feat: 提示词和 AI 参数移至数据库管理
- AiSetting 模型新增 6 个提示词字段 + 3 个阶段温度字段 - main.py 迁移逻辑改为检测 ai_settings 表的新字段并填充默认值 - generate.py 从数据库读取生成提示词，去掉硬编码 - optimizer 从 API 读取各阶段提示词和温度，删除 prompts.py - crawler 从 API 读取提取/改写提示词和温度，删除 prompts.py - settings/active 端点去掉 token 认证（供爬虫/优化器使用） - 后台设置页新增提示词编辑区和温度调节控件 - 新增 _ensure_default_settings 自动创建默认配置
This commit is contained in:
bwstudio
@@ -2,10 +2,10 @@
|
||||
<div class="setting-page">
|
||||
<h2>AI 配置</h2>
|
||||
|
||||
<!-- AI 配置表单 -->
|
||||
<!-- AI 基础配置 -->
|
||||
<el-card style="margin-top: 20px">
|
||||
<template #header>
|
||||
<span>AI 配置</span>
|
||||
<span>API 基础配置</span>
|
||||
</template>
|
||||
<el-form :model="form" label-width="120px">
|
||||
<el-form-item label="提供商">
|
||||
@@ -57,6 +57,81 @@
|
||||
</el-form>
|
||||
</el-card>
|
||||
|
||||
<!-- 提示词配置 -->
|
||||
<el-card style="margin-top: 20px">
|
||||
<template #header>
|
||||
<span>提示词模板</span>
|
||||
</template>
|
||||
<el-form :model="form" label-width="140px">
|
||||
<el-form-item label="AI 生成笑话">
|
||||
<el-input
|
||||
v-model="form.generate_prompt"
|
||||
type="textarea"
|
||||
:rows="6"
|
||||
placeholder="使用 {context}/{style}/{length_requirement} 作为占位符"
|
||||
style="width: 100%"
|
||||
/>
|
||||
</el-form-item>
|
||||
<el-divider />
|
||||
|
||||
<el-form-item label="优化器-质量检测">
|
||||
<el-input
|
||||
v-model="form.optimizer_quality_prompt"
|
||||
type="textarea"
|
||||
:rows="4"
|
||||
style="width: 100%"
|
||||
/>
|
||||
</el-form-item>
|
||||
<el-form-item label="质量检测温度">
|
||||
<el-input-number v-model="form.optimizer_quality_temperature" :min="0" :max="2" :step="0.1" />
|
||||
</el-form-item>
|
||||
<el-divider />
|
||||
|
||||
<el-form-item label="优化器-润色">
|
||||
<el-input
|
||||
v-model="form.optimizer_polish_prompt"
|
||||
type="textarea"
|
||||
:rows="4"
|
||||
style="width: 100%"
|
||||
/>
|
||||
</el-form-item>
|
||||
<el-form-item label="润色温度">
|
||||
<el-input-number v-model="form.optimizer_polish_temperature" :min="0" :max="2" :step="0.1" />
|
||||
</el-form-item>
|
||||
<el-divider />
|
||||
|
||||
<el-form-item label="优化器-评价分类">
|
||||
<el-input
|
||||
v-model="form.optimizer_evaluate_prompt"
|
||||
type="textarea"
|
||||
:rows="4"
|
||||
style="width: 100%"
|
||||
/>
|
||||
</el-form-item>
|
||||
<el-form-item label="评价温度">
|
||||
<el-input-number v-model="form.optimizer_evaluate_temperature" :min="0" :max="2" :step="0.1" />
|
||||
</el-form-item>
|
||||
<el-divider />
|
||||
|
||||
<el-form-item label="爬虫-提取笑话">
|
||||
<el-input
|
||||
v-model="form.crawler_extract_prompt"
|
||||
type="textarea"
|
||||
:rows="4"
|
||||
style="width: 100%"
|
||||
/>
|
||||
</el-form-item>
|
||||
<el-form-item label="爬虫-改写">
|
||||
<el-input
|
||||
v-model="form.crawler_rewrite_prompt"
|
||||
type="textarea"
|
||||
:rows="4"
|
||||
style="width: 100%"
|
||||
/>
|
||||
</el-form-item>
|
||||
</el-form>
|
||||
</el-card>
|
||||
|
||||
<div style="margin-top: 20px">
|
||||
<el-button type="primary" @click="handleSave" :loading="saving">保存配置</el-button>
|
||||
</div>
|
||||
@@ -106,6 +181,15 @@ const form = ref({
|
||||
crawl_enabled: false,
|
||||
crawl_keywords: '冷笑话,段子,谐音梗',
|
||||
max_pages_per_run: 3,
|
||||
generate_prompt: '',
|
||||
optimizer_quality_prompt: '',
|
||||
optimizer_polish_prompt: '',
|
||||
optimizer_evaluate_prompt: '',
|
||||
optimizer_quality_temperature: 0.3,
|
||||
optimizer_polish_temperature: 0.8,
|
||||
optimizer_evaluate_temperature: 0.3,
|
||||
crawler_extract_prompt: '',
|
||||
crawler_rewrite_prompt: '',
|
||||
})
|
||||
|
||||
const allSettings = ref([])
|
||||
@@ -126,6 +210,15 @@ const loadData = async () => {
|
||||
crawl_enabled: active.crawl_enabled,
|
||||
crawl_keywords: active.crawl_keywords || '',
|
||||
max_pages_per_run: active.max_pages_per_run,
|
||||
generate_prompt: active.generate_prompt || '',
|
||||
optimizer_quality_prompt: active.optimizer_quality_prompt || '',
|
||||
optimizer_polish_prompt: active.optimizer_polish_prompt || '',
|
||||
optimizer_evaluate_prompt: active.optimizer_evaluate_prompt || '',
|
||||
optimizer_quality_temperature: active.optimizer_quality_temperature ?? 0.3,
