zhoujie
4d83c0f4a9
- 新增热点自动采集后台线程，支持定时搜索关键词并执行 AI 分析，结果缓存至结构化状态 - 新增热点分析状态管理接口，提供线程安全的 `get_last_analysis` 和 `set_last_analysis` 方法 - 新增热点数据桥接函数 `feed_hotspot_to_engine`，将分析结果注入 TopicEngine 实现热点加权推荐 - 新增热点选题下拉组件，分析完成后自动填充推荐选题，选中后自动写入选题输入框 - 优化 `generate_from_hotspot` 函数，自动获取结构化分析摘要并增强生成上下文 - 新增热点自动采集配置节点，支持通过 `config.json` 管理关键词和采集间隔 ♻️ refactor(queue): 实现智能排期引擎并统一发布路径 - 新增智能排期引擎，基于 `AnalyticsService` 的 `time_weights` 自动计算最优发布时段 - 新增 `PublishQueue.suggest_schedule_time` 和 `auto_schedule_item` 方法，支持时段冲突检测和内容分布控制 - 修改 `generate_to_queue` 函数，新增 `auto_schedule` 和 `auto_approve` 参数，支持自动排期和自动审核 - 重构 `_scheduler_loop` 的自动发布分支，改为调用 `generate_to_queue` 通过队列发布，统一发布路径 - 重构 `auto_publish_once` 函数，移除直接发布逻辑，改为生成内容入队并返回队列信息 - 新增队列时段使用情况查询方法 `get_slot_usage`，支持 UI 热力图展示 📝 docs(openspec): 新增内容排期优化和热点探测优化规范文档 - 新增 `smart-schedule-engine` 规范，定义智能排期引擎的功能需求和场景 - 新增 `unified-publish-path` 规范，定义统一发布路径的改造方案 - 新增 `hotspot-analysis-state` 规范，定义热点分析状态存储的线程安全接口 - 新增 `hotspot-auto-collector` 规范，定义定时热点自动采集的任务流程 - 新增 `hotspot-engine-bridge` 规范，定义热点数据注入 TopicEngine 的桥接机制 - 新增 `hotspot-topic-selector` 规范，定义热点选题下拉组件的交互行为 - 更新 `services-queue`、`services-scheduler` 和 `services-hotspot` 规范，反映功能修改和新增参数 🔧 chore(config): 新增热点自动采集默认配置 - 在 `DEFAULT_CONFIG` 中新增 `hotspot_auto_collect` 配置节点，包含 `enabled`、`keywords` 和 `interval_hours` 字段 - 提供默认关键词列表 `["穿搭", "美妆", "好物"]` 和默认采集间隔 4 小时 🐛 fix(llm): 增强 JSON 解析容错能力 - 新增 `_try_fix_truncated_json` 方法，尝试修复被 token 限制截断的 JSON 输出 - 支持多种截断场景的自动补全，包括字符串值、数组和嵌套对象的截断修复 - 提高 LLM 分析热点等返回 JSON 的函数的稳定性 💄 style(ui): 优化队列管理和热点探测界面 - 在队列生成区域新增自动排期复选框，勾选后隐藏手动排期输入框 - 在日历视图旁新增推荐时段 Markdown 面板，展示各时段权重和建议热力图 - 在热点探测 Tab 新增推荐选题下拉组件，分析完成后动态填充选项 - 在热点探测 Tab 新增热点自动采集控制区域，支持启动、停止和配置采集参数
379 lines
14 KiB
Python
379 lines
14 KiB
Python
"""
|
||
services/queue_ops.py
|
||
发布队列操作：生成入队、状态管理、发布控制
|
||
"""
|
||
import os
|
||
import re
|
||
import time
|
||
import random
|
||
import logging
|
||
|
||
from PIL import Image
|
||
|
||
from .config_manager import ConfigManager, OUTPUT_DIR
|
||
from .publish_queue import (
|
||
PublishQueue, QueuePublisher,
|
||
STATUS_DRAFT, STATUS_APPROVED, STATUS_SCHEDULED, STATUS_PUBLISHING,
|
||
STATUS_PUBLISHED, STATUS_FAILED, STATUS_REJECTED, STATUS_LABELS,
|
||
