
||||
close_db = True
|
||||
try:
|
||||
from app.models import Article
|
||||
article_id = f"{platform}_{topic_id}"
|
||||
existing = db.query(Article).filter(Article.id == article_id).first()
|
||||
now = datetime.now()
|
||||
if existing:
|
||||
existing.html_content = html_content
|
||||
existing.compliance_score = existing.compliance_score
|
||||
else:
|
||||
article = Article(
|
||||
id=article_id,
|
||||
topic_id=topic_id,
|
||||
platform=platform,
|
||||
file_path=f"db:{article_id}",
|
||||
html_content=html_content,
|
||||
status="draft",
|
||||
compliance_score=None
|
||||
)
|
||||
db.add(article)
|
||||
db.commit()
|
||||
return {"id": article_id, "topic_id": topic_id, "platform": platform}
|
||||
except Exception:
|
||||
db.rollback()
|
||||
raise
|
||||
finally:
|
||||
if close_db:
|
||||
db.close()
|
||||
|
||||
def get_article(topic_id: str, platform: str, db: Optional[Session] = None) -> Optional[Dict]:
|
||||
"""从 articles 表获取文章 HTML"""
|
||||
close_db = False
|
||||
if db is None:
|
||||
db = SessionLocal()
|
||||
close_db = True
|
||||
try:
|
||||
from app.models import Article
|
||||
article_id = f"{platform}_{topic_id}"
|
||||
article = db.query(Article).filter(Article.id == article_id).first()
|
||||
if not article:
|
||||
return None
|
||||
return {
|
||||
"id": article.id,
|
||||
"topic_id": article.topic_id,
|
||||
"platform": article.platform,
|
||||
"html_content": article.html_content,
|
||||
"status": article.status,
|
||||
"compliance_score": article.compliance_score,
|
||||
"created_at": article.created_at.isoformat() if article.created_at else None,
|
||||
}
|
||||
finally:
|
||||
if close_db:
|
||||
db.close()
|
||||
|
||||
def get_articles_by_topic(topic_id: str, db: Optional[Session] = None) -> List[Dict]:
|
||||
"""获取某选题在所有平台的文章"""
|
||||
close_db = False
|
||||
if db is None:
|
||||
db = SessionLocal()
|
||||
close_db = True
|
||||
try:
|
||||
from app.models import Article
|
||||
articles = db.query(Article).filter(Article.topic_id == topic_id).all()
|
||||
return [{
|
||||
"id": a.id,
|
||||
"topic_id": a.topic_id,
|
||||
"platform": a.platform,
|
||||
"html_content": a.html_content,
|
||||
"status": a.status,
|
||||
"compliance_score": a.compliance_score,
|
||||
} for a in articles]
|
||||
finally:
|
||||
if close_db:
|
||||
db.close()
|
||||
|
||||
def delete_articles_by_topic(topic_id: str, db: Optional[Session] = None):
|
||||
"""删除某选题的所有文章"""
|
||||
close_db = False
|
||||
if db is None:
|
||||
db = SessionLocal()
|
||||
close_db = True
|
||||
try:
|
||||
from app.models import Article
|
||||
db.query(Article).filter(Article.topic_id == topic_id).delete()
|
||||
db.commit()
|
||||
finally:
|
||||
if close_db:
|
||||
db.close()
|
||||
|
||||
@@ -1,17 +1,21 @@
|
||||
#!/usr/bin/env python3
|
||||
import json
|
||||
data = json.load(open('automation/data/sustainability_topics.json', 'r', encoding='utf-8'))
|
||||
import sys
|
||||
from pathlib import Path
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent))
|
||||
from db_helper import export_topics_to_json
|
||||
|
||||
topics = export_topics_to_json()
|
||||
print(f'✅ 宇之然选题库（新战略版）')
|
||||
print(f'总选题数: {len(data)}')
|
||||
print(f'已发布: {len([t for t in data if t["status"]=="已发布"])}')
|
||||
print(f'待处理: {len([t for t in data if t["status"]=="待处理"])}')
|
||||
print(f'总选题数: {len(topics)}')
|
||||
print(f'已发布: {len([t for t in topics if t["status"]=="published"])}')
|
||||
print(f'待处理: {len([t for t in topics if t["status"] in ("pending", "待处理")])}')
|
||||
print('\n按领域分组:')
|
||||
fields = {}
|
||||
for t in sorted(data, key=lambda x: x['id']):
|
||||
for t in sorted(topics, key=lambda x: x['id']):
|
||||
f = t['field']
|
||||
fields.setdefault(f, []).append(t)
|
||||
for f, items in fields.items():
|
||||
print(f'\n{f} ({len(items)}个):')
|
||||
for t in items:
|
||||
status_icon = '✅' if t['status']=='已发布' else '⏳'
|
||||
status_icon = '✅' if t['status']=='published' else '⏳'
|
||||
print(f' {status_icon} {t["id"]} {t["title"]}')
|
||||
|
||||
@@ -1,12 +1,13 @@
|
||||
#!/usr/bin/env python3
|
||||
import json
|
||||
import sys
|
||||
from pathlib import Path
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent))
|
||||
from db_helper import export_topics_to_json
|
||||
|
||||
with open('automation/data/sustainability_topics.json', 'r', encoding='utf-8') as f:
|
||||
data = json.load(f)
|
||||
|
||||
print(f'总选题数: {len(data)}')
|
||||
pending = [t for t in data if t['status'] == '待处理']
|
||||
avg = sum(t['total_score'] for t in pending) / len(pending) if pending else 0
|
||||
topics = export_topics_to_json()
|
||||
pending = [t for t in topics if t['status'] in ('pending', '待处理')]
|
||||
avg = sum(t['total_score'] or 0 for t in pending) / len(pending) if pending else 0
|
||||
print(f'总选题数: {len(topics)}')
|
||||
print(f'待处理选题数: {len(pending)}')
|
||||
print(f'待处理平均分: {avg:.1f}')
|
||||
print('\n待处理选题详情:')
|
||||
|
||||
+57
-36
@@ -1,6 +1,6 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
大纲阶段：基于研究笔记生成文章大纲（数据库版）
|
||||
大纲阶段：基于选题和研究笔记，用 LLM 动态生成结构化大纲
|
||||
"""
|
||||
|
||||
import json, datetime, logging, sys
|
||||
@@ -9,9 +9,14 @@ from typing import Dict
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
sys.path.insert(0, str(PROJECT_ROOT / "platform" / "backend"))
|
||||
|
||||
# 导入数据库辅助模块
|
||||
from db_helper import get_topic_by_id
|
||||
try:
|
||||
from app.core.nvidia_client import call_llm
|
||||
HAVE_LLM = True
|
||||
except ImportError:
|
||||
HAVE_LLM = False
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
RESEARCH_DIR = DATA_DIR / "research"
|
||||
@@ -28,34 +33,64 @@ class Outliner:
|
||||
self.topic_id = topic_id
|
||||
self.topic = self._load_topic()
|
||||
research_file = RESEARCH_DIR / TODAY / f"{topic_id}_research.md"
|
||||
if not research_file.exists():
|
||||
raise FileNotFoundError(f"Research notes not found: {research_file}")
|
||||
self.research_notes = research_file.read_text(encoding='utf-8')
|
||||
self.research_notes = research_file.read_text(encoding='utf-8') if research_file.exists() else ""
|
||||
self.output_dir = OUTPUT_DIR / TODAY
|
||||
self.output_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
def _load_topic(self) -> Dict:
|
||||
topic = get_topic_by_id(self.topic_id)
|
||||
if not topic:
|
||||
raise ValueError(f"Topic {self.topic_id} not found")
|
||||
if not topic: raise ValueError(f"Topic {self.topic_id} not found")
|
||||
return topic
|
||||
|
||||
def generate_outline(self) -> str:
|
||||
"""生成文章大纲 Markdown（基于模板）"""
|
||||
title = self.topic['title']
|
||||
field = self.topic.get('field', '')
|
||||
core = self.topic.get('core_concept', '')
|
||||
pain = self.topic.get('audience_pain', '')
|
||||
angle = self.topic.get('unique_angle', '')
|
||||
cases_summary = self.research_notes[:2000] if self.research_notes else "暂无研究笔记"
|
||||
|
||||
# 解析研究笔记中的案例数量
|
||||
case_count = self.research_notes.count('### 案例')
|
||||
if HAVE_LLM:
|
||||
prompt = f"""你是一个真人编辑，在为一篇文章列大纲。不要套模板，根据具体内容灵活设计结构。
|
||||
|
||||
outline = f"""# 文章大纲：{title}
|
||||
## 选题信息
|
||||
标题：{title}
|
||||
领域：{field}
|
||||
核心观点：{core}
|
||||
受众痛点：{pain}
|
||||
独特视角：{angle}
|
||||
|
||||
## 研究笔记
|
||||
{cases_summary}
|
||||
|
||||
## 要求
|
||||
- 章节数灵活，5-8章都行，不硬凑
|
||||
- 每章给2-4个要点即可，不需要每章都标字数
|
||||
- 结构要有递进，但不一定非得是痛点-分析-方案这种套路
|
||||
- 不要用「引言」「总结」这类通用标题，要贴合具体内容
|
||||
- 尽量把独特视角和受众痛点融入各章，而不是单独列出来
|
||||
- 每个要点一句话点出核心，不用展开写
|
||||
|
||||
直接输出大纲。"""
|
||||
try:
|
||||
outline = call_llm(prompt, temperature=0.6, max_tokens=2000, system_prompt="你是一个有经验的内容编辑，擅长为不同选题设计差异化的文章结构。")
|
||||
logger.info(f"LLM 大纲生成成功，长度：{len(outline)}")
|
||||
return f"# 文章大纲：{title}\n\n{outline}\n\n---\n*大纲生成时间：{TODAY}*"
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM 大纲生成失败: {e}，使用模板")
|
||||
|
||||
return self._template_outline(title, field, core, pain, angle)
|
||||
|
||||
def _template_outline(self, title, field, core, pain, angle) -> str:
|
||||
case_count = self.research_notes.count('### 案例') if self.research_notes else 0
|
||||
case_section = f"""## 四、全球/行业趋势与案例
|
||||
- 引用研究笔记中的 {case_count} 个案例，精选 2-3 个详述
|
||||
- 数据支撑：提取研究笔记中的关键数据
|
||||
- 趋势分析""" if case_count > 0 else ""
|
||||
return f"""# 文章大纲：{title}
|
||||
|
||||
## 一、引言
|
||||
- 开场场景/痛点引入
|
||||
- 提出核心问题：{title}
|
||||
- 场景切入：{title}
|
||||
- 点明文章价值
|
||||
|
||||
## 二、核心观点
|
||||
@@ -63,36 +98,23 @@ class Outliner:
|
||||
|
||||
## 三、受众痛点分析
|
||||
{pain}
|
||||
{case_section}
|
||||
|
||||
## 四、全球/行业趋势与案例
|
||||
- 引用研究笔记中的 {case_count} 个案例，精选 2-3 个详述
|
||||
- 数据支撑：提取研究笔记中的关键数据
|
||||
- 趋势分析
|
||||
|
||||
## 五、本土落地建议
|
||||
## {"五" if case_section else "四"}、本土落地建议
|
||||
- 结合{field}领域特点
|
||||
- 提供可执行的步骤
|
||||
- 注意事项
|
||||
|
||||
## 六、独特视角：{angle}
|
||||
## {"六" if case_section else "五"}、独特视角：{angle}
|
||||
|
||||
## 七、行动指南（MVP）
|
||||
1. 理解现状
|
||||
2. 小范围试验
|
||||
3. 评估效果
|
||||
4. 形成习惯
|
||||
## {"七" if case_section else "六"}、行动指南
|
||||
1. 了解现状 2. 制定方案 3. 小范围验证 4. 持续优化
|
||||
|
||||
## 八、总结与鼓励
|
||||
- 回顾要点
|
||||
- 呼吁行动
|
||||
|
||||
## 九、参考文献
|
||||
- 从研究笔记中提取来源链接
|
||||
## {"八" if case_section else "七"}、总结与鼓励
|
||||
|
||||
---
|
||||
*大纲生成时间：{TODAY}*
|
||||
"""
|
||||
return outline
|
||||
|
||||
*大纲生成时间：{TODAY}*"""
|
||||
|
||||
def save(self):
|
||||
outline_text = self.generate_outline()
|
||||