|
||||
optimizer_polish_temperature: active.optimizer_polish_temperature ?? 0.8,
|
||||
optimizer_evaluate_temperature: active.optimizer_evaluate_temperature ?? 0.3,
|
||||
crawler_extract_prompt: active.crawler_extract_prompt || '',
|
||||
crawler_rewrite_prompt: active.crawler_rewrite_prompt || '',
|
||||
}
|
||||
} catch (e) {
|
||||
// 没有激活配置，使用默认值
|
||||
@@ -147,7 +240,7 @@ const handleSave = async () => {
|
||||
} else {
|
||||
const created = await createSetting(form.value)
|
||||
editingId.value = created.id
|
||||
await toggleSetting(created.id) // 设为激活
|
||||
await toggleSetting(created.id)
|
||||
ElMessage.success('创建并激活成功')
|
||||
}
|
||||
await loadData()
|
||||
@@ -170,6 +263,15 @@ const handleEdit = (row) => {
|
||||
crawl_enabled: row.crawl_enabled,
|
||||
crawl_keywords: row.crawl_keywords || '',
|
||||
max_pages_per_run: row.max_pages_per_run,
|
||||
generate_prompt: row.generate_prompt || '',
|
||||
optimizer_quality_prompt: row.optimizer_quality_prompt || '',
|
||||
optimizer_polish_prompt: row.optimizer_polish_prompt || '',
|
||||
optimizer_evaluate_prompt: row.optimizer_evaluate_prompt || '',
|
||||
optimizer_quality_temperature: row.optimizer_quality_temperature ?? 0.3,
|
||||
optimizer_polish_temperature: row.optimizer_polish_temperature ?? 0.8,
|
||||
optimizer_evaluate_temperature: row.optimizer_evaluate_temperature ?? 0.3,
|
||||
crawler_extract_prompt: row.crawler_extract_prompt || '',
|
||||
crawler_rewrite_prompt: row.crawler_rewrite_prompt || '',
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
@@ -14,6 +14,19 @@ class AiSetting(Base):
|
||||
temperature = Column(Float, nullable=False, default=0.7)
|
||||
max_tokens = Column(Integer, nullable=False, default=2048)
|
||||
|
||||
# 提示词模板
|
||||
generate_prompt = Column(Text, nullable=True, comment="AI 笑话生成提示词")
|
||||
optimizer_quality_prompt = Column(Text, nullable=True, comment="优化器-质量检测提示词")
|
||||
optimizer_polish_prompt = Column(Text, nullable=True, comment="优化器-润色提示词")
|
||||
optimizer_evaluate_prompt = Column(Text, nullable=True, comment="优化器-评价分类提示词")
|
||||
crawler_extract_prompt = Column(Text, nullable=True, comment="爬虫-笑话提取提示词")
|
||||
crawler_rewrite_prompt = Column(Text, nullable=True, comment="爬虫-改写提示词")
|
||||
|
||||
# 优化器各阶段温度
|
||||
optimizer_quality_temperature = Column(Float, nullable=False, default=0.3)
|
||||
optimizer_polish_temperature = Column(Float, nullable=False, default=0.8)
|
||||
optimizer_evaluate_temperature = Column(Float, nullable=False, default=0.3)
|
||||
|
||||
crawl_enabled = Column(Boolean, nullable=False, default=False)
|
||||
crawl_keywords = Column(Text, nullable=True, default="")
|
||||
max_pages_per_run = Column(Integer, nullable=False, default=3)
|
||||
|
||||
+33
-33
@@ -1,8 +1,8 @@
|
||||
"""智能笑话生成器 API"""
|
||||
"""AI 笑话生成器 API — 提示词从数据库读取"""
|
||||
|
||||
import json
|
||||
from datetime import datetime
|
||||
from fastapi import APIRouter, Depends, HTTPException, Query
|
||||
from fastapi import APIRouter, Depends, HTTPException
|
||||
from openai import OpenAI
|
||||
from sqlalchemy.orm import Session
|
||||
|
||||
@@ -13,9 +13,31 @@ from app.schemas.joke import GenerateRequest, GenerateResponse
|
||||
|
||||
router = APIRouter(prefix="/generate", tags=["生成器"])
|
||||
|
||||
STYLE_MAP = {
|
||||
"cold": "冷幽默 / 无厘头",
|
||||
"warm": "温馨幽默 / 暖心搞笑",
|
||||
"twist": "反转 / 神转折",
|
||||
"pun": "谐音梗 / 文字游戏",
|
||||
"sketch": "段子 / 吐槽调侃",
|
||||
"irony": "讽刺幽默 / 黑色幽默",
|
||||
}
|
||||
|
||||
# AI Prompt
|
||||
GENERATION_PROMPT = """你是一位幽默大师，专门创作轻松搞笑的短笑话。
|
||||
LENGTH_MAP = {
|
||||
"short": "30-80字，非常简短",
|
||||
"medium": "80-150字，正常长度",
|
||||
"long": "150-300字，可以描述一个小场景",
|
||||
}
|
||||
|
||||
|
||||
def _load_prompt(setting: AiSetting) -> str:
|
||||
"""从数据库读取生成提示词，没有则返回默认值"""
|
||||
prompt = setting.generate_prompt
|
||||
if not prompt or not prompt.strip():
|
||||
prompt = GENERATE_PROMPT_DEFAULT
|
||||
return prompt
|
||||
|
||||
|
||||
GENERATE_PROMPT_DEFAULT = """你是一位幽默大师，专门创作轻松搞笑的短笑话。
|
||||
|
||||
{context}
|
||||
|
||||
@@ -36,27 +58,13 @@ GENERATION_PROMPT = """你是一位幽默大师，专门创作轻松搞笑的短
|
||||
|
||||