)
|
||
from .llm_service import LLMService
|
||
from .sd_service import SDService
|
||
from .mcp_client import get_mcp_client
|
||
from .connection import _get_llm_config
|
||
from .persona import DEFAULT_TOPICS, DEFAULT_STYLES, _resolve_persona
|
||
from .content import generate_copy, generate_images
|
||
from .rate_limiter import _increment_stat, _clear_error_streak
|
||
|
||
cfg = ConfigManager()
|
||
logger = logging.getLogger("autobot")
|
||
|
||
# 模块级依赖（通过 configure() 注入）
|
||
_pub_queue: "PublishQueue | None" = None
|
||
_queue_publisher: "QueuePublisher | None" = None
|
||
_analytics = None
|
||
_log_fn = None
|
||
|
||
|
||
def configure(pub_queue, queue_publisher, analytics_svc, log_fn=None):
|
||
"""从 main.py 初始化段注入队列和分析服务"""
|
||
global _pub_queue, _queue_publisher, _analytics, _log_fn
|
||
_pub_queue = pub_queue
|
||
_queue_publisher = queue_publisher
|
||
_analytics = analytics_svc
|
||
_log_fn = log_fn
|
||
# 注册发布回调（在依赖注入完成后）
|
||
_queue_publisher.set_publish_callback(_queue_publish_callback)
|
||
_queue_publisher.set_log_callback(_log_fn or _log)
|
||
|
||
|
||
def _log(msg: str):
|
||
if _log_fn:
|
||
_log_fn(msg)
|
||
else:
|
||
logger.info("[queue] %s", msg)
|
||
|
||
# ==================================================
|
||
# 发布队列相关函数
|
||
# ==================================================
|
||
|
||
def generate_to_queue(topics_str, sd_url_val, sd_model_name, model, persona_text=None,
|
||
quality_mode_val=None, face_swap_on=False, count=1,
|
||
scheduled_time=None, auto_schedule=False, auto_approve=False):
|
||
"""批量生成内容 → 加入发布队列（不直接发布）
|
||
|
||
Args:
|
||
auto_schedule: 为每篇内容自动分配最优排期时间
|
||
auto_approve: 入队后自动审核通过
|
||
"""
|
||
try:
|
||
topics = [t.strip() for t in topics_str.split(",") if t.strip()] if topics_str else DEFAULT_TOPICS
|
||
use_weights = cfg.get("use_smart_weights", True) and _analytics.has_weights
|
||
|
||
api_key, base_url, _ = _get_llm_config()
|
||
if not api_key:
|
||
return "❌ LLM 未配置"
|
||
if not sd_url_val or not sd_model_name:
|
||
return "❌ SD WebUI 未连接或未选择模型"
|
||
|
||
count = max(1, min(int(count), 10))
|
||
results = []
|
||
|
||
for i in range(count):
|
||
try:
|
||
_log(f"📋 [队列生成] 正在生成第 {i+1}/{count} 篇...")
