feat: initial commit of AI Comic Generator
This commit is contained in:
p
+22
@@ -0,0 +1,22 @@
|
||||
# Python
|
||||
__pycache__/
|
||||
*.py[cod]
|
||||
*$py.class
|
||||
.venv/
|
||||
venv/
|
||||
*.db
|
||||
*.sqlite3
|
||||
|
||||
# Node
|
||||
node_modules/
|
||||
dist/
|
||||
npm-debug.log*
|
||||
yarn-debug.log*
|
||||
yarn-error.log*
|
||||
pnpm-debug.log*
|
||||
|
||||
# Project specific
|
||||
comic_gen/
|
||||
backend/comic_app.db
|
||||
backend/*.db
|
||||
backend/static/*.png
|
||||
@@ -0,0 +1,83 @@
|
||||
# Full-Stack AI Comic Generator Implementation Plan
|
||||
|
||||
This plan details the creation of a web-based AI comic generator using Vue 3 (Frontend) and FastAPI (Backend), separating concerns into a modular architecture.
|
||||
|
||||
## 1. Project Structure Setup
|
||||
We will create a root directory containing two main folders: `backend` and `frontend`.
|
||||
|
||||
### Backend Structure (`backend/`)
|
||||
- **Framework**: FastAPI
|
||||
- **Database**: SQLite (via SQLModel/SQLAlchemy) for storing project data and configs.
|
||||
- **Directory Layout**:
|
||||
- `app/`
|
||||
- `core/`: Configuration (env vars, DB settings).
|
||||
- `models/`: Database models (SQLModel).
|
||||
- `schemas/`: Pydantic models for request/response validation.
|
||||
- `cruds/`: Database CRUD operations.
|
||||
- `routers/`: API endpoints grouped by functionality.
|
||||
- `services/`: Business logic (AI generation, File management).
|
||||
- `utils/`: Helper functions.
|
||||
- `static/`: Serving generated images.
|
||||
- `main.py`: Application entry point.
|
||||
|
||||
### Frontend Structure (`frontend/`)
|
||||
- **Framework**: Vue 3 + Vite
|
||||
- **UI Library**: Element Plus (Dark Mode enabled for "Tech" style).
|
||||
- **Directory Layout**:
|
||||
- `src/`
|
||||
- `api/`: Axios instances for backend communication.
|
||||
- `components/`: Reusable UI components (JSON Editor, Image Cards).
|
||||
- `views/`: Main pages (Config, Workspace).
|
||||
- `stores/`: Pinia state management.
|
||||
|
||||
## 2. Backend Implementation Steps
|
||||
|
||||
### Phase 1: Core & Configuration
|
||||
1. **Environment**: Setup `requirements.txt` (FastAPI, SQLModel, Uvicorn, Google GenAI, OpenAI, python-dotenv).
|
||||
2. **Models & Schemas**:
|
||||
- `ModelConfig`: Store API keys, provider (Google/OpenAI/DeepSeek), model names.
|
||||
- `Project`: Store comic project metadata (title, status).
|
||||
- `ComicData`: Store the generated JSONs (Global Config, Characters, Storyboard).
|
||||
3. **CRUDs**: Implement basic Create/Read/Update/Delete operations for Configs and Projects.
|
||||
|
||||
### Phase 2: AI Services Integration
|
||||
1. **AI Provider Adapter**: Create a unified interface to handle different providers (Google, DeepSeek, ChatGPT, etc.).
|
||||
2. **Migration**: Refactor logic from `comic_generator.py` into `services/comic_service.py`.
|
||||
- Implement `generate_storyboard` (Text generation).
|
||||
- Implement `generate_character_image` (Image generation).
|
||||
- Implement `generate_comic_panel` (Image generation with context).
|
||||
3. **Endpoints**:
|
||||
- `POST /api/generate/json`: Generate initial JSONs from user input.
|
||||
- `POST /api/generate/image`: Generate specific image (Character or Panel).
|
||||
- `POST /api/project/{id}/export`: Package and zip output.
|
||||
|
||||
## 3. Frontend Implementation Steps
|
||||
|
||||
### Phase 1: UI Framework & Configuration
|
||||
1. **Setup**: Initialize Vue 3 project, install Element Plus, Axios, Pinia, Vue Router.