score 是 1-10 的整数评分，reason 是用一句话说明亮点。"""
|
||||
|
||||
STYLE_MAP = {
|
||||
"cold": "冷幽默 / 无厘头",
|
||||
"warm": "温馨幽默 / 暖心搞笑",
|
||||
"twist": "反转 / 神转折",
|
||||
"pun": "谐音梗 / 文字游戏",
|
||||
"sketch": "段子 / 吐槽调侃",
|
||||
"irony": "讽刺幽默 / 黑色幽默",
|
||||
}
|
||||
|
||||
LENGTH_MAP = {
|
||||
"short": "30-80字，非常简短",
|
||||
"medium": "80-150字，正常长度",
|
||||
"long": "150-300字，可以描述一个小场景",
|
||||
}
|
||||
|
||||
|
||||
def _build_prompt(
|
||||
scenarios: list[str],
|
||||
keywords: list[str],
|
||||
style: str = "twist",
|
||||
length: str = "medium",
|
||||
setting: AiSetting | None = None,
|
||||
) -> str:
|
||||
"""构建 AI prompt，支持风格和长度控制"""
|
||||
parts = []
|
||||
@@ -66,7 +74,9 @@ def _build_prompt(
|
||||
parts.append(f"关键词：{', '.join(keywords)}")
|
||||
if not parts:
|
||||
parts.append("场景：日常生活的各种趣事（不指定具体场景）")
|
||||
return GENERATION_PROMPT.format(
|
||||
|
||||
prompt_template = _load_prompt(setting) if setting else GENERATE_PROMPT_DEFAULT
|
||||
return prompt_template.format(
|
||||
context="\n".join(parts),
|
||||
style=STYLE_MAP.get(style, "反转 / 神转折"),
|
||||
length_requirement=LENGTH_MAP.get(length, "80-150字，正常长度"),
|
||||
@@ -78,7 +88,7 @@ def generate_joke(
|
||||
req: GenerateRequest,
|
||||
db: Session = Depends(get_db),
|
||||
):
|
||||
"""调用 AI 生成笑话，支持风格和长度控制"""
|
||||
"""调用 AI 生成笑话，提示词从数据库读取"""
|
||||
# 获取激活的 AI 配置
|
||||
setting = db.query(AiSetting).filter(AiSetting.is_active == True).first()
|
||||
if not setting:
|
||||
@@ -87,10 +97,9 @@ def generate_joke(
|
||||
if not setting or not setting.api_key:
|
||||
raise HTTPException(status_code=503, detail="AI 服务未配置，请联系管理员")
|
||||
|
||||
# 调用 AI
|
||||
try:
|
||||
client = OpenAI(base_url=setting.api_base, api_key=setting.api_key)
|
||||
prompt = _build_prompt(req.scenarios, req.keywords, req.style, req.length)
|
||||
prompt = _build_prompt(req.scenarios, req.keywords, req.style, req.length, setting)
|
||||
|
||||
response = client.chat.completions.create(
|
||||
model=setting.model_name,
|
||||
@@ -113,22 +122,18 @@ def _parse_json_response(raw: str | None) -> GenerateResponse:
|
||||
if not raw or not raw.strip():
|
||||
return GenerateResponse(title="生成的笑话", content="（内容生成失败，请重新生成）", score=0)
|
||||
|
||||
# 尝试从 markdown 代码块中提取 JSON
|
||||
raw = raw.strip()
|
||||
if raw.startswith("```"):
|
||||
lines = raw.split("\n")
|
||||
# 去掉第一行 ```json 和最后一行 ```
|
||||
if len(lines) >= 3:
|
||||
raw = "\n".join(lines[1:-1]).strip()
|
||||
|
||||
# 移除可能的尾部分号
|
||||
if raw.endswith(","):
|
||||
raw = raw[:-1]
|
||||
|
||||
try:
|
||||
data = json.loads(raw)
|
||||
except json.JSONDecodeError:
|
||||
# 如果 JSON 解析失败，尝试查找花括号内的内容
|
||||
try:
|
||||
start = raw.index("{")
|
||||
end = raw.rindex("}") + 1
|
||||
@@ -147,20 +152,17 @@ def _parse_json_response(raw: str | None) -> GenerateResponse:
|
||||
|
||||
|
||||
def _parse_response(raw: str | None) -> GenerateResponse:
|
||||
"""解析 AI 返回内容，提取标题和内容"""
|
||||
# 防御：处理空或 None 输入
|
||||
"""解析 AI 返回内容（非 JSON 回退）"""
|
||||
if not raw or not raw.strip():
|
||||
raise ValueError("AI 返回内容为空")
|
||||
|
||||
title = ""
|
||||
content = raw
|
||||
|
||||
# 尝试提取 "标题：xxx" 或 "标题:xxx"
|
||||
for line in raw.split("\n"):
|
||||
line = line.strip()
|
||||
if line.startswith("标题：") or line.startswith("标题:"):
|
||||
title = line.split("：", 1)[-1].split(":", 1)[-1].strip()
|
||||
# 只替换这一行，不要 replace 全局
|
||||
lines = content.split("\n")
|
||||
for i, l in enumerate(lines):
|
||||
if l.strip() == line:
|
||||
@@ -169,14 +171,12 @@ def _parse_response(raw: str | None) -> GenerateResponse:
|
||||
content = "\n".join(lines).strip()
|
||||
break
|
||||
|
||||
# 如果没有提取到标题，取第一行
|
||||
if not title:
|
||||
first_line = raw.split("\n")[0].strip()
|
||||
if first_line.startswith("标题"):
|
||||
first_line = first_line.split("：", 1)[-1].split(":", 1)[-1].strip()
|
||||
title = first_line[:30] if len(first_line) > 30 else first_line
|
||||
|
||||
# 防御：content 不能为空
|
||||
if not content.strip():
|
||||
content = "（内容生成失败，请重新生成）"
|
||||
|
||||
|
||||
@@ -22,12 +22,12 @@ def list_settings(
|
||||
@router.get("/active")
|
||||
def get_active_setting(
|
||||
db: Session = Depends(get_db),
|
||||