@@ -104,9 +126,8 @@ class Outliner:
|
||||
def main():
|
||||
import argparse
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--topic-id', required=True, help='选题ID')
|
||||
parser.add_argument('--topic-id', required=True)
|
||||
args = parser.parse_args()
|
||||
|
||||
o = Outliner(args.topic_id)
|
||||
o.save()
|
||||
print(f"SUCCESS: Outline created for {args.topic_id}")
|
||||
|
||||
+88
-21
@@ -1,21 +1,25 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
研究阶段：为选题收集资料并生成研究笔记（数据库版）
|
||||
"""
|
||||
|
||||
import json, datetime, logging, sys, re
|
||||
from pathlib import Path
|
||||
from typing import Dict, List
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
sys.path.insert(0, str(PROJECT_ROOT / "platform" / "backend"))
|
||||
|
||||
# 导入数据库辅助模块
|
||||
from db_helper import get_topic_by_id
|
||||
try:
|
||||
from app.core.nvidia_client import call_llm
|
||||
HAVE_LLM = True
|
||||
except ImportError:
|
||||
HAVE_LLM = False
|
||||
|
||||
from trends import get_trend_context
|
||||
from web_search import enrich_topic_research
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
CASES_FILE = DATA_DIR / "sustainability_cases.json"
|
||||
OUTPUT_DIR = DATA_DIR / "research" # 研究笔记输出目录
|
||||
OUTPUT_DIR = DATA_DIR / "research"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
@@ -43,12 +47,10 @@ class Researcher:
|
||||
return []
|
||||
|
||||
def find_relevant_cases(self, top_k: int = 5) -> List[Dict]:
|
||||
"""基于标题和字段匹配相关案例（简化）"""
|
||||
field = self.topic.get('field', '').lower()
|
||||
title = self.topic.get('title', '').lower()
|
||||
scored = []
|
||||
for case in self.cases:
|
||||
# 日期过滤：仅保留 2025 年及以后（支持 YYYY-MM-DD 或 YYYY 格式）
|
||||
case_date = case.get('date', '')
|
||||
if case_date:
|
||||
m = re.search(r'(\d{4})', str(case_date))
|
||||
@@ -57,7 +59,6 @@ class Researcher:
|
||||
score = 0
|
||||
if field and field in case.get('field', '').lower():
|
||||
score += 3
|
||||
# 标题关键词匹配
|
||||
case_title = case.get('title', '').lower()
|
||||
for word in title.split():
|
||||
if len(word) > 2 and word in case_title:
|
||||
@@ -67,9 +68,43 @@ class Researcher:
|
||||
scored.sort(key=lambda x: x[0], reverse=True)
|
||||
return [c for _, c in scored[:top_k]]
|
||||
|
||||
def _llm_summary(self, cases: List[Dict]) -> str:
|
||||
if not HAVE_LLM:
|
||||
return ""
|
||||
cases_text = json.dumps(cases, ensure_ascii=False, indent=2)
|
||||
# 获取实时搜索数据作为 LLM 参考
|
||||
search_data = enrich_topic_research(self.topic)
|
||||
search_section = f"\n## 实时搜索结果\n{search_data}\n" if search_data else ""
|
||||
prompt = f"""你是一个行业研究员。基于以下选题和相关案例，写一段研究发现。
|
||||
|
||||
## 选题
|
||||
标题：{self.topic['title']}
|
||||
领域：{self.topic.get('field', '')}
|
||||
核心观点：{self.topic.get('core_concept', '')}
|
||||
受众痛点：{self.topic.get('audience_pain', '')}
|
||||
独特视角：{self.topic.get('unique_angle', '')}
|
||||
{search_section}
|
||||
## 相关案例（{len(cases)}个）
|
||||
{cases_text}
|
||||
|
||||
## 相关案例（{len(cases)}个）
|
||||
{cases_text}
|
||||
|
||||
## 要求
|
||||
- 提炼2-3个真正有价值的洞察，不是每个案例都硬凑一条
|
||||
- 每条洞察1-2句话，说人话，不要列1/2/3
|
||||
- 避免「首先其次最后」「综上所述」
|
||||
- 指出1-2个你不太确定的方向，作为后续研究的提示"""
|
||||
try:
|
||||
return call_llm(prompt, temperature=0.5, max_tokens=1200, system_prompt="你是一个行业研究员，擅长从案例中发现真洞察。")
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM 研究发现摘要生成失败: {e}")
|
||||
return ""
|
||||
|
||||
def generate_notes(self) -> str:
|
||||
"""生成研究笔记 Markdown"""
|
||||
cases = self.find_relevant_cases()
|
||||
trend_context = get_trend_context(self.topic.get('field'))
|
||||
search_data = enrich_topic_research(self.topic)
|
||||
lines = [
|
||||
f"# 研究笔记：{self.topic['title']}",
|
||||
f"\n## 选题信息",
|
||||
@@ -78,6 +113,8 @@ class Researcher:
|
||||
f"- **核心观点**: {self.topic.get('core_concept', '待补充')}",
|
||||
f"- **受众痛点**: {self.topic.get('audience_pain', '待补充')}",
|
||||
f"- **独特视角**: {self.topic.get('unique_angle', '待补充')}",
|
||||
f"\n{trend_context}",
|
||||
f"\n{search_data}" if search_data else "",
|
||||
f"\n## 相关案例（{len(cases)}个）\n"
|
||||
]
|
||||
for i, case in enumerate(cases, 1):
|
||||
@@ -89,17 +126,47 @@ class Researcher:
|
||||
f"- **关键数据**: {case.get('key_metrics', '无')}",
|
||||
""
|
||||
])
|
||||
lines.extend([
|
||||
"## 研究发现摘要",
|
||||
"- 待补充：从案例中提炼的趋势和洞察",
|
||||
"- 待补充：数据支撑",
|
||||
"",
|
||||
"## 待深入研究的问题",
|
||||
"- [ ] 需要更多本土数据",
|
||||
"- [ ] 需要验证某些结论的适用性",
|
||||
"",
|
||||
f"*生成时间：{TODAY}*"
|
||||
])
|
||||
|
||||
llm_summary = self._llm_summary(cases)
|
||||
if llm_summary:
|
||||
lines.extend([
|
||||
"## 研究发现摘要（LLM 生成）",
|
||||
llm_summary,
|
||||
""
|
||||
])
|
||||
else:
|
||||
insights = []
|
||||
for case in cases[:3]:
|
||||
summary = case.get('summary', case.get('description', ''))
|
||||
metrics = case.get('key_metrics', '')
|
||||
if summary:
|
||||
insight = f"- {case.get('title', '相关案例')}：{summary[:100]}"
|
||||
if metrics:
|
||||
insight += f"（{metrics[:80]}）"
|
||||
insights.append(insight)
|
||||
if not insights:
|
||||
insights = ["- 暂未匹配到高度相关的历史案例"]
|
||||
pain_text = self.topic.get('audience_pain', '')
|
||||
topic_insights = [f"- {pain_text[:100]}"] if pain_text else []
|
||||
lines.extend([
|
||||
"## 研究发现摘要",
|
||||
"",
|
||||
])
|
||||
lines.extend(insights)
|
||||
if topic_insights:
|
||||
lines.extend(topic_insights)
|
||||
lines.extend([
|
||||
"",
|
||||
"## 待深入研究的问题",
|
||||
])
|
||||
if cases:
|
||||
lines.append("- [ ] 验证以上案例在当前选题背景下的适用性")
|
||||
lines.extend([
|
||||
"- [ ] 收集更多本土一手数据",
|
||||
"- [ ] 确认目标受众的实际反馈",
|
||||
""
|
||||
])
|
||||
lines.append(f"*生成时间：{TODAY}*")
|
||||
return "\n".join(lines)
|
||||
|
||||
def save(self):
|
||||
|
||||
@@ -1,10 +1,12 @@
|
||||
#!/usr/bin/env python3
|
||||
import json
|
||||
data = json.load(open('automation/data/sustainability_topics.json', 'r', encoding='utf-8'))
|
||||
for t in data:
|
||||
if t['id'] in ['B05', 'D01']:
|
||||
t['status'] = '待处理'
|
||||
if 'ready_at' in t:
|
||||
del t['ready_at']
|
||||
json.dump(data, open('automation/data/sustainability_topics.json', 'w', encoding='utf-8'), ensure_ascii=False, indent=2)
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
|
||||
from db_helper import update_topic_status
|
||||
|
||||
for tid in ['B05', 'D01']:
|
||||
update_topic_status(tid, 'pending')
|
||||
print('已重置 B05, D01 为待处理')
|
||||
|
||||
+9
-53
@@ -1,68 +1,24 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
重置指定选题状态为「待处理」（数据库 + JSON 备份）
|
||||
"""
|
||||
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
|
||||
try:
|
||||
from db_helper import update_topic_status, get_topic_by_id
|
||||
HAVE_DB = True
|
||||
except ImportError:
|
||||
HAVE_DB = False
|
||||
print("Warning: db_helper not available, will only update JSON")
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TOPICS_FILE = DATA_DIR / "sustainability_topics.json"
|
||||
|
||||
def reset_json_status(ids):
|
||||
try:
|
||||
with open(TOPICS_FILE, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
except:
|
||||
topics = []
|
||||
updated_ids = []
|
||||
for t in topics:
|
||||
if t['id'] in ids:
|
||||
t['status'] = 'pending'
|
||||
if 'ready_at' in t:
|
||||
del t['ready_at']
|
||||
updated_ids.append(t['id'])
|
||||
with open(TOPICS_FILE, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
return updated_ids
|
||||
|
||||
def reset_db_status(ids):
|
||||
if not HAVE_DB:
|
||||
return []
|
||||
updated = []
|
||||
for tid in ids:
|
||||
if update_topic_status(tid, 'pending'):
|
||||
updated.append(tid)
|
||||
return updated
|
||||
from db_helper import update_topic_status, get_topic_by_id
|
||||
|
||||
def main():
|
||||
# 指定要重置的ID列表
|
||||
target_ids = ['D01', 'B05'] # 可修改
|
||||
target_ids = ['D01', 'B05']
|
||||
print(f"正在重置选题状态: {target_ids}")
|
||||
|
||||
# 更新数据库
|
||||
db_updated = reset_db_status(target_ids) if HAVE_DB else []
|
||||
if db_updated:
|
||||
print(f"[DB] 已重置: {db_updated}")
|
||||
updated = []
|
||||
for tid in target_ids:
|
||||
if update_topic_status(tid, 'pending'):
|
||||
updated.append(tid)
|
||||
if updated:
|
||||
print(f"[DB] 已重置: {updated}")
|
||||
else:
|
||||
print("[DB] 未更新或数据库不可用")
|
||||
|
||||
# 更新 JSON 备份
|
||||
json_updated = reset_json_status(target_ids)
|
||||
print(f"[JSON] 已重置: {json_updated}")
|
||||
|
||||
print("[DB] 无更新")
|
||||
print("完成")
|
||||
|
||||
if __name__ == "__main__":
|
||||
import json
|
||||
main()
|
||||
|
||||
@@ -1,46 +1,50 @@
|
||||
#!/usr/bin/env python3
|
||||
import json, datetime
|
||||
import sys, datetime
|
||||
from pathlib import Path
|
||||
from collections import Counter
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
data_dir = PROJECT_ROOT / "automation" / "data"
|
||||
releases_dir = data_dir / "releases" / "2026-04-16"
|
||||
drafts_dir = data_dir / "drafts" / "2026-04-16"
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
|
||||
# 1. 选题状态
|
||||
topics = json.load(open(data_dir / "sustainability_topics.json", encoding='utf-8'))
|
||||
from db_helper import export_topics_to_json
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TODAY = datetime.date.today().isoformat()
|
||||
releases_dir = DATA_DIR / "releases" / TODAY
|
||||
drafts_dir = DATA_DIR / "drafts" / TODAY
|
||||
|
||||
topics = export_topics_to_json()
|
||||
by_status = {}
|
||||
for t in topics:
|
||||
s = t.get('status','待处理')
|
||||
s = t.get('status', '待处理')
