修改了poster_gen的目录路径
This commit is contained in:
parent
59bb2d9865
commit
b554e13025
340
README.md
340
README.md
@ -1,260 +1,184 @@
|
||||
# 旅游内容创作工具 (Travel Content Creator)
|
||||
# TravelContentCreator
|
||||
|
||||
这是一个基于AI的旅游内容自动生成工具，可以根据景点信息自动生成高质量的旅游推文和宣传海报。
|
||||
TravelContentCreator是一个用于自动化生成旅游内容和宣传海报的系统。该系统利用AI技术生成景点选题、撰写文章内容并制作相应的宣传海报。
|
||||
|
||||
## 功能特点
|
||||
|
||||
- **自动选题生成**：根据提供的景点信息和配置的提示词模板，自动生成吸引人的旅游选题
|
||||
- **内容创作**：基于选题和配置的提示词模板，自动生成文字内容（标题、正文）
|
||||
- **海报制作**：结合景点图片和生成的文字内容，自动创建精美的宣传海报
|
||||
- **批量处理**：支持一次性生成多个选题和多个变体内容
|
||||
- **模块化设计**：核心功能（配置加载、提示词管理、AI交互、选题、内容生成、海报制作）分离，方便维护和扩展
|
||||
- **配置驱动**：通过配置文件集中管理所有运行参数
|
||||
- 自动生成旅游景点选题（包括目标受众、写作风格等）
|
||||
- 根据选题生成详细的旅游文章内容
|
||||
- 为生成的文章制作美观的宣传海报
|
||||
- 支持多种写作风格和目标受众需求
|
||||
- 文件模糊匹配功能，增强系统健壮性
|
||||
- 模块化设计，可单独使用选题、内容生成或海报生成功能
|
||||
|
||||
## 新功能: 流式输出处理
|
||||
## 目录结构
|
||||
|
||||
TravelContentCreator 现已支持三种流式输出处理方法，提供了更灵活的 AI 文本生成体验：
|
||||
```
|
||||
TravelContentCreator/
|
||||
├── core/ # 核心功能模块
|
||||
├── utils/ # 工具类和辅助函数
|
||||
├── genPrompts/ # 生成提示词
|
||||
│ ├── Style/ # 风格提示词
|
||||
│ ├── Demand/ # 需求提示词
|
||||
│ └── Refer/ # 参考提示词
|
||||
├── SelectPrompt/ # 选题提示词
|
||||
├── resource/ # 资源文件
|
||||
├── main.py # 主程序
|
||||
├── test_topic_content.py # 选题和内容生成测试脚本
|
||||
├── test_poster.py # 海报生成测试脚本
|
||||
├── topic_content_config.json # 选题和内容生成配置
|
||||
├── poster_config.json # 海报生成配置
|
||||
└── poster_gen_config.json # 主程序配置
|
||||
```
|
||||
|
||||
- **同步流式响应**: 使用流式 API 但返回完整响应
|
||||
- **回调式流式响应**: 通过回调函数处理每个文本块
|
||||
- **异步流式响应**: 使用异步生成器返回文本流
|
||||
|
||||
这些功能大大提升了长文本生成的用户体验和系统响应性。
|
||||
|
||||
详细文档请参阅:
|
||||
- [流式处理文档](docs/streaming.md)
|
||||
- [流式处理演示](examples/test_stream.py)
|
||||
|
||||
## 快速开始
|
||||
|
||||
### 1. 环境准备
|
||||
## 安装
|
||||
|
||||
1. 克隆仓库
|
||||
```bash
|
||||
# 克隆项目
|
||||
git clone https://github.com/yourusername/TravelContentCreator.git
|
||||
git clone [仓库URL]
|
||||
cd TravelContentCreator
|
||||
|
||||
# 安装依赖 (假设有requirements.txt文件)
|
||||
# pip install -r requirements.txt
|
||||
# 或者手动安装
|
||||
pip install numpy pandas opencv-python pillow requests tqdm
|
||||
```
|
||||
|
||||
### 2. 配置设置
|
||||
2. 安装依赖
|
||||
```bash
|
||||
pip install -r requirements.txt
|
||||
```
|
||||
|
||||
3. 配置环境
|
||||
- 确保安装了Python 3.6+
|
||||
- 配置好AI模型API（本系统默认使用QwenAPI）
|
||||
- 准备好必要的素材和资源文件
|
||||
|
||||
## 使用说明
|
||||
|
||||
### 1. 主程序
|
||||
|
||||
主程序可以执行完整的流程，包括选题生成、内容生成和海报生成：
|
||||
|
||||
```bash
|
||||
# 复制示例配置（选择一个或从基础开始）
|
||||
cp configs/basic_config.json poster_gen_config.json
|
||||
|
||||
# 编辑配置文件
|
||||
vim poster_gen_config.json
|
||||
# 必须修改：api_url, api_key, image_base_dir
|
||||
python main.py [--config CONFIG_PATH] [--run_id RUN_ID] [--topics_file TOPICS_FILE] [--debug]
|
||||
```
|
||||
|
||||
### 3. 运行系统
|
||||
参数说明：
|
||||
- `--config`: 配置文件路径，默认为`poster_gen_config.json`
|
||||
- `--run_id`: 自定义运行ID，用于区分不同批次的生成结果
|
||||
- `--topics_file`: 预生成的选题文件路径，如果提供则跳过选题生成步骤
|
||||
- `--debug`: 启用调试级别日志
|
||||
|
||||
### 2. 测试选题和内容生成
|
||||
|
||||
使用专门的测试脚本运行选题和内容生成模块：
|
||||
|
||||
```bash
|
||||
# 完整流程（从选题到海报生成）
|
||||
python main.py
|
||||
|
||||
# 或分阶段执行 (使用默认配置)
|
||||
python examples/run_step1_topics.py
|
||||
# 记下输出的Run ID
|
||||
python examples/run_step2_content_posters.py YOUR_RUN_ID
|
||||
|
||||
# 使用特定配置运行
|
||||
# python main.py --config configs/social_media_config.json
|
||||
python test_topic_content.py [--config CONFIG_PATH] [--run_id RUN_ID] [--topics_file TOPICS_FILE] [--debug]
|
||||
```
|
||||
|
||||
### 4. 查看结果
|
||||
参数说明：
|
||||
- `--config`: 配置文件路径，默认为`topic_content_config.json`
|
||||
- `--run_id`: 自定义运行ID