current_user: AdminUser = Depends(get_current_admin_user),
|
||||
):
|
||||
"""获取当前激活的 AI 配置（爬虫调用，无需用户认证，token 校验仍保留）"""
|
||||
"""获取当前激活的 AI 配置（无需认证，供爬虫/优化器使用）"""
|
||||
setting = db.query(AiSetting).filter(AiSetting.is_active == True).first()
|
||||
if not setting:
|
||||
raise HTTPException(status_code=404, detail="未找到激活的 AI 配置")
|
||||
from app.routers.settings import _ensure_default_settings
|
||||
setting = _ensure_default_settings(db)
|
||||
return setting
|
||||
|
||||
|
||||
@@ -94,3 +94,21 @@ def delete_setting(
|
||||
db.delete(db_setting)
|
||||
db.commit()
|
||||
return {"message": "删除成功"}
|
||||
|
||||
|
||||
def _ensure_default_settings(db: Session) -> AiSetting:
|
||||
"""当没有激活配置时，创建一条默认配置"""
|
||||
existing = db.query(AiSetting).first()
|
||||
if existing:
|
||||
existing.is_active = True
|
||||
db.commit()
|
||||
return existing
|
||||
|
||||
defaults = AiSetting(
|
||||
is_active=True,
|
||||
# 默认提示词会在 main.py 迁移时填充
|
||||
)
|
||||
db.add(defaults)
|
||||
db.commit()
|
||||
db.refresh(defaults)
|
||||
return defaults
|
||||
@@ -1,4 +1,4 @@
|
||||
from pydantic import BaseModel, ConfigDict, Field
|
||||
from pydantic import BaseModel, Field
|
||||
|
||||
|
||||
class AiSettingBase(BaseModel):
|
||||
@@ -8,6 +8,20 @@ class AiSettingBase(BaseModel):
|
||||
model_name: str = Field(default="nvidia/llama-3.1-nemotron-70b-instruct")
|
||||
temperature: float = Field(default=0.7, ge=0, le=2)
|
||||
max_tokens: int = Field(default=2048, ge=1)
|
||||
|
||||
# 提示词
|
||||
generate_prompt: str = Field(default="")
|
||||
optimizer_quality_prompt: str = Field(default="")
|
||||
optimizer_polish_prompt: str = Field(default="")
|
||||
optimizer_evaluate_prompt: str = Field(default="")
|
||||
crawler_extract_prompt: str = Field(default="")
|
||||
crawler_rewrite_prompt: str = Field(default="")
|
||||
|
||||
# 优化器各阶段温度
|
||||
optimizer_quality_temperature: float = Field(default=0.3, ge=0, le=2)
|
||||
optimizer_polish_temperature: float = Field(default=0.8, ge=0, le=2)
|
||||
optimizer_evaluate_temperature: float = Field(default=0.3, ge=0, le=2)
|
||||
|
||||
crawl_enabled: bool = Field(default=False)
|
||||
crawl_keywords: str = Field(default="")
|
||||
max_pages_per_run: int = Field(default=3, ge=1)
|
||||
|
||||
+106
-3
@@ -13,13 +13,116 @@ from app.models.feedback import Feedback # 注册模型，确保 create_all 能
|
||||
Base.metadata.create_all(bind=engine)
|
||||
|
||||
# 对已有表新增字段的兼容迁移（SQLite 不支持 ALTER TABLE ADD COLUMN IF NOT EXISTS）
|
||||
# 默认提示词
|
||||
_GENERATE_PROMPT_DEFAULT = """你是一位幽默大师，专门创作轻松搞笑的短笑话。
|
||||
|
||||
{context}
|
||||
|
||||
要求：
|
||||
1. 根据场景和关键词创作一条原创笑话
|
||||
2. 笑话要有反转或意外结局
|
||||
3. 语言风格：{style}
|
||||
4. 字数要求：{length_requirement}
|
||||
5. 直接输出笑话内容，不需要解释
|
||||
|
||||
请严格按照以下 JSON 格式输出，不要加任何额外说明：
|
||||
{{
|
||||
"title": "笑话标题",
|
||||
"content": "笑话正文",
|
||||
"score": 8,
|
||||
"reason": "这个笑话巧妙结合了场景和关键词，结尾有反转"
|
||||
}}
|
||||
|
||||
score 是 1-10 的整数评分，reason 是用一句话说明亮点。"""
|
||||
|
||||
_OPTIMIZER_QUALITY_DEFAULT = """你是一个幽默内容审核专家。判断以下内容是否是一个合格的笑话/段子。
|
||||
|
||||
合格标准（满足任一即可）：
|
||||
1. 有明确的笑点或反转（punchline）
|
||||
2. 有幽默的语言表达或双关
|
||||
3. 有意外结局或情理之中意料之外
|
||||
|
||||
不合格标准（符合任一即判定不合格）：
|
||||
1. 纯粹的事实陈述，没有任何幽默元素
|
||||
2. 只是对话片段，没有笑点
|
||||
3. 普通故事或叙事，没有幽默设计
|
||||
4. 说教或道理阐述
|
||||
5. 内容不完整或难以理解
|
||||
|
||||
始终返回 JSON 格式：{"has_punchline": true/false, "reason": "简要说明判断理由"}"""
|
||||
|
||||
_OPTIMIZER_POLISH_DEFAULT = """你是一个专业的幽默文案编辑。请润色以下笑话，要求：
|
||||
1. 保持核心笑点不变
|
||||
2. 优化语言表达，使其更通顺、更精炼
|
||||
3. 增强节奏感和幽默效果，但不改变原意
|
||||
4. 字数控制在原内容的 80%-120%
|
||||
5. 不要添加额外解释或评论
|
||||
6. 直接输出润色后的内容，不要加任何前缀"""
|
||||
|
||||
_OPTIMIZER_EVALUATE_DEFAULT = """你是一个笑话分类和评价专家。对给定的笑话进行分析，返回 JSON 格式的分类和评分结果。
|
||||
|
||||
要求：
|
||||
1. types: 从提供的类型列表中选择所有匹配的类型名称（数组，可以选多个）
|
||||
2. crowds: 从提供的人群列表中选择所有匹配的人群名称（数组，可以选多个）
|
||||
3. score: 1-10 分，基于幽默程度、创意和表达效果
|
||||
4. comment: 简短评语（10字以内）
|
||||
|
||||
始终返回 JSON 格式。"""
|
||||
|
||||
_CRAWLER_EXTRACT_DEFAULT = """你是一个笑话提取专家。从给定的网页文本中识别并提取所有笑话、幽默段子或有趣内容。
|
||||
要求：
|
||||
1. 只返回真正的笑话内容，不要提取普通文章或新闻
|
||||
2. 每条笑话需要包含：title（简短标题）、content（完整笑话内容）、type（类型）、crowd（人群）
|
||||
3. 如果网页中没有笑话，返回空数组 []