|
||||
by_status.setdefault(s, []).append(t)
|
||||
|
||||
print(f"📊 宇之然内容生成系统状态报告 ({datetime.date.today()})")
|
||||
print(f"\U0001f4ca 宇之然内容生成系统状态报告 ({TODAY})")
|
||||
print(f"\n=== 1. 选题库总览 ===")
|
||||
print(f"总选题数: {len(topics)}")
|
||||
for s, arr in sorted(by_status.items()):
|
||||
print(f" {s}: {len(arr)} 个")
|
||||
|
||||
# 2. 今日生成内容
|
||||
print(f"\n=== 2. 今日 (2026-04-16) 已生成内容 ===")
|
||||
print(f"\n=== 2. 今日 ({TODAY}) 已生成内容 ===")
|
||||
if releases_dir.exists():
|
||||
zhihu = list((releases_dir / 'zhihu').glob('*.html'))
|
||||
wechat = list((releases_dir / 'wechat').glob('*.html'))
|
||||
xhs = list((releases_dir / 'xiaohongshu').glob('*.html'))
|
||||
zhihu = list((releases_dir / 'zhihu').glob('*.html')) if (releases_dir / 'zhihu').exists() else []
|
||||
wechat = list((releases_dir / 'wechat').glob('*.html')) if (releases_dir / 'wechat').exists() else []
|
||||
xhs = list((releases_dir / 'xiaohongshu').glob('*.html')) if (releases_dir / 'xiaohongshu').exists() else []
|
||||
print(f" 知乎: {len(zhihu)} 篇")
|
||||
print(f" 微信公众号: {len(wechat)} 篇")
|
||||
print(f" 小红书: {len(xhs)} 篇")
|
||||
# 列出选题ID
|
||||
topic_ids = set()
|
||||
for f in zhihu:
|
||||
topic_ids.add(f.stem.split('_')[1])
|
||||
print(f" 涉及选题ID: {', '.join(sorted(topic_ids))}")
|
||||
if topic_ids:
|
||||
print(f" 涉及选题ID: {', '.join(sorted(topic_ids))}")
|
||||
else:
|
||||
print(" 今日无发布内容")
|
||||
|
||||
# 3. 合规与优化结果
|
||||
report_file = drafts_dir / "optimization_report.json"
|
||||
if report_file.exists():
|
||||
report = json.load(open(report_file, encoding='utf-8'))
|
||||
import json
|
||||
report = json.loads(report_file.read_text(encoding='utf-8'))
|
||||
print(f"\n=== 3. 合规优化结果 ===")
|
||||
sm = report['summary']
|
||||
print(f" 总文章数: {sm['total_articles']}")
|
||||
@@ -50,14 +54,13 @@ if report_file.exists():
|
||||
if sm['need_manual'] == 0:
|
||||
print(" ✅ 所有文章均已自动合规")
|
||||
else:
|
||||
print("\n=== 3. 合规优化结果 ===")
|
||||
print(f"\n=== 3. 合规优化结果 ===")
|
||||
print(" 未找到优化报告")
|
||||
|
||||
# 4. 待发布选题（可发布）
|
||||
ready = by_status.get('ready', [])
|
||||
print(f"\n=== 4. 待发布选题（已合规）===")
|
||||
ready = by_status.get('待发布', [])
|
||||
if ready:
|
||||
for t in sorted(ready, key=lambda x: x.get('priority_score',0), reverse=True):
|
||||
for t in sorted(ready, key=lambda x: x.get('priority_score', 0), reverse=True):
|
||||
print(f" {t['id']}: {t['title'][:50]}")
|
||||
else:
|
||||
print(" 暂无待发布选题")
|
||||
|
||||
+10
-7
@@ -1,19 +1,22 @@
|
||||
#!/usr/bin/env python3
|
||||
import json
|
||||
import sys
|
||||
from collections import Counter
|
||||
from pathlib import Path
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent))
|
||||
from db_helper import export_topics_to_json
|
||||
|
||||
with open('automation/data/sustainability_topics.json', 'r', encoding='utf-8') as f:
|
||||
data = json.load(f)
|
||||
topics = export_topics_to_json()
|
||||
|
||||
print('=== 选题库状态 ===')
|
||||
print(f'总选题数: {len(data)}')
|
||||
print(f'字段: {list(data[0].keys())}')
|
||||
print(f'总选题数: {len(topics)}')
|
||||
if topics:
|
||||
print(f'字段: {list(topics[0].keys())}')
|
||||
|
||||
print('\n状态分布:')
|
||||
status_counts = Counter(t.get('status', '<无>') for t in data)
|
||||
status_counts = Counter(t.get('status', '<无>') for t in topics)
|
||||
for s, c in sorted(status_counts.items()):
|
||||
print(f' {s}: {c} 个')
|
||||
|
||||
print('\n各状态详情:')
|
||||
for t in data:
|
||||
for t in topics:
|
||||
print(f"{t['id']}: {t['title'][:40]:40} | 状态: {t.get('status', '?'):6} | 优先级: {t.get('priority_score', '-')}")
|
||||
|
||||
+6
-15
@@ -1,19 +1,10 @@
|
||||
#!/usr/bin/env python3
|
||||
import sys
|
||||
from pathlib import Path
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent))
|
||||
from db_helper import get_topic_by_id, get_topics_by_status
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
topics_file = DATA_DIR / "sustainability_topics.json"
|
||||
|
||||
print(f"Project root: {PROJECT_ROOT}")
|
||||
print(f"Looking for: {topics_file}")
|
||||
print(f"Exists: {topics_file.exists()}")
|
||||
|
||||
if topics_file.exists():
|
||||
import json
|
||||
with open(topics_file, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
print(f"Loaded {len(topics)} topics")
|
||||
if topics:
|
||||
print("First topic:", topics[0]['title'], f"score={topics[0]['priority_score']}, status={topics[0]['status']}")
|
||||
pending = get_topics_by_status('pending')
|
||||
print(f"待处理选题数: {len(pending)}")
|
||||
for t in pending[:3]:
|
||||
print(f" {t['id']}: {t['title'][:40]}, score={t['priority_score']}")
|
||||
|
||||
@@ -0,0 +1,189 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
选题引擎 v2：多趋势加权匹配 + 趋势缺口检测 + 新选题生成
|
||||
"""
|
||||
|
||||
import sys, json, logging
|
||||
from pathlib import Path
|
||||
from typing import Dict, List, Optional
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
sys.path.insert(0, str(PROJECT_ROOT / "platform" / "backend"))
|
||||
|
||||
from trends import load_trends
|
||||
from db_helper import export_topics_to_json, update_topic_status, save_topics_to_db
|
||||
from app.core.nvidia_client import call_llm
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TRENDS_FILE = DATA_DIR / "trends.json"
|
||||
TODAY = __import__('datetime').datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
TREND_DOMAIN_MAP = {
|
||||
"远程工作": "未来工作方式",
|
||||
"AI工具": "AI与效率",
|
||||
"可持续生活": "可持续生活系统",
|
||||
"知识管理": "个人知识工厂",
|
||||
"数字生活": "科技人文交叉",
|
||||
"科技人文": "科技人文交叉",
|
||||
"个人成长": "个人成长",
|
||||
"副业": "个人成长",
|
||||
"AI创作": "AI与效率",
|
||||
"未来工作": "未来工作方式",
|
||||
"效率工具": "AI与效率",
|
||||
"家庭教育": "科技人文交叉",
|
||||
}
|
||||
|
||||
def _topic_trend_score(topic: Dict, trend: Dict) -> float:
|
||||
field = (topic.get("field") or "").lower()
|
||||
title = (topic.get("title") or "").lower()
|
||||
core = (topic.get("core_concept") or "").lower()
|
||||
trend_domain = TREND_DOMAIN_MAP.get(trend.get("domain", ""), "")
|
||||
keywords = [trend.get("topic", "")] + trend.get("hot_keywords", [])
|
||||
score = 0.0
|
||||
if trend_domain and trend_domain in field:
|
||||
score += 5
|
||||
for kw in keywords:
|
||||
kw = kw.lower()
|
||||
if kw in title: score += 3
|
||||
elif kw in core: score += 2
|
||||
return score
|
||||
|
||||
def match_trends_to_topics() -> List[Dict]:
|
||||
trends = load_trends()
|
||||
topics = [t for t in export_topics_to_json() if t.get('status') in ('pending', '待处理')]
|
||||
if not trends or not topics:
|
||||
return []
|
||||
|
||||
topic_scores = {}
|
||||
for topic in topics:
|
||||
tid = topic["id"]
|
||||
per_trend = []
|
||||
total = 0.0
|
||||
for trend in trends:
|
||||
s = _topic_trend_score(topic, trend)
|
||||
if s > 0:
|
||||
per_trend.append({"trend": trend["topic"], "domain": trend.get("domain",""), "score": s})
|
||||
total += s
|
||||
if per_trend:
|
||||
topic_scores[tid] = {
|
||||
"topic_id": tid,
|
||||
"topic_title": topic["title"],
|
||||
"total_score": total + (topic.get("priority_score", 0) or 0) * 0.3,
|
||||
"matched_trends": sorted(per_trend, key=lambda x: -x["score"]),
|
||||
"matched_count": len(per_trend),
|
||||
}
|
||||
|
||||
results = sorted(topic_scores.values(), key=lambda x: -x["total_score"])
|
||||
return results
|
||||
|
||||
def detect_trend_gaps() -> List[Dict]:
|
||||
trends = load_trends()
|
||||
topics = export_topics_to_json()
|
||||
gaps = []
|
||||
for trend in trends:
|
||||
matched = False
|
||||
trend_keywords = [trend.get("topic", "").lower()] + [k.lower() for k in trend.get("hot_keywords", [])]
|
||||
for t in topics:
|
||||
title = (t.get("title") or "").lower()
|
||||
core = (t.get("core_concept") or "").lower()
|
||||
for kw in trend_keywords:
|
||||
if kw in title or kw in core:
|
||||
matched = True
|
||||
break
|
||||
if matched:
|
||||
break
|
||||
if not matched:
|
||||
gaps.append(trend)
|
||||
return gaps
|
||||
|
||||
def generate_new_topics(gaps: List[Dict]) -> List[Dict]:
|
||||
if not gaps:
|
||||
return []
|
||||
prompt = f"""以下热点话题在我们的选题库中没有匹配项。请为每个热点生成一个新选题建议。
|
||||
|
||||
热点列表：
|
||||
{chr(10).join(f'- {g["topic"]}（{g.get("domain","")}，{g.get("reason","")}）' for g in gaps)}
|
||||
|
||||
输出 JSON 数组，每个元素包含：
|
||||
- "title": 选题标题（有吸引力，20字内）
|
||||
- "field": 所属领域
|
||||
- "core_concept": 核心观点（一句话）
|
||||
- "audience_pain": 受众痛点
|
||||
- "unique_angle": 独特视角
|
||||
|
||||
只输出 JSON，不要其他文字。"""
|
||||
try:
|
||||
resp = call_llm(prompt, temperature=0.4, max_tokens=2000)
|
||||
resp = resp.strip()
|
||||
if resp.startswith("```"):
|
||||
resp = resp.split("\n", 1)[1].rsplit("\n", 1)[0]
|
||||
suggestions = json.loads(resp)
|
||||
if isinstance(suggestions, list):
|
||||
return suggestions
|
||||
except Exception as e:
|
||||
logger.warning(f"新选题生成失败: {e}")
|
||||
return []
|
||||
|
||||
def suggest_new_topics() -> List[Dict]:
|
||||
gaps = detect_trend_gaps()
|
||||
if not gaps:
|
||||
logger.info("所有趋势均有匹配选题，无需新增")
|
||||
return []
|
||||
logger.info(f"发现 {len(gaps)} 个趋势缺口: {[g['topic'] for g in gaps]}")
|
||||
new_topics = generate_new_topics(gaps)
|
||||
if new_topics:
|
||||
logger.info(f"生成 {len(new_topics)} 个新选题建议")
|
||||
report_path = DATA_DIR / "topic_suggestions.json"
|
||||
data = {"date": TODAY, "gaps": gaps, "suggestions": new_topics}
|
||||
DATA_DIR.mkdir(parents=True, exist_ok=True)
|
||||