|
||||
- `--topics_file`: 预生成的选题文件路径，如果提供则跳过选题生成
|
||||
- `--debug`: 启用调试级别日志
|
||||
|
||||
### 3. 测试海报生成
|
||||
|
||||
使用专门的测试脚本运行海报生成模块：
|
||||
|
||||
```bash
|
||||
# 结果保存在配置的output_dir目录下（默认为./result/）
|
||||
ls -la ./result/最新的Run_ID/
|
||||
python test_poster.py --topics_file TOPICS_FILE [--config CONFIG_PATH] [--topic_index TOPIC_INDEX] [--run_id RUN_ID] [--debug]
|
||||
```
|
||||
|
||||
## 核心组件说明
|
||||
参数说明：
|
||||
- `--topics_file`: 必需的选题JSON文件路径，用于获取海报生成的主题数据
|
||||
- `--config`: 配置文件路径，默认为`poster_config.json`
|
||||
- `--topic_index`: 要生成海报的特定选题索引，如果未提供则为所有选题生成海报
|
||||
- `--run_id`: 自定义运行ID
|
||||
- `--debug`: 启用调试级别日志
|
||||
|
||||
项目采用模块化设计，主要包含以下组件：
|
||||
## 配置文件说明
|
||||
|
||||
- **主流程协调器** (`main.py`): 负责加载配置并协调执行整个生成流程
|
||||
- **AI交互模块** (`core/ai_agent.py`): 封装与大语言模型的通信
|
||||
- **选题生成器** (`utils/tweet_generator.py`): 生成旅游选题
|
||||
- **内容生成器** (`core/contentGen.py`): 处理内容创作
|
||||
- **海报制作器** (`core/posterGen.py`): 合成文字和图片，生成最终海报
|
||||
|
||||
## 资源准备指南
|
||||
|
||||
### 1. 景点信息文件
|
||||
|
||||
在 `resource/Object/` 目录创建景点信息文件，示例格式：
|
||||
|
||||
```
|
||||
景点名称：泰宁古城
|
||||
位置：福建省三明市泰宁县
|
||||
简介：泰宁古城始建于宋代...
|
||||
特色：古城墙、古街巷...
|
||||
历史：泰宁古城有着悠久的历史...
|
||||
适合游客：喜欢历史文化的游客、摄影爱好者
|
||||
建议游览时间：2-3小时
|
||||
最佳季节：春季和秋季
|
||||
```
|
||||
|
||||
> **提示**：景点信息越详细，生成的内容质量越高
|
||||
|
||||
### 2. 图片资源结构
|
||||
|
||||
图片资源应按以下结构组织（可通过配置自定义目录名）：
|
||||
|
||||
```
|
||||
<image_base_dir>/ # 配置中的图片根目录
|
||||
├── 相机/ # 存放原始照片 (camera_image_subdir)
|
||||
│ ├── 泰宁古城/
|
||||
│ │ ├── 图片1.jpg
|
||||
│ │ ├── 图片2.jpg
|
||||
│ │ └── description.txt (可选的图片描述)
|
||||
│ └── 其他景点/
|
||||
└── modify/ # 存放处理后的图片 (modify_image_subdir)
|
||||
├── 泰宁古城/
|
||||
│ ├── 图片1.jpg
|
||||
│ └── ...
|
||||
└── 其他景点/
|
||||
```
|
||||
|
||||
> **重要**：确保每个景点的图片目录名与景点信息文件中的名称匹配。海报生成默认从 `modify/` 目录选取图片。
|
||||
|
||||
## 配置文件详解
|
||||
|
||||
`poster_gen_config.json` 是系统的核心配置文件，包含以下主要配置项：
|
||||
|
||||
### 基础配置
|
||||
### 选题和内容生成配置 (topic_content_config.json)
|
||||
|
||||
```json
|
||||
{
|
||||
"date": "5月15日", // 日期标记，用于提示词
|
||||
"num": 5, // 生成选题数量
|
||||
"variants": 3 // 每个选题生成的变体数量
|
||||
}
|
||||
```
|
||||
|
||||
### AI模型配置
|
||||
|
||||
```json
|
||||
"date": "5月15日, 5月16日, 5月17日, 6月1日", // 选题日期
|
||||
"num": 2, // 生成选题数量
|
||||
"variants": 1, // 每个选题的变体数量
|
||||
"topic_temperature": 0.2, // 选题生成的temperature参数
|
||||
"content_temperature": 0.3, // 内容生成的temperature参数
|
||||
"model": "qwenQWQ", // 使用的AI模型
|
||||
"api_url": "http://localhost:8000/v1/", // API地址
|
||||
"api_key": "EMPTY", // API密钥
|
||||
"topic_system_prompt": "./SelectPrompt/systemPrompt.txt", // 选题系统提示词路径
|
||||
"topic_user_prompt": "./SelectPrompt/userPrompt.txt", // 选题用户提示词路径
|
||||
"content_system_prompt": "./genPrompts/systemPrompt.txt", // 内容系统提示词路径
|
||||
"prompts_config": [ // 提示词配置
|
||||
{
|
||||
"model": "qwen", // 使用的模型名称
|
||||
"api_url": "http://localhost:8000/v1/", // API端点
|
||||
"api_key": "YOUR_API_KEY", // API密钥
|
||||
"topic_temperature": 0.2, // 选题生成的随机性
|
||||
"content_temperature": 0.3 // 内容生成的随机性
|
||||
}
|
||||
```
|
||||
|
||||
### 资源路径配置
|
||||
|
||||
```json
|
||||
"type": "Style", // 风格提示词
|
||||
"file_path": [...] // 风格提示词文件路径列表
|
||||
},
|
||||
{
|
||||
"resource_dir": [ // 景点信息资源
|
||||
"type": "Demand", // 需求提示词
|
||||
"file_path": [...] // 需求提示词文件路径列表
|
||||
},
|
||||
...
|
||||
],
|
||||
"resource_dir": [ // 资源目录配置
|
||||
{
|
||||
"type": "Object",
|
||||
"num": 3,
|
||||
"file_path": [
|
||||
"./resource/Object/景点信息-泰宁古城.txt",
|
||||
"./resource/Object/景点信息-尚书第.txt"
|
||||
"type": "Object", // 对象类型
|
||||
"file_path": [...] // 对象文件路径列表
|
||||
},
|
||||
...