|
||||
4. 永远返回合法的 JSON 格式，根节点为数组或包含 jokes 键的对象"""
|
||||
|
||||
_CRAWLER_REWRITE_DEFAULT = """你是一个幽默作家，负责润色和改写笑话。
|
||||
要求：
|
||||
1. 保持笑话的核心笑点不变
|
||||
2. 语言更通顺、更幽默
|
||||
3. 字数控制在原内容的 80%-120% 之间
|
||||
4. 不要添加任何解释说明"""
|
||||
|
||||
|
||||
def _migrate_db():
|
||||
from sqlalchemy import inspect, text
|
||||
inspector = inspect(engine)
|
||||
columns = [c["name"] for c in inspector.get_columns("jokes")]
|
||||
if "dislike_count" not in columns:
|
||||
columns = [c["name"] for c in inspector.get_columns("ai_settings")]
|
||||
|
||||
# 新增提示词字段
|
||||
prompt_fields = {
|
||||
"generate_prompt": _GENERATE_PROMPT_DEFAULT,
|
||||
"optimizer_quality_prompt": _OPTIMIZER_QUALITY_DEFAULT,
|
||||
"optimizer_polish_prompt": _OPTIMIZER_POLISH_DEFAULT,
|
||||
"optimizer_evaluate_prompt": _OPTIMIZER_EVALUATE_DEFAULT,
|
||||
"crawler_extract_prompt": _CRAWLER_EXTRACT_DEFAULT,
|
||||
"crawler_rewrite_prompt": _CRAWLER_REWRITE_DEFAULT,
|
||||
}
|
||||
for field, default_val in prompt_fields.items():
|
||||
if field not in columns:
|
||||
from sqlalchemy import Text
|
||||
col_type = "TEXT"
|
||||
with engine.connect() as conn:
|
||||
conn.execute(text("ALTER TABLE jokes ADD COLUMN dislike_count INTEGER DEFAULT 0"))
|
||||
conn.execute(text(f"ALTER TABLE ai_settings ADD COLUMN {field} {col_type}"))
|
||||
conn.commit()
|
||||
|
||||
# 新增 temperature 字段
|
||||
temp_fields = [
|
||||
("optimizer_quality_temperature", "FLOAT", "0.3"),
|
||||
("optimizer_polish_temperature", "FLOAT", "0.8"),
|
||||
("optimizer_evaluate_temperature", "FLOAT", "0.3"),
|
||||
]
|
||||
for field, col_type, default_val in temp_fields:
|
||||
if field not in columns:
|
||||
with engine.connect() as conn:
|
||||
conn.execute(text(f"ALTER TABLE ai_settings ADD COLUMN {field} {col_type} DEFAULT {default_val}"))
|
||||
conn.commit()
|
||||
|
||||
# 给新增字段填充默认值（已有记录）
|
||||
with engine.connect() as conn:
|
||||
for field, default_val in prompt_fields.items():
|
||||
conn.execute(text(f"UPDATE ai_settings SET {field} = :val WHERE {field} IS NULL"),
|
||||
{"val": default_val})
|
||||
conn.commit()
|
||||
|
||||
_migrate_db()
|
||||
|
||||
+80
-20
@@ -1,35 +1,90 @@
|
||||
"""LLM 处理：调用 NVIDIA NIM（OpenAI 兼容 API）进行笑话提取、改写和分类。"""
|
||||
"""LLM 处理：调用 NVIDIA NIM（OpenAI 兼容 API）进行笑话提取、改写。"""
|
||||
|
||||
import json
|
||||
import os
|
||||
import httpx
|
||||
from openai import OpenAI
|
||||
|
||||
|
||||
class AiService:
|
||||
def __init__(self, api_base: str, api_key: str, model_name: str, temperature: float = 0.7, max_tokens: int = 2048):
|
||||
self.client = OpenAI(base_url=api_base, api_key=api_key)
|
||||
self.model = model_name
|
||||
self.temperature = temperature
|
||||
self.max_tokens = max_tokens
|
||||
def __init__(self, api_base: str, username: str, password: str):
|
||||
self.api_base = api_base.rstrip("/")
|
||||
self.username = username
|
||||
self.password = password
|
||||
self.token = None
|
||||
self.client = None
|
||||
self.model = ""
|
||||
self.ai_config = None
|
||||
|
||||
def _login(self) -> str:
|
||||
resp = httpx.post(
|
||||
f"{self.api_base}/api/auth/login",
|
||||
json={"username": self.username, "password": self.password},
|
||||
timeout=30,
|
||||
)
|
||||
resp.raise_for_status()
|
||||
return resp.json()["access_token"]
|
||||
|
||||
def _get(self, path: str) -> dict | list:
|
||||
resp = httpx.get(
|
||||
f"{self.api_base}{path}",
|
||||
headers={"Authorization": f"Bearer {self.token}"},
|
||||
timeout=30,
|
||||
)
|
||||
resp.raise_for_status()
|
||||
return resp.json()
|
||||
|
||||
def setup(self):
|
||||
"""登录并从 API 获取 AI 配置（含提示词和温度）"""
|
||||
self.token = self._login()
|
||||
self.ai_config = self._get("/api/admin/settings/active")
|
||||
self.client = OpenAI(
|
||||
base_url=self.ai_config["api_base"],
|
||||
api_key=self.ai_config["api_key"],
|
||||
)
|
||||
self.model = self.ai_config["model_name"]
|
||||
|
||||
def _get_prompt(self, key: str, default: str) -> str:
|
||||
if self.ai_config:
|
||||
val = self.ai_config.get(key)
|
||||
if val and val.strip():
|
||||
return val
|
||||
return default
|
||||
|
||||