report_path.write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
|
||||
return new_topics
|
||||
|
||||
def select_best_topic() -> Optional[Dict]:
|
||||
matches = match_trends_to_topics()
|
||||
if matches:
|
||||
best = matches[0]
|
||||
logger.info(f"选题推荐: {best['topic_id']} {best['topic_title']} "
|
||||
f"(命中 {best['matched_count']} 个趋势, 综合分: {best['total_score']:.1f})")
|
||||
topic = __import__('db_helper').get_topic_by_id(best['topic_id'])
|
||||
if topic:
|
||||
return topic
|
||||
|
||||
from db_helper import get_next_topic
|
||||
fallback = get_next_topic(priority='高') or get_next_topic()
|
||||
if fallback:
|
||||
logger.info(f"无趋势匹配，退回到优先级最高选题: {fallback['id']}")
|
||||
return fallback
|
||||
|
||||
def print_report():
|
||||
matches = match_trends_to_topics()
|
||||
trends = load_trends()
|
||||
gaps = detect_trend_gaps()
|
||||
print(f"=== 选题匹配报告 ({TODAY}) ===\n")
|
||||
print(f"当前趋势数: {len(trends)}")
|
||||
print(f"待处理选题: {len([t for t in export_topics_to_json() if t.get('status') in ('pending', '待处理')])}")
|
||||
print(f"趋势匹配数: {len(matches)}")
|
||||
print(f"趋势缺口: {len(gaps)}\n")
|
||||
|
||||
if matches:
|
||||
print(f"{'排名':>4} {'选题ID':6} {'综合分':>6} {'命中趋势':>6} {'趋势详情':30} {'选题标题'}")
|
||||
print("-" * 100)
|
||||
for i, m in enumerate(matches[:10], 1):
|
||||
trends_str = ", ".join(t["trend"][:10] for t in m["matched_trends"][:3])
|
||||
print(f"{i:>4} {m['topic_id']:6} {m['total_score']:>6.1f} {m['matched_count']:>6} {trends_str:30} {m['topic_title'][:40]}")
|
||||
|
||||
if gaps:
|
||||
print(f"\n=== 趋势缺口（无匹配选题）===")
|
||||
for g in gaps:
|
||||
print(f" ⚠️ {g['topic']}（{g.get('domain','')}）— {g.get('reason','')}")
|
||||
|
||||
if __name__ == "__main__":
|
||||
logging.basicConfig(level=logging.INFO)
|
||||
print_report()
|
||||
print()
|
||||
new = suggest_new_topics()
|
||||
if new:
|
||||
print(f"\n新选题建议已保存到 automation/data/topic_suggestions.json")
|
||||
@@ -0,0 +1,103 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
热点趋势感知模块
|
||||
- LLM 生成当前领域热点话题
|
||||
- 可扩展接入外部热搜 API
|
||||
"""
|
||||
|
||||
import json, datetime, logging, sys
|
||||
from pathlib import Path
|
||||
from typing import List, Dict
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
sys.path.insert(0, str(PROJECT_ROOT / "platform" / "backend"))
|
||||
|
||||
from app.core.nvidia_client import call_llm
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TRENDS_FILE = DATA_DIR / "trends.json"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
logging.basicConfig(
|
||||
level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s',
|
||||
handlers=[logging.FileHandler(LOGS_DIR / f"trends_{TODAY}.log"), logging.StreamHandler()]
|
||||
)
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
DOMAINS = ["远程工作", "AI工具", "可持续生活", "知识管理", "数字生活", "科技人文"]
|
||||
|
||||
def fetch_llm_trends() -> List[Dict]:
|
||||
prompt = f"""你是社交媒体趋势分析师。列出今天（2026年5月）中文互联网上最热的10个话题，要求：
|
||||
|
||||
1. 覆盖以下领域：{', '.join(DOMAINS)}
|
||||
2. 每个话题包含：领域、话题名称、热度原因（1句话）、相关热搜词（3个）
|
||||
3. 优先选择在知乎/微博/小红书上有讨论度的话题
|
||||
4. 输出 JSON 数组，格式：
|
||||
[{{"domain": "领域", "topic": "话题名", "reason": "热度原因", "hot_keywords": ["词1","词2","词3"], "platform": "知乎/微博/小红书"}}]
|
||||
|
||||
只输出 JSON，不要其他文字。"""
|
||||
try:
|
||||
resp = call_llm(prompt, temperature=0.4, max_tokens=2000)
|
||||
resp = resp.strip()
|
||||
if resp.startswith("```"):
|
||||
resp = resp.split("\n", 1)[1].rsplit("\n", 1)[0]
|
||||
trends = json.loads(resp)
|
||||
if isinstance(trends, list):
|
||||
return trends
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM 趋势获取失败: {e}")
|
||||
return []
|
||||
|
||||
def save_trends(trends: List[Dict]):
|
||||
data = {
|
||||
"date": TODAY,
|
||||
"updated_at": datetime.datetime.now().isoformat(),
|
||||
"trends": trends
|
||||
}
|
||||
DATA_DIR.mkdir(parents=True, exist_ok=True)
|
||||
TRENDS_FILE.write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding='utf-8')
|
||||
logger.info(f"趋势数据已保存: {len(trends)} 条")
|
||||
|
||||
def load_trends() -> List[Dict]:
|
||||
if TRENDS_FILE.exists():
|
||||
try:
|
||||
data = json.loads(TRENDS_FILE.read_text(encoding='utf-8'))
|
||||
if data.get("date") == TODAY:
|
||||
return data.get("trends", [])
|
||||
except:
|
||||
pass
|
||||
return []
|
||||
|
||||
def get_trending_topics(domain: str = None, top_k: int = 5) -> List[Dict]:
|
||||
trends = load_trends()
|
||||
if domain:
|
||||
trends = [t for t in trends if domain in t.get("domain", "") or t.get("domain", "") in domain]
|
||||
return trends[:top_k]
|
||||
|
||||
def get_trend_context(domain: str = None) -> str:
|
||||
trends = get_trending_topics(domain, top_k=3)
|
||||
if not trends:
|
||||
return "暂无趋势数据"
|
||||
lines = ["## 当前热点趋势", ""]
|
||||
for t in trends:
|
||||
keywords = ", ".join(t.get("hot_keywords", []))
|
||||
lines.append(f"- **{t['topic']}**（{t.get('platform','')}）：{t.get('reason','')}")
|
||||
if keywords:
|
||||
lines.append(f" 热搜词：{keywords}")
|
||||
return "\n".join(lines)
|
||||
|
||||
def main():
|
||||
logger.info("开始获取热点趋势...")
|
||||
trends = fetch_llm_trends()
|
||||
if trends:
|
||||
save_trends(trends)
|
||||
for t in trends:
|
||||
print(f" [{t.get('domain','?')}] {t['topic']} — {t.get('platform','')}")
|
||||
else:
|
||||
print("未获取到趋势数据")
|
||||
print(f"完成，共 {len(trends)} 条")
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,6 +1,10 @@
|
||||
#!/usr/bin/env python3
|
||||
import json
|
||||
data = json.load(open('automation/data/sustainability_topics.json', 'r', encoding='utf-8'))
|
||||
print(f"Total topics: {len(data)}")
|
||||
for t in data[:5]:
|
||||
import sys
|
||||
from pathlib import Path
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent))
|
||||
from db_helper import export_topics_to_json
|
||||
|
||||
topics = export_topics_to_json()
|
||||
print(f"Total topics: {len(topics)}")
|
||||
for t in topics[:5]:
|
||||
print(f"- {t.get('id')}: {t.get('title','')[:40]}")
|
||||
|
||||
@@ -0,0 +1,152 @@
|
||||
#!/usr/bin/env python3
|
||||
"""实时搜索模块：使用 Bing 搜索获取真实结果，为研究提供数据支撑"""
|
||||
|
||||
import sys, json, logging, re
|
||||
from pathlib import Path
|
||||
from typing import List, Dict
|
||||
from datetime import datetime
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
CACHE_FILE = DATA_DIR / "search_cache.json"
|
||||
CACHE_TTL = 3600 * 6
|
||||
_skip_domains = ['baike.baidu.com', 'zdic.net', 'hanyu.baidu.com', 'dict.cn']
|
||||
|
||||
_cache = None
|
||||
|
||||
def _load_cache() -> Dict:
|
||||
global _cache
|
||||
if _cache is not None:
|
||||
return _cache
|
||||
if CACHE_FILE.exists():
|
||||
try:
|
||||
_cache = json.loads(CACHE_FILE.read_text(encoding='utf-8'))
|
||||
return _cache
|
||||
except:
|
||||
pass
|
||||
_cache = {}
|
||||
return _cache
|
||||
|
||||
def _save_cache():
|
||||
global _cache
|
||||
CACHE_FILE.parent.mkdir(parents=True, exist_ok=True)
|
||||
CACHE_FILE.write_text(json.dumps(_cache, ensure_ascii=False, indent=2), encoding='utf-8')
|
||||
|
||||
def _is_relevant(result: Dict, query: str) -> bool:
|
||||
url = result.get('url', '')
|
||||
if any(d in url for d in _skip_domains):
|
||||
return False
|
||||
text = result.get('title', '') + result.get('snippet', '')
|
||||
query_words = [w for w in re.split(r'[\s,，]+', query) if len(w) >= 2]
|
||||
if not query_words:
|
||||
return True
|
||||
match_count = sum(1 for w in query_words if w in text)
|
||||
return match_count >= max(1, int(len(query_words) * 0.3))
|
||||
|
||||
def search(query: str, max_results: int = 5, use_cache: bool = True) -> List[Dict]:
|
||||
cache_key = f"{query}_{max_results}"
|
||||
if use_cache:
|
||||
cache = _load_cache()
|
||||
if cache_key in cache:
|
||||
entry = cache[cache_key]
|
||||
if datetime.now().timestamp() - entry.get('ts', 0) < CACHE_TTL:
|
||||
logger.info(f"缓存命中: {query[:30]}")
|
||||
return entry['results']
|
||||
try:
|
||||
from bs4 import BeautifulSoup
|
||||
import requests
|
||||
headers = {
|
||||
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36',
|
||||
'Accept-Language': 'zh-CN,zh;q=0.9,en;q=0.8',
|
||||
}
|
||||
url = 'https://cn.bing.com/search?q=' + requests.utils.quote(query) + '&setlang=zh-cn&cc=cn'
|
||||
r = requests.get(url, headers=headers, timeout=15)
|
||||
soup = BeautifulSoup(r.text, 'html.parser')
|
||||
results = []
|
||||
for li in soup.find_all('li', class_='b_algo'):
|
||||
if len(results) >= max_results:
|
||||
break
|
||||
h2 = li.find('h2')
|
||||
if not h2:
|
||||
continue
|
||||
a = h2.find('a')
|
||||
if not a or not a.get('href'):
|
||||
continue
|
||||
title = a.get_text(strip=True)
|
||||
href = a['href']
|
||||
p = li.find('p')
|
||||
snippet = p.get_text(strip=True) if p else ''
|
||||
if title and href and not href.startswith('javascript'):
|
||||
r = {'title': title, 'url': href, 'snippet': snippet[:200]}
|
||||
if _is_relevant(r, query):
|
||||
results.append(r)
|
||||
if use_cache:
|
||||
_load_cache()
|
||||
_cache[cache_key] = {'ts': datetime.now().timestamp(), 'results': results}
|
||||
_save_cache()
|
||||
logger.info(f"搜索完成: {query[:40]} -> {len(results)}条")
|
||||
return results
|
||||
except Exception as e:
|
||||
logger.warning(f"搜索失败: {query[:30]} -> {e}")
|
||||
return []
|
||||
|
||||
def _extract_keywords(text: str, max_words: int = 4) -> str:
|
||||
sep = r'[：:，。！？\s()（）""\u201c\u201d#+\-*\d]'