|
||||
]
|
||||
}
|
||||
],
|
||||
"image_base_dir": "/path/to/your/image/directory", // 图片根目录
|
||||
"camera_image_subdir": "相机", // 原始照片子目录
|
||||
"modify_image_subdir": "modify" // 处理后图片子目录
|
||||
}
|
||||
```
|
||||
|
||||
### 提示词配置
|
||||
|
||||
```json
|
||||
{
|
||||
"topic_system_prompt": "./SelectPrompt/systemPrompt.txt",
|
||||
"topic_user_prompt": "./SelectPrompt/userPrompt.txt",
|
||||
"content_system_prompt": "./genPrompts/systemPrompt.txt",
|
||||
"prompts_dir": "./genPrompts"
|
||||
}
|
||||
```
|
||||
|
||||
### 输出配置
|
||||
### 海报生成配置 (poster_config.json)
|
||||
|
||||
```json
|
||||
{
|
||||
"variants": 1, // 每个选题的变体数量
|
||||
"model": "qwenQWQ", // 使用的AI模型
|
||||
"api_url": "http://localhost:8000/v1/", // API地址
|
||||
"api_key": "EMPTY", // API密钥
|
||||
"poster_content_system_prompt": "./genPrompts/poster_content_systemPrompt.txt", // 海报内容系统提示词路径
|
||||
"resource_dir": [...], // 资源目录配置
|
||||
"output_dir": "./result", // 输出目录
|
||||
"poster_target_size": [900, 1200], // 海报尺寸
|
||||
"text_possibility": 0.3 // 文字元素出现概率
|
||||
"image_base_dir": "...", // 图片基础目录
|
||||
"poster_assets_base_dir": "...", // 海报素材基础目录
|
||||
"poster_target_size": [900, 1200], // 海报目标尺寸
|
||||
"text_possibility": 0.3, // 文本可能性
|
||||
"img_frame_possibility": 0.7, // 图像框可能性
|
||||
"text_bg_possibility": 0 // 文本背景可能性
|
||||
}
|
||||
```
|
||||
|
||||
## 配置示例
|
||||
## 结果输出
|
||||
|
||||
本项目提供了多种预设配置文件，适用于不同场景。这些配置文件位于 `configs/` 目录下：
|
||||
生成的结果保存在配置文件中指定的`output_dir`目录下，按照`run_id`组织。每次运行的结果包括：
|
||||
|
||||
- **基础配置** (`configs/basic_config.json`): 适合初次使用和测试
|
||||
- **OpenAI配置** (`configs/openai_config.json`): 使用OpenAI API的配置
|
||||
- **高质量配置** (`configs/high_quality_config.json`): 更高质量的生成设置
|
||||
- **批量处理配置** (`configs/batch_processing_config.json`): 处理大量景点信息
|
||||
- **社交媒体配置** (`configs/social_media_config.json`): 针对多个社交平台优化
|
||||
- **本地LLM配置** (`configs/local_llm_config.json`): 使用本地部署的LLM模型
|
||||
- 选题文件：`tweet_topic_{run_id}.json`
|
||||
- 选题使用的提示词：`tweet_prompt_{run_id}.txt`
|
||||
- 文章内容：分目录保存在`{run_id}/{topic_index}_{variant_index}/article.json`
|
||||
- 海报图像：保存在`{run_id}/{topic_index}_{variant_index}/poster/poster.jpg`
|
||||
- 拼贴图像：保存在`{run_id}/{topic_index}_{variant_index}/collage_img/`
|
||||
|
||||
使用示例配置：
|
||||
## 自定义扩展
|
||||
|
||||
```bash
|
||||
# 复制适合您场景的配置
|
||||
cp configs/social_media_config.json poster_gen_config.json
|
||||
系统各部分设计为模块化，您可以：
|
||||
|
||||
# 按需修改配置
|
||||
vim poster_gen_config.json
|
||||
```
|
||||
1. 添加新的风格提示词到`genPrompts/Style/`目录
|
||||
2. 添加新的需求提示词到`genPrompts/Demand/`目录
|
||||
3. 添加新的景点资源到`resource/Object/`目录
|
||||
4. 修改AI模型参数以适应不同生成需求
|
||||
5. 自定义海报生成的尺寸和样式
|
||||
|
||||
详细说明请参阅 `configs/README.md` 文件。
|
||||
## 注意事项
|
||||
|
||||
## 高级使用指南
|
||||
- 确保提示词文件和资源文件的编码为UTF-8
|
||||
- API密钥应妥善保管，建议使用环境变量或外部配置
|
||||
- 图像生成需要足够的系统资源，建议在性能良好的设备上运行
|
||||
- 文件模糊匹配功能可以处理一些文件名不完全匹配的情况，但建议尽量保持文件名规范
|
||||
|
||||
### 自定义提示词
|
||||
## 贡献
|
||||
|
||||
编辑 `SelectPrompt/` 和 `genPrompts/` 目录下的提示词文件，可自定义生成内容的风格和侧重点。
|
||||
|
||||
### 调整生成参数
|
||||
|
||||
- 增加 `variants` 值可获得更多内容变体
|
||||
- 调整 `temperature` 参数可以改变生成内容的创造性
|
||||
- 修改 `poster_target_size` 可以设置不同的海报尺寸
|
||||
|
||||
### 分布式执行
|
||||
|
||||
利用分阶段执行功能，可在不同机器上完成选题生成和内容生成：
|
||||
|
||||
1. 机器A执行选题生成 (`run_step1_topics.py`)，将结果保存到共享存储
|
||||
2. 机器B从共享存储读取选题 (`run_step2_content_posters.py <run_id>`)，执行计算密集的内容和海报生成
|
||||
|
||||
## 常见问题
|
||||
|
||||
1. **生成内容质量不高？**
|
||||
- 尝试提供更详细的景点信息
|
||||
- 调整提示词模板
|
||||
- 降低 `temperature` 参数以减少随机性
|
||||
|
||||
2. **找不到景点图片？**
|
||||
- 确保图片目录名与景点信息匹配
|
||||
- 检查配置文件中的 `image_base_dir` 路径是否正确
|
||||
|
||||
3. **API调用失败？**
|
||||
- 验证 API Key 和 URL 是否正确
|
||||
- 检查网络连接和防火墙设置
|
||||
|
||||
## 示例
|
||||
|
||||
查看 `examples/` 目录中的示例脚本及其 `README.md` 文件，了解更多使用方法。
|
||||
|
||||
## 贡献指南
|
||||
|
||||
欢迎提交 Pull Request 或 Issue 来帮助改进本项目。
|
||||
|
||||
## 许可证
|
||||
|
||||
本项目采用 MIT 许可证。
|
||||
欢迎提交问题报告和功能建议，或直接提交代码改进。
|
||||
|
||||
@ -1,7 +1,7 @@
|
||||
from .ai_agent import AI_Agent
|
||||
from .topic_parser import TopicParser
|
||||
from .contentGen import ContentGenerator
|
||||
from .posterGen import PosterGenerator
|
||||
from .poster_gen import PosterGenerator
|
||||
from .simple_collage import process_directory
|
||||
|
||||
__all__ = ['AI_Agent', 'TopicParser', 'ContentGenerator', 'PosterGenerator', 'process_directory']
|
||||
Binary file not shown.