def extract_jokes(self, page_content: str, known_types: list[str], known_crowds: list[str]) -> list[dict]:
|
||||
"""从页面内容中提取笑话，返回结构化数据。"""
|
||||
from crawler.prompts import EXTRACTION_SYSTEM_PROMPT, EXTRACTION_USER_PROMPT
|
||||
system_prompt = self._get_prompt("crawler_extract_prompt",
|
||||
"""你是一个笑话提取专家。从给定的网页文本中识别并提取所有笑话、幽默段子或有趣内容。
|
||||
要求：
|
||||
1. 只返回真正的笑话内容，不要提取普通文章或新闻
|
||||
2. 每条笑话需要包含：title（简短标题）、content（完整笑话内容）、type（类型）、crowd（人群）
|
||||
3. 如果网页中没有笑话，返回空数组 []
|
||||
4. 永远返回合法的 JSON 格式，根节点为数组或包含 jokes 键的对象""")
|
||||
|
||||
user_prompt = EXTRACTION_USER_PROMPT.format(
|
||||
page_content=page_content[:8000],
|
||||
known_types=", ".join(known_types),
|
||||
known_crowds=", ".join(known_crowds),
|
||||
)
|
||||
user_prompt = f"""网页内容：
|
||||
---
|
||||
{page_content[:8000]}
|
||||
---
|
||||
|
||||
已知笑话类型：{', '.join(known_types)}
|
||||
已知人群分类：{', '.join(known_crowds)}
|
||||
|
||||
请提取所有笑话，以 JSON 格式返回，示例：
|
||||
[
|
||||
{{"title": "程序员的幽默", "content": "程序员去相亲...", "types": ["谐音梗", "段子"], "crowds": ["职场", "大学生"]}},
|
||||
{{"title": "...", "content": "...", "types": ["..."], "crowds": ["..."]}}
|
||||
]
|
||||
|
||||
注意：types 和 crowds 是数组，可以填多个。
|
||||
只返回 JSON，不要其他文字。"""
|
||||
|
||||
response = self.client.chat.completions.create(
|
||||
model=self.model,
|
||||
messages=[
|
||||
{"role": "system", "content": EXTRACTION_SYSTEM_PROMPT},
|
||||
{"role": "system", "content": system_prompt},
|
||||
{"role": "user", "content": user_prompt},
|
||||
],
|
||||
temperature=self.temperature,
|
||||
max_tokens=self.max_tokens,
|
||||
temperature=self.ai_config.get("temperature", 0.7) if self.ai_config else 0.7,
|
||||
max_tokens=self.ai_config.get("max_tokens", 2048) if self.ai_config else 2048,
|
||||
)
|
||||
|
||||
raw = response.choices[0].message.content
|
||||
@@ -37,13 +92,19 @@ class AiService:
|
||||
|
||||
def rewrite_joke(self, content: str) -> str:
|
||||
"""润色单条笑话内容。"""
|
||||
from crawler.prompts import REWRITE_SYSTEM_PROMPT, REWRITE_USER_PROMPT
|
||||
system_prompt = self._get_prompt("crawler_rewrite_prompt",
|
||||
"""你是一个幽默作家，负责润色和改写笑话。
|
||||
要求：
|
||||
1. 保持笑话的核心笑点不变
|
||||
2. 语言更通顺、更幽默
|
||||
3. 字数控制在原内容的 80%-120% 之间
|
||||
4. 不要添加任何解释说明""")
|
||||
|
||||
response = self.client.chat.completions.create(
|
||||
model=self.model,
|
||||
messages=[
|
||||
{"role": "system", "content": REWRITE_SYSTEM_PROMPT},
|
||||
{"role": "user", "content": REWRITE_USER_PROMPT.format(content=content)},
|
||||
{"role": "system", "content": system_prompt},
|
||||
{"role": "user", "content": f"请润色以下笑话：\n\n{content}"},
|
||||
],
|
||||
temperature=0.8,
|
||||
max_tokens=500,
|
||||
@@ -60,7 +121,6 @@ class AiService:
|
||||
jokes = data
|
||||
return [j for j in jokes if isinstance(j, dict) and j.get("content")]
|
||||
except json.JSONDecodeError:
|
||||
# 尝试提取 markdown 代码块
|
||||
if "```json" in raw:
|
||||
raw = raw.split("```json")[1].split("```")[0]
|
||||
elif "```" in raw:
|
||||
|
||||
@@ -57,15 +57,13 @@ class Processor:
|
||||
self.token = self._login()
|
||||
print("[*] 登录成功")
|
||||
|
||||
ai_config = self._get("/api/admin/settings/active")
|
||||
self.ai = AiService(
|
||||
api_base=ai_config["api_base"],
|
||||
api_key=ai_config["api_key"],
|
||||
model_name=ai_config["model_name"],
|
||||
temperature=ai_config.get("temperature", 0.7),
|
||||
max_tokens=ai_config.get("max_tokens", 2048),
|
||||
api_base=self.api_base,
|
||||
username=self.username,
|
||||
password=self.password,
|
||||
)
|
||||
print(f"[*] AI 配置: {ai_config['model_name']}")
|
||||
self.ai.setup()
|
||||
print(f"[*] AI 配置: {self.ai.model}")
|
||||
|
||||
self.types = self._get("/api/categories/types")
|
||||
self.crowds = self._get("/api/categories/crowds")
|
||||
|
||||
@@ -1,41 +0,0 @@
|
||||
"""AI 提示词模板。"""
|
||||
|
||||
# ===== 笑话提取 =====
|
||||
EXTRACTION_SYSTEM_PROMPT = """你是一个笑话提取专家。从给定的网页文本中识别并提取所有笑话、幽默段子或有趣内容。
|
||||
要求：
|
||||
1. 只返回真正的笑话内容，不要提取普通文章或新闻
|
||||
2. 每条笑话需要包含：title（简短标题）、content（完整笑话内容）、type（类型）、crowd（人群）
|
||||
3. 如果网页中没有笑话，返回空数组 []
|
||||
4. 永远返回合法的 JSON 格式，根节点为数组或包含 jokes 键的对象"""
|
||||
|
||||
EXTRACTION_USER_PROMPT = """网页内容：
|
||||
---
|
||||
{page_content}
|
||||
---
|
||||