|
||||
words = [w.strip() for w in re.split(sep, text) if len(w.strip()) >= 2]
|
||||
seen = set()
|
||||
result = []
|
||||
for w in words:
|
||||
if w not in seen:
|
||||
seen.add(w)
|
||||
result.append(w)
|
||||
return ' '.join(result[:max_words])
|
||||
|
||||
def _make_queries(topic: Dict) -> List[str]:
|
||||
title = topic.get('title', '')
|
||||
field = topic.get('field', '')
|
||||
core = topic.get('core_concept', '')
|
||||
pain = topic.get('audience_pain', '')
|
||||
queries = []
|
||||
kw = _extract_keywords(title)
|
||||
if kw:
|
||||
queries.append(kw)
|
||||
core_kw = _extract_keywords(core, 3)
|
||||
if core_kw and core_kw != kw:
|
||||
queries.append(core_kw)
|
||||
if field and kw:
|
||||
queries.append(field + ' ' + kw.split()[0] if kw.split() else field)
|
||||
if pain:
|
||||
pain_kw = _extract_keywords(pain, 3)
|
||||
if pain_kw and pain_kw not in queries:
|
||||
queries.append(pain_kw)
|
||||
if kw:
|
||||
queries.append(kw + ' 2025 2026')
|
||||
queries.append(kw + ' 案例')
|
||||
return queries
|
||||
|
||||
def enrich_topic_research(topic: Dict) -> str:
|
||||
queries = _make_queries(topic)
|
||||
seen_urls = set()
|
||||
all_results = []
|
||||
for q in queries:
|
||||
for r in search(q, max_results=3):
|
||||
if r['url'] not in seen_urls:
|
||||
seen_urls.add(r['url'])
|
||||
all_results.append(r)
|
||||
if not all_results:
|
||||
return ""
|
||||
lines = ["\n## 实时搜索数据\n"]
|
||||
for r in all_results[:6]:
|
||||
lines.append("- **" + r['title'] + "**")
|
||||
lines.append(" " + r['url'])
|
||||
if r.get('snippet'):
|
||||
lines.append(" > " + r['snippet'][:200])
|
||||
lines.append("")
|
||||
return '\n'.join(lines)
|
||||
|
||||
if __name__ == "__main__":
|
||||
logging.basicConfig(level=logging.INFO)
|
||||
results = search("Obsidian AI 第二大脑 搭建")
|
||||
print(json.dumps(results, ensure_ascii=False, indent=2))
|
||||
+298
-87
@@ -1,32 +1,20 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
撰写阶段：基于大纲和选题生成完整文章（三平台版本）- 数据库版
|
||||
"""
|
||||
|
||||
import json, datetime, logging, sys, re, subprocess
|
||||
import json, datetime, logging, sys, re
|
||||
from pathlib import Path
|
||||
from typing import Dict, List
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
# 添加项目根和 backend 路径，以导入 app.core.llm_client
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
sys.path.insert(0, str(PROJECT_ROOT / 'platform' / 'backend'))
|
||||
|
||||
# 导入 LLM 客户端（NVIDIA）
|
||||
from app.database import SessionLocal
|
||||
from app.models import Topic
|
||||
from db_helper import get_topic_by_id, update_topic_status
|
||||
try:
|
||||
from app.core.modelscope_client import expand_content_with_llm # type: ignore
|
||||
HAVE_LLM = True # ModelScope
|
||||
except ImportError as e:
|
||||
logging.warning(f"LLM client unavailable: {e}")
|
||||
from app.core.nvidia_client import call_llm
|
||||
HAVE_LLM = True
|
||||
except ImportError:
|
||||
HAVE_LLM = False
|
||||
|
||||
# 导入数据库辅助模块
|
||||
from db_helper import get_topic_by_id, update_topic_status
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
import mistune
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
OUTLINE_DIR = DATA_DIR / "outlines"
|
||||
@@ -46,6 +34,23 @@ logging.basicConfig(
|
||||
)
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
_md_parser = mistune.create_markdown()
|
||||
|
||||
PLATFORM_CONFIG = {
|
||||
"zhihu": {
|
||||
"max_chars": 3000,
|
||||
"style": "深度长文分析",
|
||||
},
|
||||
"wechat": {
|
||||
"max_chars": 1500,
|
||||
"style": "亲切口语化",
|
||||
},
|
||||
"xiaohongshu": {
|
||||
"max_chars": 800,
|
||||
"style": "图文笔记，emoji+标签",
|
||||
},
|
||||
}
|
||||
|
||||
class Writer:
|
||||
def __init__(self, topic_id: str):
|
||||
self.topic_id = topic_id
|
||||
@@ -56,7 +61,6 @@ class Writer:
|
||||
self.outline_content = outline_file.read_text(encoding='utf-8')
|
||||
self.release_dir = RELEASE_DIR / TODAY
|
||||
self.release_dir.mkdir(parents=True, exist_ok=True)
|
||||
# 加载研究笔记（作为 LLM 上下文）
|
||||
research_file = DATA_DIR / "research" / TODAY / f"{topic_id}_research.md"
|
||||
self.research_notes = research_file.read_text(encoding='utf-8') if research_file.exists() else ""
|
||||
|
||||
@@ -67,16 +71,12 @@ class Writer:
|
||||
return topic
|
||||
|
||||
def _clean_title(self, title: str) -> str:
|
||||
"""去除标题中的指导性文字（如字数说明、MVP标记等）"""
|
||||
import re
|
||||
# 去掉括号中的说明：约200字、约300字、MVP、试行等
|
||||
title = re.sub(r'[（(]约\s*\d+字[）)]', '', title)
|
||||
title = re.sub(r'[（(]MVP[）)]', '', title)
|
||||
title = re.sub(r'[（(][^）)]*?[）)]', '', title) # 保守移除任意括号内容（可能误伤，但大纲通常不包含重要括号信息）
|
||||
title = re.sub(r'[（(][^）)]*?[）)]', '', title)
|
||||
return title.strip()
|
||||
|
||||
def _parse_outline_sections(self) -> List[Dict]:
|
||||
"""将大纲 Markdown 解析为结构化列表，保留层级和内容"""
|
||||
sections = []
|
||||
current = None
|
||||
for line in self.outline_content.splitlines():
|
||||
@@ -100,110 +100,321 @@ class Writer:
|
||||
return sections
|
||||
|
||||
def _expand_section(self, section: Dict) -> str:
|
||||
"""将大纲中的简短描述扩展为完整段落"""
|
||||
content = section.get('content', '').strip()
|
||||
# 如果有足够内容（>200字），直接返回
|
||||
if len(content) > 200:
|
||||
return content
|
||||
# 如果内容极少，需要 LLM 扩写
|
||||
if HAVE_LLM and len(content) < 150:
|
||||
logger.info(f"使用 LLM 扩写章节: {section['title']}")
|
||||
prompt = f"""你是一个真人写作者，在写一篇关于「{self.topic['title']}」的文章。现在要写「{section['title']}」这一节，你的笔记要点如下：
|
||||
|
||||
{content}
|
||||
|
||||
要求：
|
||||
- 200-300字，用自己的话展开
|
||||
- 用「你」或「我们」视角，不要用「我」
|
||||
- 读起来像人在自然说话，不是AI在组装文字
|
||||
- 避免「首先」「其次」「总的来说」「综上所述」这类套路表达
|
||||
- 如果合适，可以加一句反问
|
||||
|
||||
直接输出段落正文。"""
|
||||
try:
|
||||
expanded = expand_content_with_llm(
|
||||
topic=self.topic,
|
||||
section_title=section['title'],
|
||||
section_content=content,
|
||||
context=self.research_notes
|
||||
)
|
||||
expanded = call_llm(prompt, temperature=0.6, max_tokens=1500)
|
||||
if expanded and len(expanded.strip()) > len(content):
|
||||
return expanded.strip()
|
||||
else:
|
||||
logger.warning("LLM 扩写失败，返回占位")
|
||||
raise ValueError("Empty expansion")
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM 扩写失败: {e}，使用占位内容")
|
||||
# 返回占位内容，保持流程继续
|
||||
return f"{content}\n\n（本段内容需要人工补充：当前模型调用失败或未配置）"
|
||||
# 否则返回原内容
|
||||
logger.warning(f"LLM 扩写失败: {e}")
|
||||
|
||||
# Fallback: 将 bullet points 展开为段落
|
||||
lines = [l.strip() for l in content.split('\n') if l.strip()]
|
||||
if lines:
|
||||
sentences = []
|
||||
for line in lines:
|
||||
text = line.lstrip('- *').strip()
|
||||
if text:
|
||||
for prefix in ['- ', '* ', '1. ', '2. ', '3. ', '4. ', '5. ']:
|
||||
if line.startswith(prefix):
|
||||
text = line[len(prefix):]
|
||||
break
|
||||
if text[-1] not in '。！？；':
|
||||
text += '。'
|
||||
sentences.append(text)
|
||||
if sentences:
|
||||
return ' '.join(sentences)
|
||||
return content
|
||||
|
||||
def generate_full_markdown(self) -> str:
|
||||
"""根据大纲生成完整 Markdown 正文"""
|
||||
sections = self._parse_outline_sections()
|
||||
parts = []
|
||||
|
||||
for sec in sections:
|
||||
if sec['level'] == 1:
|
||||
continue
|
||||
heading = f"{'#' * sec['level']} {sec['title']}"
|
||||
parts.append(heading)
|
||||
if sec.get('content'):
|
||||
expanded = self._expand_section(sec)
|
||||
parts.append(expanded + "\n\n")
|
||||
|
||||
parts.append(expanded + "\n")
|
||||
full_md = "\n".join(parts).strip()
|
||||
return full_md
|
||||
|
||||
def generate_platform_html(self, markdown: str, platform: str) -> str:
|
||||
"""将 Markdown 转换为平台 HTML（基于模板）"""
|
||||
title = self.topic['title']
|
||||
def _adapt_for_platform(self, markdown: str, platform: str) -> str:
|
||||
cfg = PLATFORM_CONFIG[platform]
|
||||
|
||||
# 加载模板
|
||||
platform_prompts = {
|
||||
"zhihu": f"""你是一个知乎答主。把以下文章改写成知乎回答。
|
||||
|
||||
## 知乎回答的特点
|
||||
- 知乎用户习惯理性分析、数据支撑、逻辑递进
|
||||
- 开头直击问题本质，不绕弯子
|
||||
- 每一段讲一个观点，段与段之间有逻辑推进
|
||||
|
||||
## 改写要求
|
||||
- 用第三人称或「我们」视角，不要用「我」
|
||||
- 不要编个人经历——知乎读者在意的是分析质量，不是故事
|
||||
- 开头可以抛数据、抛现象、抛一个矛盾点
|
||||
- 不用"首先其次最后"，用自然过渡
|
||||
- 避免AI感表达：「总的来说」「综上所述」「值得注意的是」
|
||||
- 字数：{cfg['max_chars']}字以内
|
||||
|
||||
## 原文
|
||||
{markdown[:3000]}
|
||||
|
||||
## 输出
|
||||
直接输出改写后的完整内容（仅正文），每段后空一行。""",
|
||||
|
||||
"wechat": f"""你是一个公众号作者。把以下文章改写成公众号推文。
|
||||
|
||||
## 公众号推文的特点
|
||||
- 开头必须制造共鸣或好奇心，让读者愿意往下读
|
||||
- 短段落，有节奏感，每段2-3行
|
||||
- 核心观点加粗突出
|
||||
|
||||
## 改写要求
|
||||
- 用「你」视角，**通篇不允许出现「我」字**
|
||||
- 把原文中所有的「我」改成「你」或「很多人」或「有人」
|
||||
- 结构可以和原文完全不同——公众号不需要全面的分析，抓住1-2个痛点打透就行
|
||||
- 可以删减原文内容，保留最有力的观点
|
||||
- 避免「综上所述」「值得注意的是」「换言之」
|
||||
- 字数：{cfg['max_chars']}字以内
|
||||
|
||||
## 原文
|
||||
{markdown[:3000]}
|
||||
|
||||
## 输出
|
||||
直接输出改写后的完整内容（仅正文），每段后空一行。""",
|
||||
|
||||
"xiaohongshu": f"""你是一个小红书用户。把以下文章改写成小红书笔记。
|
||||
|
||||
## 小红书笔记的特点
|
||||
- 极短！极短！极短！小红书用户没耐心看长文
|
||||
- 直接给结论、给清单、给步骤
|
||||
- 原文通常很深很全，但笔记只取最关键的2-3个点
|
||||
|
||||
## 改写要求
|
||||
- **全文控制在 {cfg['max_chars']} 字以内**，多一个字都不要
|
||||
- 用「你」视角，不要用「我」
|
||||
- 开头一句抓住注意力，可以是结论、可以是一个反常识的观点
|
||||
- 正文每段1-2句，可以完全打乱原文结构
|
||||
- emoji每人段最多1个点缀，不要堆砌（✨💡✅🔸）
|
||||
- 结尾加 #话题标签 3-5个
|
||||
- 原文的深度分析全部砍掉，只留最 actionable 的内容
|
||||
|
||||
## 原文
|
||||
{markdown[:3000]}
|
||||
|
||||
## 输出
|
||||
直接输出改写后的完整内容（仅正文）。""",
|
||||
}
|
||||
|
||||
if HAVE_LLM:
|
||||
prompt = platform_prompts.get(platform, f"""将以下文章改写为适合{platform}平台版本，{cfg['max_chars']}字以内：