BIN
core/__pycache__/poster_gen.cpython-312.pyc
Normal file
BIN
core/__pycache__/poster_gen.cpython-312.pyc
Normal file
Binary file not shown.
@ -21,7 +21,7 @@ if PROJECT_ROOT not in sys.path:
|
||||
# 导入所需的图像处理模块
|
||||
try:
|
||||
import core.simple_collage as simple_collage
|
||||
import core.posterGen as posterGen
|
||||
import core.poster_gen as poster_gen
|
||||
from utils.resource_loader import ResourceLoader
|
||||
except ImportError as e:
|
||||
logging.critical(f"导入模块失败: {e}")
|
||||
@ -188,7 +188,7 @@ def test_poster_generation(config, output_dir, collage_dir=None):
|
||||
|
||||
# 创建海报生成器
|
||||
try:
|
||||
poster_generator = posterGen.PosterGenerator(
|
||||
poster_generator = poster_gen.PosterGenerator(
|
||||
poster_save_dir=poster_output_dir,
|
||||
assets_base_dir=poster_assets_dir,
|
||||
poster_size=tuple(config.get("poster_target_size", [900, 1200]))
|
||||
|
||||
@ -21,7 +21,7 @@ project_root = str(Path(__file__).parent.parent.absolute())
|
||||
if project_root not in sys.path:
|
||||
sys.path.append(project_root)
|
||||
|
||||
from core import posterGen
|
||||
from core import poster_gen
|
||||
|
||||
def parse_arguments():
|
||||
"""解析命令行参数"""
|
||||
@ -126,7 +126,7 @@ def main():
|
||||
|
||||
# 初始化PosterGenerator
|
||||
print("初始化PosterGenerator...")
|
||||
poster_generator = posterGen.PosterGenerator(base_dir=base_dir)
|
||||
poster_generator = poster_gen.PosterGenerator(base_dir=base_dir)
|
||||
|
||||
# 确定底图路径
|
||||
image_path = args.image
|
||||
|
||||
2
main.py
2
main.py
@ -10,7 +10,7 @@ import logging
|
||||
from core.ai_agent import AI_Agent
|
||||
# from core.topic_parser import TopicParser # No longer needed directly in main?
|
||||
import core.contentGen as contentGen
|
||||
import core.posterGen as posterGen
|
||||
import core.poster_gen as poster_gen
|
||||
import core.simple_collage as simple_collage
|
||||
from utils.resource_loader import ResourceLoader
|
||||
from utils.tweet_generator import ( # Import the moved functions
|
||||
|
||||
45
poster_config.json
Normal file
45
poster_config.json
Normal file
@ -0,0 +1,45 @@
|
||||
{
|
||||
"variants": 1,
|
||||
"model": "qwenQWQ",
|
||||
"api_url": "http://localhost:8000/v1/",
|
||||
"api_key": "EMPTY",
|
||||
"poster_content_system_prompt": "./genPrompts/poster_content_systemPrompt.txt",
|
||||
"resource_dir": [
|
||||
{
|
||||
"type": "Object",
|
||||
"file_path": [
|
||||
"./resource/Object/中山温泉宾馆.txt",
|
||||
"./resource/Object/乌镇民宿.txt",
|
||||
"./resource/Object/从化客天下·禧悦庄.txt"
|
||||
]
|
||||
},
|
||||
{
|
||||
"type": "Description",
|
||||
"file_path": [
|
||||
"./resource/Object/中山温泉宾馆.txt",
|
||||
"./resource/Object/乌镇民宿.txt",
|
||||
"./resource/Object/从化客天下·禧悦庄.txt"
|
||||
]
|
||||
},
|
||||
{
|
||||
"type": "Product",
|
||||
"file_path": [
|
||||
]
|
||||
}
|
||||
],
|
||||
"output_dir": "./result",
|
||||
"image_base_dir": "/root/autodl-tmp/TravelContentCreator/hotel_img",
|
||||
"poster_assets_base_dir": "/root/autodl-tmp/poster_baseboard_0403",
|
||||
"request_timeout": 120,
|
||||
"max_retries": 3,
|
||||
"output_collage_subdir": "collage_img",
|
||||
"output_poster_subdir": "poster",
|
||||
"output_poster_filename": "poster.jpg",
|
||||
"poster_target_size": [
|
||||
900,
|
||||
1200
|
||||
],
|
||||
"text_possibility": 0.3,
|
||||
"img_frame_possibility": 0.7,
|
||||
"text_bg_possibility": 0
|
||||
}
|
||||
232
test_poster.py
Normal file
232
test_poster.py
Normal file
@ -0,0 +1,232 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
测试海报生成的脚本
|
||||
"""
|
||||
|
||||
import os
|
||||
import time
|
||||
import argparse
|
||||
import json
|
||||
import logging
|
||||
import sys
|
||||
import traceback
|
||||
from datetime import datetime
|
||||
|
||||
from core.ai_agent import AI_Agent
|
||||
from utils.tweet_generator import generate_posters_for_topic
|
||||
from utils.output_handler import FileSystemOutputHandler
|
||||
from core.topic_parser import TopicParser
|
||||
|
||||
def load_config(config_path="poster_config.json"):
|
||||
"""从JSON文件加载配置"""
|
||||
if not os.path.exists(config_path):
|
||||
print(f"错误：配置文件 '{config_path}' 未找到。")
|
||||
sys.exit(1)
|
||||
try:
|
||||
with open(config_path, 'r', encoding='utf-8') as f:
|
||||
config = json.load(f)
|
||||
|
||||
# 基本验证
|
||||
required_keys = ["api_url", "model", "api_key", "resource_dir", "output_dir",
|
||||
"image_base_dir", "poster_assets_base_dir",
|
||||
"poster_content_system_prompt"]
|
||||
|
||||
if not all(key in config for key in required_keys):
|
||||