|
||||
已知笑话类型：{known_types}
|
||||
已知人群分类：{known_crowds}
|
||||
|
||||
请提取所有笑话，以 JSON 格式返回，示例：
|
||||
[
|
||||
{{"title": "程序员的幽默", "content": "程序员去相亲...", "types": ["谐音梗", "段子"], "crowds": ["职场", "大学生"]}},
|
||||
{{"title": "...", "content": "...", "types": ["..."], "crowds": ["..."]}}
|
||||
]
|
||||
|
||||
注意：types 和 crowds 是数组，可以填多个。
|
||||
只返回 JSON，不要其他文字。"""
|
||||
|
||||
|
||||
# ===== 笑话改写 =====
|
||||
REWRITE_SYSTEM_PROMPT = """你是一个幽默作家，负责润色和改写笑话。
|
||||
要求：
|
||||
1. 保持笑话的核心笑点不变
|
||||
2. 语言更通顺、更幽默
|
||||
3. 字数控制在原内容的 80%-120% 之间
|
||||
4. 不要添加任何解释说明"""
|
||||
|
||||
REWRITE_USER_PROMPT = """请润色以下笑话：
|
||||
|
||||
{content}
|
||||
|
||||
只返回润色后的笑话文字，不要其他内容。"""
|
||||
+76
-23
@@ -8,15 +8,6 @@ import time
|
||||
import httpx
|
||||
from openai import OpenAI
|
||||
|
||||
from optimizer.prompts import (
|
||||
QUALITY_CHECK_SYSTEM_PROMPT,
|
||||
QUALITY_CHECK_USER_PROMPT,
|
||||
POLISH_SYSTEM_PROMPT,
|
||||
POLISH_USER_PROMPT,
|
||||
EVALUATE_SYSTEM_PROMPT,
|
||||
EVALUATE_USER_PROMPT,
|
||||
)
|
||||
|
||||
|
||||
class Optimizer:
|
||||
def __init__(self, api_base: str, username: str, password: str):
|
||||
@@ -26,6 +17,7 @@ class Optimizer:
|
||||
self.token = None
|
||||
self.ai_client = None
|
||||
self.model_name = ""
|
||||
self.ai_config = None # 完整 AI 配置（含提示词和温度）
|
||||
self.types = []
|
||||
self.crowds = []
|
||||
# 统计
|
||||
@@ -67,6 +59,7 @@ class Optimizer:
|
||||
print("[*] 登录成功")
|
||||
|
||||
ai_config = self._get("/api/admin/settings/active")
|
||||
self.ai_config = ai_config
|
||||
self.ai_client = OpenAI(
|
||||
base_url=ai_config["api_base"],
|
||||
api_key=ai_config["api_key"],
|
||||
@@ -78,6 +71,22 @@ class Optimizer:
|
||||
self.crowds = self._get("/api/categories/crowds")
|
||||
print(f"[*] 分类: {len(self.types)} 种类型, {len(self.crowds)} 种人群")
|
||||
|
||||
def _get_prompt(self, key: str, default: str) -> str:
|
||||
"""从数据库配置中读取提示词，没有则返回默认"""
|
||||
if self.ai_config:
|
||||
val = self.ai_config.get(key)
|
||||
if val and val.strip():
|
||||
return val
|
||||
return default
|
||||
|
||||
def _get_temp(self, key: str, default: float) -> float:
|
||||
"""从数据库配置中读取温度"""
|
||||
if self.ai_config:
|
||||
val = self.ai_config.get(key)
|
||||
if val is not None:
|
||||
return float(val)
|
||||
return default
|
||||
|
||||
# === 读取笑话 ===
|
||||
def get_jokes(self, status: str | None = None, limit: int | None = None,
|
||||
ids: list[int] | None = None) -> list[dict]:
|
||||
@@ -120,14 +129,32 @@ class Optimizer:
|
||||
# === Stage 1: 质量检测 ===
|
||||
def quality_check(self, content: str) -> dict:
|
||||
"""判断笑话是否有笑点，返回 {"has_punchline": bool, "reason": str}"""
|
||||
system_prompt = self._get_prompt("optimizer_quality_prompt",
|
||||
"""你是一个幽默内容审核专家。判断以下内容是否是一个合格的笑话/段子。
|
||||
|
||||
合格标准（满足任一即可）：
|
||||
1. 有明确的笑点或反转（punchline）
|
||||
2. 有幽默的语言表达或双关
|
||||
3. 有意外结局或情理之中意料之外
|
||||
|
||||
不合格标准（符合任一即判定不合格）：
|
||||
1. 纯粹的事实陈述，没有任何幽默元素
|
||||
2. 只是对话片段，没有笑点
|
||||
3. 普通故事或叙事，没有幽默设计
|
||||
4. 说教或道理阐述
|
||||
5. 内容不完整或难以理解
|
||||
|
||||
始终返回 JSON 格式：{"has_punchline": true/false, "reason": "简要说明判断理由"}""")
|
||||
temperature = self._get_temp("optimizer_quality_temperature", 0.3)
|
||||
|
||||
def _call():
|
||||
return self.ai_client.chat.completions.create(
|
||||
model=self.model_name,
|
||||
messages=[
|
||||
{"role": "system", "content": QUALITY_CHECK_SYSTEM_PROMPT},
|
||||
{"role": "user", "content": QUALITY_CHECK_USER_PROMPT.format(content=content[:2000])},
|
||||
{"role": "system", "content": system_prompt},
|
||||
{"role": "user", "content": f"请判断以下内容是否为合格笑话：\n\n{content[:2000]}\n\n返回 JSON 格式。"},
|
||||
],
|
||||
temperature=0.3,
|
||||
temperature=temperature,
|
||||
max_tokens=200,
|
||||
)
|
||||
resp = self._safe_api_call(_call)
|
||||
@@ -137,14 +164,24 @@ class Optimizer:
|
||||
# === Stage 2: AI 润色 ===
|
||||
def polish(self, content: str) -> str:
|
||||
"""润色笑话内容"""
|
||||
system_prompt = self._get_prompt("optimizer_polish_prompt",
|
||||
"""你是一个专业的幽默文案编辑。请润色以下笑话，要求：
|
||||
1. 保持核心笑点不变
|
||||
2. 优化语言表达，使其更通顺、更精炼
|
||||
3. 增强节奏感和幽默效果，但不改变原意
|
||||
4. 字数控制在原内容的 80%-120%
|
||||
5. 不要添加额外解释或评论
|
||||
6. 直接输出润色后的内容，不要加任何前缀""")