|
||
|
||
if use_weights:
|
||
topic = _analytics.get_weighted_topic(topics)
|
||
style = _analytics.get_weighted_style(DEFAULT_STYLES)
|
||
else:
|
||
topic = random.choice(topics)
|
||
style = random.choice(DEFAULT_STYLES)
|
||
|
||
svc = LLMService(api_key, base_url, model)
|
||
persona = _resolve_persona(persona_text) if persona_text else None
|
||
|
||
if use_weights:
|
||
weight_insights = f"高权重主题: {', '.join(list(_analytics._weights.get('topic_weights', {}).keys())[:5])}\n"
|
||
weight_insights += f"权重摘要: {_analytics.weights_summary}"
|
||
title_advice = _analytics.get_title_advice()
|
||
hot_tags = ", ".join(_analytics.get_top_tags(8))
|
||
try:
|
||
data = svc.generate_weighted_copy(topic, style, weight_insights, title_advice, hot_tags, sd_model_name=sd_model_name, persona=persona)
|
||
except Exception:
|
||
data = svc.generate_copy(topic, style, sd_model_name=sd_model_name, persona=persona)
|
||
else:
|
||
data = svc.generate_copy(topic, style, sd_model_name=sd_model_name, persona=persona)
|
||
|
||
title = (data.get("title", "") or "")[:20]
|
||
content = data.get("content", "")
|
||
sd_prompt = data.get("sd_prompt", "")
|
||
tags = data.get("tags", [])
|
||
|
||
if use_weights:
|
||
top_tags = _analytics.get_top_tags(5)
|
||
for t in top_tags:
|
||
if t not in tags:
|
||
tags.append(t)
|
||
tags = tags[:10]
|
||
|
||
if not title:
|
||
_log(f"⚠️ 第 {i+1} 篇文案生成失败，跳过")
|
||
continue
|
||
|
||
# 生成图片
|
||
sd_svc = SDService(sd_url_val)
|
||
face_image = None
|
||
if face_swap_on:
|
||
face_image = SDService.load_face_image()
|
||
images = sd_svc.txt2img(prompt=sd_prompt, model=sd_model_name,
|
||
face_image=face_image,
|
||
quality_mode=quality_mode_val or "快速 (约30秒)",
|
||
persona=persona)
|
||
if not images:
|
||
_log(f"⚠️ 第 {i+1} 篇图片生成失败，跳过")
|
||
continue
|
||
|
||
# 保存备份
|
||
ts = int(time.time())
|
||
safe_title = re.sub(r'[\\/*?:"<>|]', "", title)[:20]
|
||
backup_dir = os.path.join(OUTPUT_DIR, f"{ts}_{safe_title}")
|
||
os.makedirs(backup_dir, exist_ok=True)
|
||
|
||
with open(os.path.join(backup_dir, "文案.txt"), "w", encoding="utf-8") as f:
|
||
f.write(f"标题: {title}\n风格: {style}\n主题: {topic}\n\n{content}\n\n标签: {', '.join(tags)}\n\nSD Prompt: {sd_prompt}")
|
||
|
||
image_paths = []
|
||
for idx, img in enumerate(images):
|
||
if isinstance(img, Image.Image):
|
||
path = os.path.abspath(os.path.join(backup_dir, f"图{idx+1}.jpg"))
|
||
if img.mode != "RGB":
|
||
img = img.convert("RGB")
|
||
img.save(path, format="JPEG", quality=95)
|
||
image_paths.append(path)
|
||
|
||
if not image_paths:
|
||
continue
|
||
|
||
# 加入队列
|
||
item_id = _pub_queue.add(
|
||
title=title, content=content, sd_prompt=sd_prompt,
|
||
tags=tags, image_paths=image_paths, backup_dir=backup_dir,
|
||
topic=topic, style=style, persona=persona or "",
|
||
status=STATUS_DRAFT, scheduled_time=scheduled_time,
|
||
)
|
||
# 自动排期
|
||
sched_msg = ""
|
||
if auto_schedule and _analytics:
|
||
ok = _pub_queue.auto_schedule_item(item_id, _analytics)
|
||
if ok:
|
||
item = _pub_queue.get(item_id)
|
||