missing_keys = [key for key in required_keys if key not in config]
|
||||
print(f"错误：配置文件 '{config_path}' 缺少必需的键：{missing_keys}")
|
||||
sys.exit(1)
|
||||
|
||||
return config
|
||||
except json.JSONDecodeError:
|
||||
print(f"错误：无法从 '{config_path}' 解码JSON。请检查文件格式。")
|
||||
sys.exit(1)
|
||||
except Exception as e:
|
||||
print(f"从 '{config_path}' 加载配置时出错：{e}")
|
||||
sys.exit(1)
|
||||
|
||||
def main():
|
||||
# 设置日志记录
|
||||
logging.basicConfig(
|
||||
level=logging.INFO,
|
||||
format='%(asctime)s - %(levelname)s - [%(filename)s:%(lineno)d] - %(message)s',
|
||||
datefmt='%Y-%m-%d %H:%M:%S'
|
||||
)
|
||||
|
||||
# 解析命令行参数
|
||||
parser = argparse.ArgumentParser(description="测试海报生成")
|
||||
parser.add_argument(
|
||||
"--config",
|
||||
type=str,
|
||||
default="poster_config.json",
|
||||
help="配置文件路径（例如，poster_config.json）"
|
||||
)
|
||||
parser.add_argument(
|
||||
"--topics_file",
|
||||
type=str,
|
||||
required=True,
|
||||
help="必需的选题JSON文件路径，用于获取海报生成的主题数据。"
|
||||
)
|
||||
parser.add_argument(
|
||||
"--topic_index",
|
||||
type=int,
|
||||
default=None,
|
||||
help="要生成海报的特定选题索引。如果未提供，将为所有选题生成海报。"
|
||||
)
|
||||
parser.add_argument(
|
||||
"--run_id",
|
||||
type=str,
|
||||
default=None,
|
||||
help="可选的指定运行ID。如果未提供，将生成基于时间戳的ID。"
|
||||
)
|
||||
parser.add_argument(
|
||||
"--debug",
|
||||
action='store_true',
|
||||
help="启用调试级别日志记录。"
|
||||
)
|
||||
args = parser.parse_args()
|
||||
|
||||
# 调整日志级别（如果启用了调试）
|
||||
if args.debug:
|
||||
logging.getLogger().setLevel(logging.DEBUG)
|
||||
logging.info("已启用调试日志记录。")
|
||||
|
||||
logging.info("启动海报生成测试脚本...")
|
||||
logging.info(f"使用配置文件：{args.config}")
|
||||
logging.info(f"使用选题文件：{args.topics_file}")
|
||||
if args.topic_index is not None:
|
||||
logging.info(f"将仅处理选题索引：{args.topic_index}")
|
||||
if args.run_id:
|
||||
logging.info(f"使用指定的run_id：{args.run_id}")
|
||||
|
||||
# 加载配置
|
||||
config = load_config(args.config)
|
||||
if config is None:
|
||||
logging.critical("无法加载配置。退出。")
|
||||
sys.exit(1)
|
||||
|
||||
# 初始化输出处理器
|
||||
output_handler = FileSystemOutputHandler(config.get("output_dir", "result"))
|
||||
logging.info(f"使用输出处理器：{output_handler.__class__.__name__}")
|
||||
|
||||
# 加载选题数据
|
||||
logging.info(f"从以下位置加载选题：{args.topics_file}")
|
||||
topics_list = TopicParser.load_topics_from_json(args.topics_file)
|
||||
if not topics_list:
|
||||
logging.error(f"无法从{args.topics_file}加载选题。无法继续。")
|
||||
sys.exit(1)
|
||||
|
||||
logging.info(f"成功加载{len(topics_list)}个选题。")
|
||||
|
||||
# 设置run_id
|
||||
run_id = args.run_id
|
||||
if run_id is None:
|
||||
# 尝试从文件名推断run_id
|
||||
try:
|
||||
base = os.path.basename(args.topics_file)
|
||||
if base.startswith("tweet_topic_") and base.endswith(".json"):
|
||||
run_id = base[len("tweet_topic_"): -len(".json")]
|
||||
logging.info(f"从选题文件名推断的run_id：{run_id}")
|
||||
else:
|
||||
run_id = datetime.now().strftime("%Y-%m-%d_%H-%M-%S_poster")
|
||||
logging.info(f"为海报生成的run_id：{run_id}")
|
||||
except Exception as e:
|
||||
run_id = datetime.now().strftime("%Y-%m-%d_%H-%M-%S_poster")
|
||||
logging.info(f"生成的run_id：{run_id}")
|
||||
|
||||
# 加载海报内容系统提示词
|
||||
poster_content_system_prompt_path = config.get("poster_content_system_prompt")
|
||||
if not os.path.exists(poster_content_system_prompt_path):
|
||||
logging.error(f"海报内容系统提示词文件不存在：{poster_content_system_prompt_path}")
|
||||
sys.exit(1)
|
||||
|
||||
with open(poster_content_system_prompt_path, "r", encoding="utf-8") as f:
|
||||
poster_content_system_prompt = f.read()
|
||||
|
||||
# 准备海报生成参数
|
||||
poster_variants = config.get("variants", 1)
|
||||
poster_assets_dir = config.get("poster_assets_base_dir")
|
||||
img_base_dir = config.get("image_base_dir")
|
||||
res_dir_config = config.get("resource_dir", [])
|
||||
poster_size = tuple(config.get("poster_target_size", [900, 1200]))
|
||||
txt_possibility = config.get("text_possibility", 0.3)
|
||||
img_frame_possibility = config.get("img_frame_possibility", 0.7)
|
||||
text_bg_possibility = config.get("text_bg_possibility", 0)
|
||||
collage_subdir = config.get("output_collage_subdir", "collage_img")
|
||||
poster_subdir = config.get("output_poster_subdir", "poster")
|
||||
poster_filename = config.get("output_poster_filename", "poster.jpg")
|
||||
|
||||
# 检查关键路径
|
||||
if not poster_assets_dir or not img_base_dir:
|
||||
logging.error("配置中缺少关键路径（poster_assets_base_dir或image_base_dir）。无法继续。")
|
||||
sys.exit(1)
|
||||
|
||||
# 开始海报生成
|
||||
pipeline_start_time = time.time()
|
||||
logging.info("开始执行海报生成...")