|
||||
|
||||
{markdown[:3000]}""")
|
||||
try:
|
||||
adapted = call_llm(prompt, temperature=0.7, max_tokens=2000)
|
||||
if adapted and len(adapted.strip()) > 50:
|
||||
adapted = adapted.strip()
|
||||
# 字数强校验
|
||||
max_c = cfg['max_chars']
|
||||
if platform == "xiaohongshu" and len(adapted) > max_c * 1.2:
|
||||
adapted = adapted[:max_c]
|
||||
last_break = max(adapted.rfind('。'), adapted.rfind('\n'), adapted.rfind('！'), adapted.rfind('？'))
|
||||
if last_break > max_c // 2:
|
||||
adapted = adapted[:last_break + 1]
|
||||
# 微信/知乎清除残留的"我"
|
||||
if platform in ("wechat", "zhihu"):
|
||||
adapted = adapted.replace('我', '你')
|
||||
logger.info(f"LLM 平台适配完成: {platform} ({len(adapted)}字, \"我\"x{adapted.count('我')})")
|
||||
return adapted
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM {platform} 适配失败: {e}")
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM {platform} 适配失败: {e}")
|
||||
|
||||
if platform == "xiaohongshu":
|
||||
lines = markdown.split('\n')
|
||||
result = []
|
||||
char_count = 0
|
||||
for line in lines:
|
||||
if char_count >= cfg['max_chars']:
|
||||
break
|
||||
if line.startswith('## '):
|
||||
line = f"## ✨ {line[3:]}"
|
||||
elif line.startswith('### '):
|
||||
line = f"### 💡 {line[4:]}"
|
||||
|
||||
result.append(line)
|
||||
char_count += len(line)
|
||||
adapted = '\n'.join(result)
|
||||
if adapted.count('#') == 0:
|
||||
adapted = f"# {self.topic['title']}\n\n{adapted}"
|
||||
return adapted
|
||||
return markdown
|
||||
|
||||
def _get_platform_tags(self, platform: str) -> str:
|
||||
field = self.topic.get('field', '')
|
||||
title = self.topic.get('title', '')
|
||||
core = self.topic.get('core_concept', '')
|
||||
|
||||
tag_prompts = {
|
||||
"zhihu": f"根据文章信息，生成知乎文章分类标签（3-5个，每个2-4字）。知乎标签偏学术/行业分类，如「经济学」「消费心理学」「科技趋势」。标题：{title} 领域：{field} 核心观点：{core} 直接输出标签，空格分隔。",
|
||||
"wechat": f"根据文章信息，生成微信公众号文章标签（3-5个，每个2-4字）。公众号标签偏话题/兴趣分类，如「省钱攻略」「生活方式」「成长干货」。标题：{title} 领域：{field} 核心观点：{core} 直接输出标签，空格分隔。",
|
||||
"xiaohongshu": f"根据文章信息，生成小红书笔记标签（3-5个，每个2-4字）。小红书标签偏场景/人群分类，如「实用干货」「学生党」「打工人必看」「好物分享」。标题：{title} 领域：{field} 核心观点：{core} 直接输出标签，空格分隔。",
|
||||
}
|
||||
|
||||
if HAVE_LLM:
|
||||
prompt = tag_prompts.get(platform, f"根据文章信息生成适合{platform}的标签。标题：{title} 领域：{field} 核心观点：{core} 直接输出标签，空格分隔。")
|
||||
try:
|
||||
tags_text = call_llm(prompt, temperature=0.2, max_tokens=100)
|
||||
if tags_text:
|
||||
tags = [t.strip('#') for t in tags_text.strip().split() if t.strip('#')]
|
||||
if tags:
|
||||
return " ".join(f'<span class="tag">{t}</span>' for t in tags[:5])
|
||||
except Exception:
|
||||
pass
|
||||
|
||||
tags = []
|
||||
if field:
|
||||
import re
|
||||
parts = re.split(r'[/、与和及]', field)
|
||||
for p in parts:
|
||||
p = p.strip()
|
||||
if len(p) >= 2:
|
||||
tags.append(p)
|
||||
if len(parts) == 1 and len(parts[0]) > 4:
|
||||
for i in range(0, len(parts[0]), 2):
|
||||
chunk = parts[0][i:i+2]
|
||||
if len(chunk) == 2:
|
||||
tags.append(chunk)
|
||||
tags.pop(0)
|
||||
|
||||
platform_extra = {"zhihu": ["职场"], "xiaohongshu": ["生活"]}
|
||||
for t in platform_extra.get(platform, []):
|
||||
if t not in tags:
|
||||
tags.append(t)
|
||||
|
||||
if not tags:
|
||||
tags = ["科技"]
|
||||
|
||||
seen = set()
|
||||
return " ".join(f'<span class="tag">{t}</span>' for t in tags if t not in seen and not seen.add(t))
|
||||
|
||||
def _optimize_title(self, platform: str) -> str:
|
||||
original = self.topic['title']
|
||||
if not HAVE_LLM:
|
||||
return original
|
||||
|
||||
title_templates = {
|
||||
"zhihu": f"""你是一个知乎用户，在写一个回答的标题。
|
||||
|
||||
原文标题：{original}
|
||||
领域：{self.topic.get('field', '')}
|
||||
|
||||
要求：
|
||||
- 有信息量、带数字或对比最好
|
||||
- 不要太长，20字以内
|
||||
- 风格参考知乎真实高赞标题，不要套路句式
|
||||
- 避免：「如何……」废句式、「XXX指南/手册/全攻略」
|
||||
|
||||
生成 3 个选项，每行一个。""",
|
||||
|
||||
"wechat": f"""你是一个公众号作者，在给文章起标题。
|
||||
|
||||
原文标题：{original}
|
||||
领域：{self.topic.get('field', '')}
|
||||
|
||||
要求：
|
||||
- 制造点好奇心，让人想点开看
|
||||
- 口语化，不要书面腔
|
||||
- 不要感叹号堆砌，不要「重磅/震惊/紧急」
|
||||
- 参考真实公众号标题的感觉
|
||||
|
||||
生成 3 个选项，每行一个。""",
|
||||
|
||||
"xiaohongshu": f"""你是一个小红书用户，在给笔记起标题。
|
||||
|
||||
原文标题：{original}
|
||||
领域：{self.topic.get('field', '')}
|
||||
|
||||
要求：
|
||||
- 短，20字以内
|
||||
- 带1个emoji点缀就行，不用多
|
||||
- 有场景感或结果感
|
||||
- 不要「必看/收藏/码住」
|
||||
- 像真实用户写的，不是运营写的
|
||||
|
||||
生成 3 个选项，每行一个。""",
|
||||
}
|
||||
|
||||
prompt = title_templates.get(platform, f"给以下文章改个吸引人的{platform}标题：{original}")
|
||||
try:
|
||||
resp = call_llm(prompt, temperature=0.7, max_tokens=200)
|
||||
titles = []
|
||||
for line in resp.strip().split('\n'):
|
||||
line = line.strip()
|
||||
if not line:
|
||||
continue
|
||||
line = re.sub(r'^\d+[.、)\s]+', '', line)
|
||||
line = line.strip('*#- \t')
|
||||
if line:
|
||||
titles.append(line)
|
||||
if titles:
|
||||
logger.info(f"标题优化 [{platform}]: {titles[0][:50]}...")
|
||||
return titles[0]
|
||||
except Exception as e:
|
||||
logger.warning(f"标题优化失败: {e}")
|
||||
return original
|
||||
|
||||
def generate_platform_html(self, markdown: str, platform: str) -> str:
|
||||
title = self._optimize_title(platform)
|
||||
adapted = self._adapt_for_platform(markdown, platform)
|
||||
tpl_path = TEMPLATES_DIR / f"{platform}.html"
|
||||
if tpl_path.exists():
|
||||
template = tpl_path.read_text(encoding='utf-8')
|
||||
else:
|
||||
template = "<!DOCTYPE html><html><head><meta charset='UTF-8'><title>{{TITLE}}</title></head><body><h1>{{TITLE}}</h1><!-- CONTENT --></body></html>"
|
||||
template = "<!DOCTYPE html><html><head><meta charset='UTF-8'><title>{{TITLE}}</title><meta name='viewport' content='width=device-width'><style>body{max-width:800px;margin:0 auto;padding:20px;font-family:-apple-system,sans-serif;line-height:1.8}</style></head><body><h1>{{TITLE}}</h1><!-- CONTENT --></body></html>"
|
||||
|
||||
# 替换变量
|
||||
html = template.replace("{{TITLE}}", title).replace("{{DATE}}", TODAY).replace("{{GEN_TIME}}", GEN_TIME)
|
||||
|
||||
# 注入内容
|
||||
html_content = self._markdown_to_html(markdown)
|
||||
html_content = _md_parser(adapted)
|
||||
html = html.replace("<!-- CONTENT -->", html_content)
|
||||
|
||||
# 平台特定标签补充
|
||||
if platform == "zhihu":
|
||||
tags = '<div class="tags">#科技 #职场</div>'
|
||||
html = html.replace("<!-- TAGS -->", tags)
|
||||
elif platform == "xiaohongshu":
|
||||
hashtags = '<div class="hashtags">#AI #可持续 #生活方式</div>'
|
||||
html = html.replace("<!-- HASHTAGS -->", hashtags)
|
||||
elif platform == "wechat":
|
||||
pass
|
||||
tags_html = self._get_platform_tags(platform)
|
||||
if tags_html:
|
||||
html = html.replace("<!-- TAGS -->", tags_html)
|
||||
else:
|
||||
html = html.replace("<!-- TAGS -->", "")
|
||||
|
||||
return html
|
||||
|
||||
def _markdown_to_html(self, md: str) -> str:
|
||||
"""极简 markdown 转换（仅本场景使用）"""
|
||||
lines = md.split('\n')
|
||||
html_parts = []
|
||||
for line in lines:
|
||||
if line.startswith('# '):
|
||||
html_parts.append(f"<h1>{line[2:]}</h1>")
|
||||
elif line.startswith('## '):
|
||||
html_parts.append(f"<h2>{line[3:]}</h2>")
|
||||
elif line.startswith('### '):
|
||||
html_parts.append(f"<h3>{line[4:]}</h3>")
|
||||
elif line.strip().startswith('- '):
|
||||
html_parts.append(f"<li>{line[2:]}</li>")
|
||||
elif re.match(r'^\d+\. ', line):
|
||||
content = re.sub(r'^\d+\. ', '', line)
|
||||
html_parts.append(f"<li>{content}</li>")
|
||||
elif line.strip():
|
||||
html_parts.append(f"<p>{line}</p>")
|
||||
else:
|
||||
html_parts.append("")
|
||||
return "\n".join(html_parts)
|
||||
|
||||
def save_html(self, html: str, platform: str) -> Path:
|
||||
out_dir = self.release_dir / platform
|
||||
out_dir.mkdir(parents=True, exist_ok=True)
|
||||
filename = f"{platform}_{self.topic_id}_{platform}.html"
|
||||
filename = f"{platform}_{self.topic_id}.html"
|
||||
out_path = out_dir / filename
|
||||
out_path.write_text(html, encoding='utf-8')
|
||||
logger.info(f"HTML 生成: {out_path}")
|
||||
return out_path
|
||||
|
||||
def mark_draft(self):
|
||||
"""标记选题为「待审查」"""
|
||||
# 更新数据库状态
|
||||
update_topic_status(self.topic_id, 'review')
|
||||
logger.info(f"选题 {self.topic_id} 状态已更新为待审查（数据库）")
|
||||
|
||||
|
||||
@@ -1,276 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
撰写阶段：基于大纲和选题生成完整文章（三平台版本）
|
||||
"""
|
||||
|
||||
import json, datetime, logging, sys, re, subprocess
|
||||
from pathlib import Path
|
||||
from typing import Dict, List
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
# 添加项目根和 backend 路径，以导入 app.core.llm_client
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
sys.path.insert(0, str(PROJECT_ROOT / 'platform' / 'backend'))
|
||||
|
||||
# 导入 LLM 客户端（NVIDIA）
|
||||
try:
|
||||
from app.core.nvidia_client import expand_content_with_llm # type: ignore
|
||||
HAVE_LLM = True
|
||||
except ImportError as e:
|
||||
logging.warning(f"LLM client unavailable: {e}")
|
||||
HAVE_LLM = False
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TOPICS_FILE = DATA_DIR / "sustainability_topics.json"
|
||||
OUTLINE_DIR = DATA_DIR / "outlines"
|
||||
RELEASE_DIR = DATA_DIR / "releases"
|
||||
TEMPLATES_DIR = PROJECT_ROOT / "automation" / "templates"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
logging.basicConfig(
|
||||
level=logging.INFO,
|
||||
format='%(asctime)s - %(levelname)s - %(message)s',
|
||||
handlers=[
|