sched_msg = f" ⏰{item['scheduled_time'][:16]}" if item else ""
|
||
_log(f"🕐 #{item_id} 自动排期{sched_msg}")
|
||
|
||
# 自动审核
|
||
if auto_approve:
|
||
_pub_queue.approve(item_id)
|
||
_log(f"✅ #{item_id} 自动审核通过")
|
||
|
||
results.append(f"#{item_id} {title}{sched_msg}")
|
||
_log(f"📋 已加入队列 #{item_id}: {title}")
|
||
|
||
# 多篇间隔
|
||
if i < count - 1:
|
||
time.sleep(2)
|
||
|
||
except Exception as e:
|
||
_log(f"⚠️ 第 {i+1} 篇生成异常: {e}")
|
||
continue
|
||
|
||
if not results:
|
||
return "❌ 所有内容生成失败，请检查配置"
|
||
|
||
return f"✅ 已生成 {len(results)} 篇内容加入队列:\n" + "\n".join(f" - {r}" for r in results)
|
||
|
||
except Exception as e:
|
||
return f"❌ 批量生成异常: {e}"
|
||
|
||
|
||
def _queue_publish_callback(item: dict) -> tuple[bool, str]:
|
||
"""队列发布回调: 从队列项数据发布到小红书"""
|
||
try:
|
||
mcp_url = cfg.get("mcp_url", "http://localhost:18060/mcp")
|
||
client = get_mcp_client(mcp_url)
|
||
title = item.get("title", "")
|
||
content = item.get("content", "")
|
||
image_paths = item.get("image_paths", [])
|
||
tags = item.get("tags", [])
|
||
|
||
if not title or not image_paths:
|
||
return False, "标题或图片缺失"
|
||
|
||
# 验证图片文件存在
|
||
valid_paths = [p for p in image_paths if os.path.isfile(p)]
|
||
if not valid_paths:
|
||
return False, "所有图片文件不存在"
|
||
|
||
result = client.publish_content(
|
||
title=title, content=content, images=valid_paths, tags=tags,
|
||
)
|
||
if "error" in result:
|
||
return False, result["error"]
|
||
|
||
_increment_stat("publishes")
|
||
_clear_error_streak()
|
||
return True, result.get("text", "发布成功")
|
||
|
||
except Exception as e:
|
||
return False, str(e)
|
||
|
||
|
||
|
||
|
||
def queue_format_table():
|
||
"""返回当前队列的完整表格（不过滤）"""
|
||
return _pub_queue.format_queue_table() if _pub_queue else ""
|
||
|
||
|
||
def queue_format_calendar():
|
||
"""返回未来14天的日历视图"""
|
||
return _pub_queue.format_calendar(14) if _pub_queue else ""
|
||
|
||
|
||
def queue_refresh_table(status_filter):
|
||
"""刷新队列表格"""
|
||
statuses = None
|
||
if status_filter and status_filter != "全部":
|
||
status_map = {v: k for k, v in STATUS_LABELS.items()}
|
||
if status_filter in status_map:
|
||
statuses = [status_map[status_filter]]
|
||
return _pub_queue.format_queue_table(statuses)
|
||
|
||
|
||
def queue_refresh_calendar():
|
||
"""刷新日历视图"""
|
||
return _pub_queue.format_calendar(14)
|
||
|
||
|
||
def queue_preview_item(item_id_str):
|
||
"""预览队列项"""
|
||
try:
|
||
item_id = int(str(item_id_str).strip().replace("#", ""))
|
||
return _pub_queue.format_preview(item_id)
|
||
except (ValueError, TypeError):
|
||
return "❌ 请输入有效的队列项 ID（数字）"
|
||
|
||
|
||
def queue_approve_item(item_id_str, scheduled_time_str):
|
||
"""审核通过"""
|
||
try:
|
||
item_id = int(str(item_id_str).strip().replace("#", ""))
|
||
sched = scheduled_time_str.strip() if scheduled_time_str else None
|
||
ok = _pub_queue.approve(item_id, scheduled_time=sched)
|
||
if ok:
|
||
status = "已排期" if sched else "待发布"
|
||
return f"✅ #{item_id} 已审核通过 → {status}"
|
||
return f"❌ #{item_id} 无法审核（可能不是草稿/失败状态）"
|
||
except (ValueError, TypeError):
|
||
return "❌ 请输入有效的 ID"
|
||
|
||
|