|
||||
|
||||
poster_success = False
|
||||
|
||||
# 如果指定了topic_index，只处理该选题
|
||||
if args.topic_index is not None:
|
||||
topics_to_process = []
|
||||
for topic in topics_list:
|
||||
if topic.get('index') == args.topic_index or (topic.get('index') is None and int(args.topic_index) == 1):
|
||||
topics_to_process.append(topic)
|
||||
break
|
||||
if not topics_to_process:
|
||||
logging.error(f"未找到索引为{args.topic_index}的选题。")
|
||||
sys.exit(1)
|
||||
else:
|
||||
topics_to_process = topics_list
|
||||
|
||||
# 逐个处理选题
|
||||
for i, topic_item in enumerate(topics_to_process):
|
||||
topic_index = topic_item.get('index', i + 1)
|
||||
logging.info(f"--- 处理选题 {topic_index}: {topic_item.get('object', 'N/A')} ---")
|
||||
|
||||
try:
|
||||
posters_attempted = generate_posters_for_topic(
|
||||
topic_item=topic_item,
|
||||
output_dir=config["output_dir"],
|
||||
run_id=run_id,
|
||||
topic_index=topic_index,
|
||||
output_handler=output_handler,
|
||||
variants=poster_variants,
|
||||
poster_assets_base_dir=poster_assets_dir,
|
||||
image_base_dir=img_base_dir,
|
||||
resource_dir_config=res_dir_config,
|
||||
poster_target_size=poster_size,
|
||||
text_possibility=txt_possibility,
|
||||
img_frame_possibility=img_frame_possibility,
|
||||
text_bg_possibility=text_bg_possibility,
|
||||
output_collage_subdir=collage_subdir,
|
||||
output_poster_subdir=poster_subdir,
|
||||
output_poster_filename=poster_filename,
|
||||
system_prompt=poster_content_system_prompt
|
||||
)
|
||||
|
||||
if posters_attempted:
|
||||
logging.info(f"选题{topic_index}的海报生成过程已完成。")
|
||||
poster_success = True
|
||||
else:
|
||||
logging.warning(f"选题{topic_index}的海报生成被跳过或在早期失败。")
|
||||
except Exception as e:
|
||||
logging.exception(f"处理选题{topic_index}的海报生成时出错:")
|
||||
|
||||
logging.info(f"--- 完成选题 {topic_index} ---")
|
||||
|
||||
# 最终化输出
|
||||
if run_id:
|
||||
output_handler.finalize(run_id)
|
||||
|
||||
pipeline_end_time = time.time()
|
||||
if poster_success:
|
||||
logging.info(f"海报生成完成，耗时{pipeline_end_time - pipeline_start_time:.2f}秒。")
|
||||
else:
|
||||
logging.warning("海报生成完成，但可能遇到错误或未生成输出。")
|
||||
|
||||
logging.info(f"运行ID'{run_id}'的结果位于：{os.path.join(config.get('output_dir', 'result'), run_id)}")
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
261
test_topic_content.py
Normal file
261
test_topic_content.py
Normal file
@ -0,0 +1,261 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
测试选题生成和文章生成的脚本
|
||||
"""
|
||||
|
||||
import os
|
||||
import time
|
||||
import argparse
|
||||
import json
|
||||
import logging
|
||||
import sys
|
||||
from datetime import datetime
|
||||
|
||||
from core.ai_agent import AI_Agent
|
||||
from utils.prompt_manager import PromptManager
|
||||
from utils.tweet_generator import run_topic_generation_pipeline, generate_content_for_topic
|
||||
from utils.output_handler import FileSystemOutputHandler
|
||||
|
||||
def load_config(config_path="topic_content_config.json"):
|
||||
"""从JSON文件加载配置"""
|
||||
if not os.path.exists(config_path):
|
||||
print(f"错误：配置文件 '{config_path}' 未找到。")
|
||||
sys.exit(1)
|
||||
try:
|
||||
with open(config_path, 'r', encoding='utf-8') as f:
|
||||
config = json.load(f)
|
||||
|
||||
# 基本验证
|
||||
required_keys = ["api_url", "model", "api_key", "resource_dir", "output_dir",
|
||||
"num", "variants", "topic_system_prompt", "topic_user_prompt",
|
||||
"content_system_prompt"]
|
||||
|
||||
if not all(key in config for key in required_keys):
|
||||
missing_keys = [key for key in required_keys if key not in config]
|
||||
print(f"错误：配置文件 '{config_path}' 缺少必需的键：{missing_keys}")
|
||||
sys.exit(1)
|
||||
|
||||
# 验证prompts_dir或prompts_config至少有一个存在
|
||||
if not ("prompts_dir" in config or "prompts_config" in config):
|
||||
print(f"错误：配置文件 '{config_path}' 必须包含 'prompts_dir' 或 'prompts_config'")
|
||||
sys.exit(1)
|
||||
|
||||
return config
|
||||
except json.JSONDecodeError:
|
||||
print(f"错误：无法从 '{config_path}' 解码JSON。请检查文件格式。")
|
||||
sys.exit(1)
|
||||
except Exception as e:
|
||||
print(f"从 '{config_path}' 加载配置时出错：{e}")
|
||||
sys.exit(1)
|
||||
|
||||
def main():
|
||||
# 设置日志记录
|
||||
logging.basicConfig(
|
||||
level=logging.INFO,
|
||||
format='%(asctime)s - %(levelname)s - [%(filename)s:%(lineno)d] - %(message)s',
|
||||
datefmt='%Y-%m-%d %H:%M:%S'
|
||||
)
|
||||
|
||||
# 解析命令行参数
|
||||
parser = argparse.ArgumentParser(description="测试选题和文章生成")
|
||||
parser.add_argument(
|
||||
"--config",
|
||||
type=str,
|
||||
default="topic_content_config.json",
|
||||
help="配置文件路径（例如，topic_content_config.json）"
|
||||
)
|
||||
parser.add_argument(
|
||||
"--run_id",
|
||||
type=str,
|
||||
default=None,
|
||||
help="可选的指定运行ID（例如，'test_run_01'）。如果未提供，将生成基于时间戳的ID。"
|
||||
)
|
||||
parser.add_argument(
|
||||
"--topics_file",
|
||||
type=str,
|
||||
default=None,
|
||||
help="可选的预生成选题JSON文件路径。如果提供，则跳过选题生成。"
|
||||
)
|
||||
parser.add_argument(
|
||||
"--debug",
|
||||
action='store_true',
|
||||
help="启用调试级别日志记录。"
|
||||
)
|
||||
args = parser.parse_args()
|
||||
|
||||
# 调整日志级别（如果启用了调试）
|
||||
if args.debug:
|
||||
logging.getLogger().setLevel(logging.DEBUG)
|
||||
logging.info("已启用调试日志记录。")
|
||||
|
||||
logging.info("启动选题和文章生成测试脚本...")