||||
logging.FileHandler(LOGS_DIR / f"writer_{TODAY}.log"),
|
||||
logging.StreamHandler()
|
||||
]
|
||||
)
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
class Writer:
|
||||
def __init__(self, topic_id: str):
|
||||
self.topic_id = topic_id
|
||||
self.topic = self._load_topic()
|
||||
outline_file = OUTLINE_DIR / TODAY / f"{topic_id}_outline.md"
|
||||
if not outline_file.exists():
|
||||
raise FileNotFoundError(f"Outline not found: {outline_file}")
|
||||
self.outline_content = outline_file.read_text(encoding='utf-8')
|
||||
self.release_dir = RELEASE_DIR / TODAY
|
||||
self.release_dir.mkdir(parents=True, exist_ok=True)
|
||||
# 加载研究笔记（作为 LLM 上下文）
|
||||
research_file = DATA_DIR / "research" / TODAY / f"{topic_id}_research.md"
|
||||
self.research_notes = research_file.read_text(encoding='utf-8') if research_file.exists() else ""
|
||||
|
||||
def _load_topic(self) -> Dict:
|
||||
topics = json.loads(TOPICS_FILE.read_text(encoding='utf-8'))
|
||||
for t in topics:
|
||||
if t['id'] == self.topic_id:
|
||||
return t
|
||||
raise ValueError(f"Topic {self.topic_id} not found")
|
||||
|
||||
def _clean_title(self, title: str) -> str:
|
||||
"""去除标题中的指导性文字（如字数说明、MVP标记等）"""
|
||||
import re
|
||||
# 去掉括号中的说明：约200字、约300字、MVP、试行等
|
||||
title = re.sub(r'[（(]约\s*\d+字[）)]', '', title)
|
||||
title = re.sub(r'[（(]MVP[）)]', '', title)
|
||||
title = re.sub(r'[（(][^）)]*?[）)]', '', title) # 保守移除任意括号内容（可能误伤，但大纲通常不包含重要括号信息）
|
||||
return title.strip()
|
||||
|
||||
def _parse_outline_sections(self) -> List[Dict]:
|
||||
"""将大纲 Markdown 解析为结构化列表，保留层级和内容"""
|
||||
sections = []
|
||||
current = None
|
||||
for line in self.outline_content.splitlines():
|
||||
if line.startswith("# "):
|
||||
if current:
|
||||
sections.append(current)
|
||||
current = {"level": 1, "title": line[2:].strip(), "content": ""}
|
||||
elif line.startswith("## "):
|
||||
if current:
|
||||
sections.append(current)
|
||||
current = {"level": 2, "title": line[3:].strip(), "content": ""}
|
||||
elif line.startswith("### "):
|
||||
if current:
|
||||
sections.append(current)
|
||||
current = {"level": 3, "title": line[4:].strip(), "content": ""}
|
||||
else:
|
||||
if current and line.strip():
|
||||
current['content'] = current.get('content', '') + line + "\n"
|
||||
if current:
|
||||
sections.append(current)
|
||||
return sections
|
||||
|
||||
def _expand_section(self, section: Dict) -> str:
|
||||
"""将大纲中的简短描述扩展为完整段落"""
|
||||
content = section.get('content', '').strip()
|
||||
# 如果有足够内容（>200字），直接返回
|
||||
if len(content) > 200:
|
||||
return content
|
||||
# 如果内容极少，需要 LLM 扩写
|
||||
if HAVE_LLM and len(content) < 150:
|
||||
logger.info(f"使用 LLM 扩写章节: {section['title']}")
|
||||
try:
|
||||
expanded = expand_content_with_llm(
|
||||
topic=self.topic,
|
||||
section_title=section['title'],
|
||||
section_content=content,
|
||||
context=self.research_notes
|
||||
)
|
||||
if expanded and len(expanded.strip()) > len(content):
|
||||
return expanded.strip()
|
||||
else:
|
||||
logger.warning("LLM 扩写结果为空或过短，使用占位")
|
||||
raise ValueError("Empty expansion")
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM 扩写失败: {e}，使用占位内容")
|
||||
# 返回占位内容，保持流程继续
|
||||
return f"{content}\n\n（本段内容需要人工补充：当前模型调用失败或未配置）"
|
||||
# 否则返回原内容
|
||||
return content
|
||||
|
||||
def generate_full_markdown(self) -> str:
|
||||
"""根据大纲生成完整 Markdown 正文（不用原标题，全部由 LLM 扩写生成）"""
|
||||
sections = self._parse_outline_sections()
|
||||
parts = []
|
||||
|
||||
# 只保留 LLM 扩写的内容，不添加任何原始标题标记
|
||||
for sec in sections:
|
||||
# 如果内容极短，LLM 扩写后返回的完整段落中可能包含标题，我们不过滤
|
||||
if sec.get('content'):
|
||||
expanded = self._expand_section(sec)
|
||||
parts.append(expanded + "\n\n")
|
||||
|
||||
full_md = "\n".join(parts).strip()
|
||||
|
||||
# 添加文末声明
|
||||
full_md += f"\n<p>（本文由宇之然AI助手生成，数据来源可靠，内容经合规审查）</p>\n"
|
||||
full_md += f"<p><em>生成时间：{TODAY}</em></p>\n"
|
||||
return full_md
|
||||
|
||||
def generate_platform_html(self, markdown: str, platform: str) -> str:
|
||||
"""将 Markdown 转换为平台 HTML（基于模板）"""
|
||||
title = self.topic['title']
|
||||
|
||||
# 加载模板
|
||||
tpl_path = TEMPLATES_DIR / f"{platform}.html"
|
||||
if tpl_path.exists():
|
||||
template = tpl_path.read_text(encoding='utf-8')
|
||||
else:
|
||||
template = "<!DOCTYPE html><html><head><meta charset='UTF-8'><title>{{TITLE}}</title></head><body><h1>{{TITLE}}</h1><!-- CONTENT --></body></html>"
|
||||
|
||||
# 替换变量
|
||||
html = template.replace("{{TITLE}}", title).replace("{{DATE}}", TODAY)
|
||||
|
||||
# 注入内容 (简单处理：markdown 转 HTML 可以用 marked.js 或 simple转换，这里暂时用 <pre> 包裹或简单段落化)
|
||||
# 为了快速展示，我们将 markdown 的段落转换为 <p> 标签
|
||||
# 实际中建议使用 markdown 库（如 python-markdown）转换
|
||||
html_content = self._markdown_to_html(markdown)
|
||||
html = html.replace("<!-- CONTENT -->", html_content)
|
||||
|
||||
# 平台特定标签补充
|
||||
if platform == "zhihu":
|
||||
tags = '<div class="tags">#科技 #职场</div>'
|
||||
html = html.replace("<!-- TAGS -->", tags)
|
||||
elif platform == "xiaohongshu":
|
||||
hashtags = '<div class="hashtags">#AI #可持续 #生活方式</div>'
|
||||
html = html.replace("<!-- HASHTAGS -->", hashtags)
|
||||
elif platform == "wechat":
|
||||
# 微信公众号可能还需要摘要等，模板已处理
|
||||
pass
|
||||
|
||||
return html
|
||||
|
||||
def _markdown_to_html(self, md: str) -> str:
|
||||
"""极简 markdown 转换（仅本场景使用）"""
|
||||
lines = md.split('\n')
|
||||
html_parts = []
|
||||
for line in lines:
|
||||
if line.startswith('# '):
|
||||
html_parts.append(f"<h1>{line[2:]}</h1>")
|
||||
elif line.startswith('## '):
|
||||
html_parts.append(f"<h2>{line[3:]}</h2>")
|
||||
elif line.startswith('### '):
|
||||
html_parts.append(f"<h3>{line[4:]}</h3>")
|
||||
elif line.strip().startswith('- '):
|
||||
html_parts.append(f"<li>{line[2:]}</li>")
|
||||
elif re.match(r'^\d+\. ', line):
|
||||
content = re.sub(r'^\d+\. ', '', line)
|
||||
html_parts.append(f"<li>{content}</li>")
|
||||
elif line.strip():
|
||||
html_parts.append(f"<p>{line}</p>")
|
||||
else:
|
||||
html_parts.append("") # 空行
|
||||
return "\n".join(html_parts)
|
||||
|
||||
def save_html(self, html: str, platform: str) -> Path:
|
||||
out_dir = self.release_dir / platform
|
||||
out_dir.mkdir(parents=True, exist_ok=True)
|
||||
filename = f"{platform}_{self.topic_id}_{platform}.html"
|
||||
out_path = out_dir / filename
|
||||
out_path.write_text(html, encoding='utf-8')
|
||||
logger.info(f"HTML 生成: {out_path}")
|
||||
return out_path
|
||||
|
||||
def mark_draft(self):
|
||||
"""标记选题为「待发布」，同时更新数据库"""
|
||||
# 更新 JSON 文件
|
||||
with open(TOPICS_FILE, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
updated = False
|
||||
for t in topics:
|
||||
if t.get('id') == self.topic_id:
|
||||
t[status'] = 'draft'
|
||||
updated = True
|
||||
break
|
||||
if updated:
|
||||
with open(TOPICS_FILE, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
|
||||
# 更新数据库
|
||||
db = SessionLocal()
|
||||
try:
|
||||
topic_db = db.query(Topic).filter(Topic.id == self.topic_id).first()
|
||||
if topic_db:
|
||||
topic_db.status = 'draft'
|
||||
db.commit()
|
||||
logger.info(f"选题 {self.topic_id} 状态已更新为 draft（数据库）")
|
||||
else:
|
||||
logger.warning(f"数据库中未找到选题 {self.topic_id}")
|
||||
except Exception as e:
|
||||
logger.error(f"更新数据库失败: {e}")
|
||||
db.rollback()
|
||||
finally:
|
||||
db.close()
|
||||
|
||||
logger.info(f"选题 {self.topic_id} 状态更新为「待发布」（JSON）")
|
||||
"""标记选题为「待发布」"""
|
||||
with open(TOPICS_FILE, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
for t in topics:
|
||||
if t.get('id') == self.topic_id:
|
||||
t['status'] = 'draft'
|
||||
# ready_at 留空，待合规审核通过后设置
|
||||
break
|
||||
with open(TOPICS_FILE, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
logger.info(f"选题 {self.topic_id} 状态更新为「待发布」")
|
||||
|
||||
def run(self):
|
||||
logger.info("开始撰写阶段")
|
||||
markdown = self.generate_full_markdown()
|
||||
results = {}
|
||||
for platform in ["zhihu", "wechat", "xiaohongshu"]:
|
||||
html = self.generate_platform_html(markdown, platform)
|
||||
results[platform] = str(self.save_html(html, platform))
|
||||
self.mark_draft()
|
||||
logger.info(f"撰写完成，状态改为 draft，待合规审核")
|
||||
return {"ok": True, "files": results}
|
||||
|
||||
def main():
|
||||
import argparse
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--topic-id', required=True, help='选题ID')
|
||||
args = parser.parse_args()
|
||||
|
||||
w = Writer(args.topic_id)
|
||||
result = w.run()
|
||||
print(json.dumps(result, ensure_ascii=False))
|
||||
sys.exit(0 if result['ok'] else 1)
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,278 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
撰写阶段：基于大纲和选题生成完整文章（三平台版本）
|
||||
"""
|
||||
|
||||
import json, datetime, logging, sys, re, subprocess
|
||||
from pathlib import Path
|
||||
from typing import Dict, List
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
# 添加项目根和 backend 路径，以导入 app.core.llm_client
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
sys.path.insert(0, str(PROJECT_ROOT / 'platform' / 'backend'))
|
||||
|
||||
# 导入 LLM 客户端（NVIDIA）
|
||||
try:
|
||||
from app.core.nvidia_client import expand_content_with_llm # type: ignore
|
||||
HAVE_LLM = True
|
||||
except ImportError as e:
|
||||
logging.warning(f"LLM client unavailable: {e}")
|
||||
HAVE_LLM = False
|
||||
|
||||
PROJECT_ROOT = Path(__file__).parent.parent
|
||||