|
||||
temperature = self._get_temp("optimizer_polish_temperature", 0.8)
|
||||
|
||||
def _call():
|
||||
return self.ai_client.chat.completions.create(
|
||||
model=self.model_name,
|
||||
messages=[
|
||||
{"role": "system", "content": POLISH_SYSTEM_PROMPT},
|
||||
{"role": "user", "content": POLISH_USER_PROMPT.format(content=content)},
|
||||
{"role": "system", "content": system_prompt},
|
||||
{"role": "user", "content": f"请润色以下笑话：\n\n{content}"},
|
||||
],
|
||||
temperature=0.8,
|
||||
temperature=temperature,
|
||||
max_tokens=1024,
|
||||
)
|
||||
resp = self._safe_api_call(_call)
|
||||
@@ -156,24 +193,40 @@ class Optimizer:
|
||||
type_names = [t.get("name", "") for t in self.types]
|
||||
crowd_names = [c.get("name", "") for c in self.crowds]
|
||||
|
||||
system_prompt = self._get_prompt("optimizer_evaluate_prompt",
|
||||
"""你是一个笑话分类和评价专家。对给定的笑话进行分析，返回 JSON 格式的分类和评分结果。
|
||||
|
||||
要求：
|
||||
1. types: 从提供的类型列表中选择所有匹配的类型名称（数组，可以选多个）
|
||||
2. crowds: 从提供的人群列表中选择所有匹配的人群名称（数组，可以选多个）
|
||||
3. score: 1-10 分，基于幽默程度、创意和表达效果
|
||||
4. comment: 简短评语（10字以内）
|
||||
|
||||
始终返回 JSON 格式。""")
|
||||
temperature = self._get_temp("optimizer_evaluate_temperature", 0.3)
|
||||
|
||||
user_content = f"""笑话内容：
|
||||
{content[:2000]}
|
||||
|
||||
可选类型：{', '.join(type_names)}
|
||||
可选人群：{', '.join(crowd_names)}
|
||||
|
||||
返回 JSON 格式：{{"types": ["类型1", "类型2"], "crowds": ["人群1", "人群2"], "score": 8, "comment": "简短评语"}}"""
|
||||
|
||||
def _call():
|
||||
return self.ai_client.chat.completions.create(
|
||||
model=self.model_name,
|
||||
messages=[
|
||||
{"role": "system", "content": EVALUATE_SYSTEM_PROMPT},
|
||||
{"role": "user", "content": EVALUATE_USER_PROMPT.format(
|
||||
content=content[:2000],
|
||||
known_types=", ".join(type_names),
|
||||
known_crowds=", ".join(crowd_names),
|
||||
)},
|
||||
{"role": "system", "content": system_prompt},
|
||||
{"role": "user", "content": user_content},
|
||||
],
|
||||
temperature=0.3,
|
||||
temperature=temperature,
|
||||
max_tokens=300,
|
||||
)
|
||||
resp = self._safe_api_call(_call)
|
||||
raw = resp.choices[0].message.content.strip()
|
||||
result = self._parse_json(raw, {"types": [], "crowds": [], "score": 5, "comment": ""})
|
||||
# Backward compatibility: if LLM returns old single format, convert to array
|
||||
# Backward compatibility
|
||||
if isinstance(result.get("types"), str):
|
||||
result["types"] = [result["types"]] if result["types"] else []
|
||||
if isinstance(result.get("crowds"), str):
|
||||
|
||||
@@ -1,58 +0,0 @@
|
||||
"""AI 提示词模板 — 笑话质量检测、润色、评价分类。"""
|
||||
|
||||
# ===== Stage 1: 质量检测 =====
|
||||
QUALITY_CHECK_SYSTEM_PROMPT = """你是一个幽默内容审核专家。判断以下内容是否是一个合格的笑话/段子。
|
||||
|
||||
合格标准（满足任一即可）：
|
||||
1. 有明确的笑点或反转（punchline）
|
||||
2. 有幽默的语言表达或双关
|
||||
3. 有意外结局或情理之中意料之外
|
||||
|
||||
不合格标准（符合任一即判定不合格）：
|
||||
1. 纯粹的事实陈述，没有任何幽默元素
|
||||
2. 只是对话片段，没有笑点
|
||||
3. 普通故事或叙事，没有幽默设计
|
||||
4. 说教或道理阐述
|
||||
5. 内容不完整或难以理解
|
||||
|
||||
始终返回 JSON 格式：{"has_punchline": true/false, "reason": "简要说明判断理由"}"""
|
||||
|
||||
QUALITY_CHECK_USER_PROMPT = """请判断以下内容是否为合格笑话：
|
||||
|
||||
{content}
|
||||
|
||||
返回 JSON 格式。"""
|
||||
|
||||
# ===== Stage 2: AI 润色 =====
|
||||
POLISH_SYSTEM_PROMPT = """你是一个专业的幽默文案编辑。请润色以下笑话，要求：
|
||||
1. 保持核心笑点不变
|
||||
2. 优化语言表达，使其更通顺、更精炼
|
||||
3. 增强节奏感和幽默效果，但不改变原意
|
||||
4. 字数控制在原内容的 80%-120%
|
||||
5. 不要添加额外解释或评论
|
||||
6. 直接输出润色后的内容，不要加任何前缀"""
|
||||
|
||||
POLISH_USER_PROMPT = """请润色以下笑话：
|
||||
|
||||
{content}
|
||||
|
||||
只输出润色后的笑话内容。"""
|
||||
|
||||
# ===== Stage 3: 评价分类 =====
|
||||
EVALUATE_SYSTEM_PROMPT = """你是一个笑话分类和评价专家。对给定的笑话进行分析，返回 JSON 格式的分类和评分结果。
|
||||
|
||||
要求：
|
||||
1. types: 从提供的类型列表中选择所有匹配的类型名称（数组，可以选多个）
|
||||
2. crowds: 从提供的人群列表中选择所有匹配的人群名称（数组，可以选多个）
|
||||
3. score: 1-10 分，基于幽默程度、创意和表达效果
|
||||
4. comment: 简短评语（10字以内）
|
||||
|
||||
始终返回 JSON 格式。"""
|
||||
|
||||
EVALUATE_USER_PROMPT = """笑话内容：
|
||||
{content}
|
||||
|
||||
可选类型：{known_types}
|
||||
可选人群：{known_crowds}
|
||||
|
||||
返回 JSON 格式：{{"types": ["类型1", "类型2"], "crowds": ["人群1", "人群2"], "score": 8, "comment": "简短评语"}}"""
|
||||
Reference in New Issue
Block a user