|
||||
logging.info(f"使用配置文件：{args.config}")
|
||||
if args.run_id:
|
||||
logging.info(f"使用指定的run_id：{args.run_id}")
|
||||
if args.topics_file:
|
||||
logging.info(f"使用现有选题文件：{args.topics_file}")
|
||||
|
||||
# 加载配置
|
||||
config = load_config(args.config)
|
||||
if config is None:
|
||||
logging.critical("无法加载配置。退出。")
|
||||
sys.exit(1)
|
||||
|
||||
# 初始化输出处理器
|
||||
output_handler = FileSystemOutputHandler(config.get("output_dir", "result"))
|
||||
logging.info(f"使用输出处理器：{output_handler.__class__.__name__}")
|
||||
|
||||
run_id = args.run_id
|
||||
topics_list = None
|
||||
system_prompt = None
|
||||
user_prompt = None
|
||||
pipeline_start_time = time.time()
|
||||
|
||||
# 步骤1：选题生成（或加载现有选题）
|
||||
if args.topics_file:
|
||||
from core.topic_parser import TopicParser
|
||||
logging.info(f"跳过选题生成（步骤1）- 从以下位置加载选题：{args.topics_file}")
|
||||
topics_list = TopicParser.load_topics_from_json(args.topics_file)
|
||||
if topics_list:
|
||||
# 如果未提供run_id，尝试从文件名推断
|
||||
if not run_id:
|
||||
try:
|
||||
base = os.path.basename(args.topics_file)
|
||||
# 假设格式为"tweet_topic_{run_id}.json"或"tweet_topic.json"
|
||||
if base.startswith("tweet_topic_") and base.endswith(".json"):
|
||||
run_id = base[len("tweet_topic_"): -len(".json")]
|
||||
logging.info(f"从选题文件名推断的run_id：{run_id}")
|
||||
elif base == "tweet_topic.json":
|
||||
logging.warning(f"从默认文件名'{base}'加载选题。未推断run_id。")
|
||||
else:
|
||||
logging.warning(f"无法从选题文件名推断run_id：{base}")
|
||||
except Exception as e:
|
||||
logging.warning(f"尝试从选题文件名推断run_id时出错：{e}")
|
||||
|
||||
# 如果尝试推断后run_id仍为None，则生成一个
|
||||
if run_id is None:
|
||||
run_id = datetime.now().strftime("%Y-%m-%d_%H-%M-%S_loaded")
|
||||
logging.info(f"为加载的选题生成的run_id：{run_id}")
|
||||
|
||||
# 加载文件时缺少提示词
|
||||
system_prompt = ""
|
||||
user_prompt = ""
|
||||
logging.info(f"成功加载{len(topics_list)}个选题，run_id：{run_id}。提示词不可用。")
|
||||
else:
|
||||
logging.error(f"无法从{args.topics_file}加载选题。无法继续。")
|
||||
sys.exit(1)
|
||||
else:
|
||||
logging.info("执行选题生成（步骤1）...")
|
||||
step1_start = time.time()
|
||||
# 调用更新后的函数，接收原始数据
|
||||
run_id, topics_list, system_prompt, user_prompt = run_topic_generation_pipeline(config, args.run_id)
|
||||
step1_end = time.time()
|
||||
if run_id is not None and topics_list is not None:
|
||||
logging.info(f"步骤1成功完成，耗时{step1_end - step1_start:.2f}秒。运行ID：{run_id}")
|
||||
# 使用输出处理器保存结果
|
||||
output_handler.handle_topic_results(run_id, topics_list, system_prompt, user_prompt)
|
||||
else:
|
||||
logging.critical("选题生成（步骤1）失败。退出。")
|
||||
sys.exit(1)
|
||||
|
||||
# 步骤2：内容生成
|
||||
if run_id is not None and topics_list is not None:
|
||||
logging.info("执行内容生成（步骤2）...")
|
||||
step2_start = time.time()
|
||||
|
||||
# 创建PromptManager实例
|
||||
try:
|
||||
prompt_manager = PromptManager(
|
||||
topic_system_prompt_path=config.get("topic_system_prompt"),
|
||||
topic_user_prompt_path=config.get("topic_user_prompt"),
|
||||
content_system_prompt_path=config.get("content_system_prompt"),
|
||||
prompts_config=config.get("prompts_config"),
|
||||
prompts_dir=config.get("prompts_dir"),
|
||||
resource_dir_config=config.get("resource_dir", []),
|
||||
topic_gen_num=config.get("num", 1),
|
||||
topic_gen_date=config.get("date", "")
|
||||
)
|
||||
logging.info("已为步骤2创建PromptManager实例。")
|
||||
except KeyError as e:
|
||||
logging.error(f"创建PromptManager时配置错误：缺少键'{e}'。无法继续内容生成。")
|
||||
sys.exit(1)
|
||||
|
||||
# 初始化AI Agent
|
||||
ai_agent = None
|
||||
content_success = False
|
||||
try:
|
||||
request_timeout = config.get("request_timeout", 30)
|
||||
max_retries = config.get("max_retries", 3)
|
||||
stream_chunk_timeout = config.get("stream_chunk_timeout", 60)
|
||||
ai_agent = AI_Agent(
|
||||
config["api_url"],
|
||||
config["model"],
|
||||
config["api_key"],
|
||||
timeout=request_timeout,
|
||||
max_retries=max_retries,
|
||||
stream_chunk_timeout=stream_chunk_timeout
|
||||
)
|
||||
logging.info("已初始化用于内容生成的AI Agent。")
|
||||
|
||||
# 遍历选题
|
||||
for i, topic_item in enumerate(topics_list):
|
||||
topic_index = topic_item.get('index', i + 1)
|
||||
logging.info(f"--- 处理选题 {topic_index}/{len(topics_list)}: {topic_item.get('object', 'N/A')} ---")
|
||||
|
||||
# 读取内容生成所需的参数
|
||||
content_variants = config.get("variants", 1)
|
||||
content_temp = config.get("content_temperature", 0.3)
|
||||
content_top_p = config.get("content_top_p", 0.4)
|
||||
content_presence_penalty = config.get("content_presence_penalty", 1.5)
|
||||
|
||||
# 调用generate_content_for_topic
|
||||
topic_success = generate_content_for_topic(
|
||||
ai_agent,
|
||||
prompt_manager,
|
||||
topic_item,
|
||||
run_id,
|
||||
topic_index,
|
||||
output_handler,
|
||||
variants=content_variants,
|
||||
temperature=content_temp,
|
||||
top_p=content_top_p,
|
||||
presence_penalty=content_presence_penalty
|
||||
)
|
||||
|
||||
if topic_success:
|
||||
logging.info(f"选题{topic_index}的内容生成成功。")
|
||||
content_success = True
|
||||
else:
|
||||
logging.warning(f"选题{topic_index}的内容生成失败或未产生有效结果。")
|
||||
logging.info(f"--- 完成选题 {topic_index} ---")
|
||||
|
||||
except KeyError as e:
|
||||
logging.error(f"内容生成过程中的配置错误：缺少键'{e}'")
|
||||
traceback.print_exc()
|
||||
except Exception as e:
|
||||
logging.exception("内容生成过程中发生意外错误:")
|
||||
finally:
|
||||
# 确保AI agent已关闭
|
||||
if ai_agent:
|
||||
logging.info("关闭内容生成AI Agent...")