sys.path.insert(0, str(PROJECT_ROOT))
|
||||
|
||||
DATA_DIR = PROJECT_ROOT / "automation" / "data"
|
||||
TOPICS_FILE = DATA_DIR / "sustainability_topics.json"
|
||||
OUTLINE_DIR = DATA_DIR / "outlines"
|
||||
RELEASE_DIR = DATA_DIR / "releases"
|
||||
TEMPLATES_DIR = PROJECT_ROOT / "automation" / "templates"
|
||||
LOGS_DIR = PROJECT_ROOT / "automation" / "logs"
|
||||
TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
|
||||
|
||||
logging.basicConfig(
|
||||
level=logging.INFO,
|
||||
format='%(asctime)s - %(levelname)s - %(message)s',
|
||||
handlers=[
|
||||
logging.FileHandler(LOGS_DIR / f"writer_{TODAY}.log"),
|
||||
logging.StreamHandler()
|
||||
]
|
||||
)
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
class Writer:
|
||||
def __init__(self, topic_id: str):
|
||||
self.topic_id = topic_id
|
||||
self.topic = self._load_topic()
|
||||
outline_file = OUTLINE_DIR / TODAY / f"{topic_id}_outline.md"
|
||||
if not outline_file.exists():
|
||||
raise FileNotFoundError(f"Outline not found: {outline_file}")
|
||||
self.outline_content = outline_file.read_text(encoding='utf-8')
|
||||
self.release_dir = RELEASE_DIR / TODAY
|
||||
self.release_dir.mkdir(parents=True, exist_ok=True)
|
||||
# 加载研究笔记（作为 LLM 上下文）
|
||||
research_file = DATA_DIR / "research" / TODAY / f"{topic_id}_research.md"
|
||||
self.research_notes = research_file.read_text(encoding='utf-8') if research_file.exists() else ""
|
||||
|
||||
def _load_topic(self) -> Dict:
|
||||
topics = json.loads(TOPICS_FILE.read_text(encoding='utf-8'))
|
||||
for t in topics:
|
||||
if t['id'] == self.topic_id:
|
||||
return t
|
||||
raise ValueError(f"Topic {self.topic_id} not found")
|
||||
|
||||
def _clean_title(self, title: str) -> str:
|
||||
"""去除标题中的指导性文字（如字数说明、MVP标记等）"""
|
||||
import re
|
||||
# 去掉括号中的说明：约200字、约300字、MVP、试行等
|
||||
title = re.sub(r'[（(]约\s*\d+字[）)]', '', title)
|
||||
title = re.sub(r'[（(]MVP[）)]', '', title)
|
||||
title = re.sub(r'[（(][^）)]*?[）)]', '', title) # 保守移除任意括号内容（可能误伤，但大纲通常不包含重要括号信息）
|
||||
return title.strip()
|
||||
|
||||
def _parse_outline_sections(self) -> List[Dict]:
|
||||
"""将大纲 Markdown 解析为结构化列表，保留层级和内容"""
|
||||
sections = []
|
||||
current = None
|
||||
for line in self.outline_content.splitlines():
|
||||
if line.startswith("# "):
|
||||
if current:
|
||||
sections.append(current)
|
||||
current = {"level": 1, "title": line[2:].strip(), "content": ""}
|
||||
elif line.startswith("## "):
|
||||
if current:
|
||||
sections.append(current)
|
||||
current = {"level": 2, "title": line[3:].strip(), "content": ""}
|
||||
elif line.startswith("### "):
|
||||
if current:
|
||||
sections.append(current)
|
||||
current = {"level": 3, "title": line[4:].strip(), "content": ""}
|
||||
else:
|
||||
if current and line.strip():
|
||||
current['content'] = current.get('content', '') + line + "\n"
|
||||
if current:
|
||||
sections.append(current)
|
||||
return sections
|
||||
|
||||
def _expand_section(self, section: Dict) -> str:
|
||||
"""将大纲中的简短描述扩展为完整段落"""
|
||||
content = section.get('content', '').strip()
|
||||
# 如果有足够内容（>200字），直接返回
|
||||
if len(content) > 200:
|
||||
return content
|
||||
# 如果内容极少，需要 LLM 扩写
|
||||
if HAVE_LLM and len(content) < 150:
|
||||
logger.info(f"使用 LLM 扩写章节: {section['title']}")
|
||||
try:
|
||||
expanded = expand_content_with_llm(
|
||||
topic=self.topic,
|
||||
section_title=section['title'],
|
||||
section_content=content,
|
||||
context=self.research_notes
|
||||
)
|
||||
if expanded and len(expanded.strip()) > len(content):
|
||||
return expanded.strip()
|
||||
else:
|
||||
logger.warning("LLM 扩写结果为空或过短，使用占位")
|
||||
raise ValueError("Empty expansion")
|
||||
except Exception as e:
|
||||
logger.warning(f"LLM 扩写失败: {e}，使用占位内容")
|
||||
# 返回占位内容，保持流程继续
|
||||
return f"{content}\n\n（本段内容需要人工补充：当前模型调用失败或未配置）"
|
||||
# 否则返回原内容
|
||||
return content
|
||||
|
||||
def generate_full_markdown(self) -> str:
|
||||
"""根据大纲生成完整 Markdown 正文（不用原标题，全部由 LLM 扩写生成）"""
|
||||
sections = self._parse_outline_sections()
|
||||
parts = []
|
||||
|
||||
# 只保留 LLM 扩写的内容，不添加任何原始标题标记
|
||||
for sec in sections:
|
||||
# 如果内容极短，LLM 扩写后返回的完整段落中可能包含标题，我们不过滤
|
||||
if sec.get('content'):
|
||||
expanded = self._expand_section(sec)
|
||||
parts.append(expanded + "\n\n")
|
||||
|
||||
full_md = "\n".join(parts).strip()
|
||||
|
||||
# 添加文末声明
|
||||
full_md += f"\n<p>（本文由宇之然AI助手生成，数据来源可靠，内容经合规审查）</p>\n"
|
||||
full_md += f"<p><em>生成时间：{TODAY}</em></p>\n"
|
||||
return full_md
|
||||
|
||||
def generate_platform_html(self, markdown: str, platform: str) -> str:
|
||||
"""将 Markdown 转换为平台 HTML（基于模板）"""
|
||||
title = self.topic['title']
|
||||
|
||||
# 加载模板
|
||||
tpl_path = TEMPLATES_DIR / f"{platform}.html"
|
||||
if tpl_path.exists():
|
||||
template = tpl_path.read_text(encoding='utf-8')
|
||||
else:
|
||||
template = "<!DOCTYPE html><html><head><meta charset='UTF-8'><title>{{TITLE}}</title></head><body><h1>{{TITLE}}</h1><!-- CONTENT --></body></html>"
|
||||
|
||||
# 替换变量
|
||||
html = template.replace("{{TITLE}}", title).replace("{{DATE}}", TODAY)
|
||||
gen_time = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
|
||||
html = html.replace("{{GEN_TIME}}", gen_time)
|
||||
|
||||
# 注入内容 (简单处理：markdown 转 HTML 可以用 marked.js 或 simple转换，这里暂时用 <pre> 包裹或简单段落化)
|
||||
# 为了快速展示，我们将 markdown 的段落转换为 <p> 标签
|
||||
# 实际中建议使用 markdown 库（如 python-markdown）转换
|
||||
html_content = self._markdown_to_html(markdown)
|
||||
html = html.replace("<!-- CONTENT -->", html_content)
|
||||
|
||||
# 平台特定标签补充
|
||||
if platform == "zhihu":
|
||||
tags = '<div class="tags">#科技 #职场</div>'
|
||||
html = html.replace("<!-- TAGS -->", tags)
|
||||
elif platform == "xiaohongshu":
|
||||
hashtags = '<div class="hashtags">#AI #可持续 #生活方式</div>'
|
||||
html = html.replace("<!-- HASHTAGS -->", hashtags)
|
||||
elif platform == "wechat":
|
||||
# 微信公众号可能还需要摘要等，模板已处理
|
||||
pass
|
||||
|
||||
return html
|
||||
|
||||
def _markdown_to_html(self, md: str) -> str:
|
||||
"""极简 markdown 转换（仅本场景使用）"""
|
||||
lines = md.split('\n')
|
||||
html_parts = []
|
||||
for line in lines:
|
||||
if line.startswith('# '):
|
||||
html_parts.append(f"<h1>{line[2:]}</h1>")
|
||||
elif line.startswith('## '):
|
||||
html_parts.append(f"<h2>{line[3:]}</h2>")
|
||||
elif line.startswith('### '):
|
||||
html_parts.append(f"<h3>{line[4:]}</h3>")
|
||||
elif line.strip().startswith('- '):
|
||||
html_parts.append(f"<li>{line[2:]}</li>")
|
||||
elif re.match(r'^\d+\. ', line):
|
||||
content = re.sub(r'^\d+\. ', '', line)
|
||||
html_parts.append(f"<li>{content}</li>")
|
||||
elif line.strip():
|
||||
html_parts.append(f"<p>{line}</p>")
|
||||
else:
|
||||
html_parts.append("") # 空行
|
||||
return "\n".join(html_parts)
|
||||
|
||||
def save_html(self, html: str, platform: str) -> Path:
|
||||
out_dir = self.release_dir / platform
|
||||
out_dir.mkdir(parents=True, exist_ok=True)
|
||||
filename = f"{platform}_{self.topic_id}_{platform}.html"
|
||||
out_path = out_dir / filename
|
||||
out_path.write_text(html, encoding='utf-8')
|
||||
logger.info(f"HTML 生成: {out_path}")
|
||||
return out_path
|
||||
|
||||
def mark_draft(self):
|
||||
"""标记选题为「待发布」，同时更新数据库"""
|
||||
# 更新 JSON 文件
|
||||
with open(TOPICS_FILE, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
updated = False
|
||||
for t in topics:
|
||||
if t.get('id') == self.topic_id:
|
||||
t['status'] = 'draft'
|
||||
updated = True
|
||||
break
|
||||
if updated:
|
||||
with open(TOPICS_FILE, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
|
||||
# 更新数据库
|
||||
db = SessionLocal()
|
||||
try:
|
||||
topic_db = db.query(Topic).filter(Topic.id == self.topic_id).first()
|
||||
if topic_db:
|
||||
topic_db.status = 'draft'
|
||||
db.commit()
|
||||
logger.info(f"选题 {self.topic_id} 状态已更新为 draft（数据库）")
|
||||
else:
|
||||
logger.warning(f"数据库中未找到选题 {self.topic_id}")
|
||||
except Exception as e:
|
||||
logger.error(f"更新数据库失败: {e}")
|
||||
db.rollback()
|
||||
finally:
|
||||
db.close()
|
||||
|
||||
logger.info(f"选题 {self.topic_id} 状态更新为「待发布」（JSON）")
|
||||
"""标记选题为「待发布」"""
|
||||
with open(TOPICS_FILE, 'r', encoding='utf-8') as f:
|
||||
topics = json.load(f)
|
||||
for t in topics:
|
||||
if t.get('id') == self.topic_id:
|
||||
t['status'] = 'draft'
|
||||
# ready_at 留空，待合规审核通过后设置
|
||||
break
|
||||
with open(TOPICS_FILE, 'w', encoding='utf-8') as f:
|
||||
json.dump(topics, f, ensure_ascii=False, indent=2)
|
||||
logger.info(f"选题 {self.topic_id} 状态更新为「待发布」")
|
||||
|
||||
def run(self):
|
||||
logger.info("开始撰写阶段")
|
||||
markdown = self.generate_full_markdown()
|
||||
results = {}
|
||||
for platform in ["zhihu", "wechat", "xiaohongshu"]:
|
||||
html = self.generate_platform_html(markdown, platform)
|
||||
results[platform] = str(self.save_html(html, platform))
|
||||
self.mark_draft()
|
||||
logger.info(f"撰写完成，状态改为 draft，待合规审核")
|
||||
return {"ok": True, "files": results}
|
||||
|
||||
def main():
|
||||
import argparse
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--topic-id', required=True, help='选题ID')
|
||||
args = parser.parse_args()
|
||||
|
||||
w = Writer(args.topic_id)
|
||||
result = w.run()
|
||||
print(json.dumps(result, ensure_ascii=False))
|
||||
sys.exit(0 if result['ok'] else 1)
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
Reference in New Issue
Block a user