||
def queue_reject_item(item_id_str):
|
||
"""拒绝队列项"""
|
||
try:
|
||
item_id = int(str(item_id_str).strip().replace("#", ""))
|
||
ok = _pub_queue.reject(item_id)
|
||
return f"✅ #{item_id} 已拒绝" if ok else f"❌ #{item_id} 无法拒绝"
|
||
except (ValueError, TypeError):
|
||
return "❌ 请输入有效的 ID"
|
||
|
||
|
||
def queue_delete_item(item_id_str):
|
||
"""删除队列项"""
|
||
try:
|
||
item_id = int(str(item_id_str).strip().replace("#", ""))
|
||
ok = _pub_queue.delete(item_id)
|
||
return f"✅ #{item_id} 已删除" if ok else f"❌ #{item_id} 无法删除（可能正在发布中）"
|
||
except (ValueError, TypeError):
|
||
return "❌ 请输入有效的 ID"
|
||
|
||
|
||
def queue_retry_item(item_id_str):
|
||
"""重试失败项"""
|
||
try:
|
||
item_id = int(str(item_id_str).strip().replace("#", ""))
|
||
ok = _pub_queue.retry(item_id)
|
||
return f"✅ #{item_id} 已重新加入待发布" if ok else f"❌ #{item_id} 无法重试（不是失败状态）"
|
||
except (ValueError, TypeError):
|
||
return "❌ 请输入有效的 ID"
|
||
|
||
|
||
def queue_publish_now(item_id_str):
|
||
"""立即发布队列项"""
|
||
try:
|
||
item_id = int(str(item_id_str).strip().replace("#", ""))
|
||
return _queue_publisher.publish_now(item_id)
|
||
except (ValueError, TypeError):
|
||
return "❌ 请输入有效的 ID"
|
||
|
||
|
||
def queue_start_processor():
|
||
"""启动队列后台处理器"""
|
||
if _queue_publisher.is_running:
|
||
return "⚠️ 队列处理器已在运行中"
|
||
_queue_publisher.start(check_interval=60)
|
||
return "✅ 队列处理器已启动，每分钟检查待发布项"
|
||
|
||
|
||
def queue_stop_processor():
|
||
"""停止队列后台处理器"""
|
||
if not _queue_publisher.is_running:
|
||
return "⚠️ 队列处理器未在运行"
|
||
_queue_publisher.stop()
|
||
return "🛑 队列处理器已停止"
|
||
|
||
|
||
def queue_get_status():
|
||
"""获取队列状态摘要"""
|
||
counts = _pub_queue.count_by_status()
|
||
running = "🟢 运行中" if _queue_publisher.is_running else "⚪ 未启动"
|
||
parts = [f"**队列处理器**: {running}"]
|
||
for s, label in STATUS_LABELS.items():
|
||
cnt = counts.get(s, 0)
|
||
if cnt > 0:
|
||
parts.append(f"{label}: {cnt}")
|
||
total = sum(counts.values())
|
||
parts.append(f"**合计**: {total} 项")
|
||
return " · ".join(parts)
|
||
|
||
|
||
def queue_batch_approve(status_filter):
|
||
"""批量审核通过所有草稿"""
|
||
items = _pub_queue.list_by_status([STATUS_DRAFT])
|
||
if not items:
|
||
return "📭 没有待审核的草稿"
|
||
approved = 0
|
||
for item in items:
|
||
if _pub_queue.approve(item["id"]):
|
||
approved += 1
|
||
return f"✅ 已批量审核通过 {approved} 项"
|
||
|
||
|
||
def queue_generate_and_refresh(topics_str, sd_url_val, sd_model_name, model,
|
||
persona_text, quality_mode_val, face_swap_on,
|
||
gen_count, gen_schedule_time, auto_schedule=False):
|
||
"""生成内容到队列 + 刷新表格"""
|
||
msg = generate_to_queue(
|
||
topics_str, sd_url_val, sd_model_name, model,
|
||
persona_text=persona_text, quality_mode_val=quality_mode_val,
|
||
face_swap_on=face_swap_on, count=gen_count,
|
||
scheduled_time=gen_schedule_time.strip() if gen_schedule_time else None,
|
||
auto_schedule=auto_schedule,
|
||
)
|
||
table = _pub_queue.format_queue_table()
|
||
calendar = _pub_queue.format_calendar(14)
|
||
status = queue_get_status()
|
||
return msg, table, calendar, status
|
||
|
||
|
||
# 调度器下次执行时间追踪
|