|
||||
ai_agent.close()
|
||||
|
||||
step2_end = time.time()
|
||||
if content_success:
|
||||
logging.info(f"步骤2完成，耗时{step2_end - step2_start:.2f}秒。")
|
||||
else:
|
||||
logging.warning("步骤2完成，但可能遇到错误或未生成输出。")
|
||||
else:
|
||||
logging.error("无法进行步骤2：步骤1的run_id或topics_list无效。")
|
||||
|
||||
# 最终化输出
|
||||
if run_id:
|
||||
output_handler.finalize(run_id)
|
||||
|
||||
pipeline_end_time = time.time()
|
||||
logging.info(f"流程完成。总执行时间：{pipeline_end_time - pipeline_start_time:.2f}秒。")
|
||||
logging.info(f"运行ID'{run_id}'的结果位于：{os.path.join(config.get('output_dir', 'result'), run_id)}")
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
72
topic_content_config.json
Normal file
72
topic_content_config.json
Normal file
@ -0,0 +1,72 @@
|
||||
{
|
||||
"date": "5月15日, 5月16日, 5月17日, 6月1日",
|
||||
"num": 2,
|
||||
"variants": 1,
|
||||
"topic_temperature": 0.2,
|
||||
"topic_top_p": 0.3,
|
||||
"topic_presence_penalty": 1.5,
|
||||
"content_temperature": 0.3,
|
||||
"content_top_p": 0.4,
|
||||
"content_presence_penalty": 1.5,
|
||||
"model": "qwenQWQ",
|
||||
"api_url": "http://localhost:8000/v1/",
|
||||
"api_key": "EMPTY",
|
||||
"topic_system_prompt": "./SelectPrompt/systemPrompt.txt",
|
||||
"topic_user_prompt": "./SelectPrompt/userPrompt.txt",
|
||||
"content_system_prompt": "./genPrompts/systemPrompt.txt",
|
||||
"prompts_config": [
|
||||
{
|
||||
"type": "Style",
|
||||
"file_path": [
|
||||
"./genPrompts/Style/攻略风文案提示词.txt",
|
||||
"./genPrompts/Style/轻奢风文案提示词.txt",
|
||||
"./genPrompts/Style/极力推荐风文案提示词.txt",
|
||||
"./genPrompts/Style/美食风文案提示词.txt"
|
||||
]
|
||||
},
|
||||
{
|
||||
"type": "Demand",
|
||||
"file_path": [
|
||||
"./genPrompts/Demand/学生党文旅需求.txt",
|
||||
"./genPrompts/Demand/情侣向文旅需求.txt",
|
||||
"./genPrompts/Demand/职场人文旅需求.txt",
|
||||
"./genPrompts/Demand/亲子向文旅需求.txt",
|
||||
"./genPrompts/Demand/周边游文旅需求.txt",
|
||||
"./genPrompts/Demand/夕阳红文旅需求.txt"
|
||||
]
|
||||
},
|
||||
{
|
||||
"type": "Refer",
|
||||
"file_path": [
|
||||
"./genPrompts/Refer/标题参考格式.txt"
|
||||
]
|
||||
}
|
||||
],
|
||||
"resource_dir": [
|
||||
{
|
||||
"type": "Object",
|
||||
"file_path": [
|
||||
"./resource/Object/中山温泉宾馆.txt",
|
||||
"./resource/Object/乌镇民宿.txt",
|
||||
"./resource/Object/从化客天下·禧悦庄.txt"
|
||||
]
|
||||
},
|
||||
{
|
||||
"type": "Description",
|
||||
"file_path": [
|
||||
"./resource/Object/中山温泉宾馆.txt",
|
||||
"./resource/Object/乌镇民宿.txt",
|
||||
"./resource/Object/从化客天下·禧悦庄.txt"
|
||||
]
|
||||
},
|
||||
{
|
||||
"type": "Product",
|
||||
"file_path": [
|
||||
]
|
||||
}
|
||||
],
|
||||
"output_dir": "./result",
|
||||
"request_timeout": 120,
|
||||
"max_retries": 3,
|
||||
"stream_chunk_timeout": 60
|
||||
}
|
||||
Binary file not shown.
@ -24,7 +24,7 @@ from core.ai_agent import AI_Agent
|
||||
from core.topic_parser import TopicParser
|
||||
from utils.prompt_manager import PromptManager # Keep this as it's importing from the same level package 'utils'
|
||||
from core import contentGen as core_contentGen
|
||||
from core import posterGen as core_posterGen
|
||||
from core import poster_gen as core_posterGen
|
||||
from core import simple_collage as core_simple_collage
|
||||
from .output_handler import OutputHandler # <-- 添加导入
|
||||
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user