|
||||
2. **Theme**: Configure Element Plus for Dark Mode/Tech style.
|
||||
3. **Model Configuration Page**:
|
||||
- Form to add/edit API keys and select models for Text and Image generation.
|
||||
|
||||
### Phase 2: Comic Workflow Page
|
||||
1. **Step 1: Concept & JSON**:
|
||||
- Input field for story idea.
|
||||
- "Generate" button.
|
||||
- **JSON Editor**: Integrated code editor (e.g., Monaco Editor) to modify generated JSONs (Global Config, Characters, Storyboard).
|
||||
2. **Step 2: Character Studio**:
|
||||
- Display list of characters from JSON.
|
||||
- "Generate/Regenerate" button for each character.
|
||||
- Support "Add Character" manually.
|
||||
3. **Step 3: Comic Board**:
|
||||
- Display storyboard panels.
|
||||
- "Generate/Regenerate" button for each panel (4-grid or single).
|
||||
- Support modifying prompt per panel.
|
||||
4. **Step 4: Export**:
|
||||
- Button to download the complete comic package.
|
||||
|
||||
## 4. Execution Strategy
|
||||
1. **Backend First**: I will build the FastAPI backend, ensuring the API is functional and can replicate the existing script's logic.
|
||||
2. **Frontend Second**: I will build the Vue frontend and connect it to the backend.
|
||||
3. **Verification**: I will test the full flow: Config -> Story Input -> Edit JSON -> Generate Images -> Export.
|
||||
|
||||
I will begin by setting up the backend structure and dependencies.
|
||||
@@ -0,0 +1,148 @@
|
||||
# AI Comic Generator
|
||||
|
||||
[English](./README.md) | [中文](./README_CN.md)
|
||||
|
||||
An open-source AI-powered manga creation tool that transforms text stories into fully illustrated comics using Google Gemini models. Features include automatic storyboard generation, character consistency checks, and a visual editor.
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
## ✨ Core Features
|
||||
|
||||
* **Project Management**: Supports multi-project management, with each project independently saving story, characters, and storyboard data.
|
||||
* **Intelligent Scriptwriting**: Input simple story ideas, and AI automatically expands the plot and breaks it down into professional comic storyboard scripts (JSON format).
|
||||
* **Character Workshop**:
|
||||
* Automatically extracts characters from the story and generates detailed character settings (three views).
|
||||
* **Character Consistency**: Automatically references character setting images as references (Image-to-Image) when drawing.
|
||||
* **Merge & Deduplication**: Supports manual merging of duplicate generated characters (e.g., "Butler Ma" and "Old Ma").
|
||||
* **Storyboard Editor**:
|
||||
* Visually edit prompts, characters, and actions for each panel.
|
||||
* Supports single-panel redrawing and batch generation.
|
||||
* **Context Awareness**: Automatically reads previous storyboard panels and character images when generating storyboards to maintain style and plot continuity.
|
||||
* **Style Control**:
|
||||
* Global style configuration (e.g., "Cyberpunk", "Ink Style"), forcing AI to follow settings.
|
||||
* Supports custom dialog box and border styles.
|
||||
* **Background Tasks**: Time-consuming batch drawing tasks run in the background, supporting real-time progress viewing.
|
||||
|
||||
## 🛠️ Tech Stack
|
||||
|
||||
### Backend
|
||||
* **Framework**: FastAPI
|
||||
* **Database**: SQLite + SQLModel
|
||||
* **AI Service**: Google Gemini (Currently only supports Google's latest models)
|
||||
* **Text Model**: `gemini-3-flash-preview`
|
||||
* **Image Model**: `gemini-3-pro-image-preview`
|
||||
* **Task Queue**: FastAPI BackgroundTasks
|
||||
|
||||
### Frontend
|
||||
* **Framework**: Vue 3 + Vite
|
||||
* **UI Library**: Element Plus
|
||||
* **State Management**: Pinia
|
||||
* **HTTP Client**: Axios
|
||||
* **Package Manager**: pnpm
|
||||
|
||||
## 🚀 Quick Start
|
||||
|
||||
### Prerequisites
|
||||
* Python 3.9+
|
||||
* Node.js 16+
|
||||
* pnpm
|
||||
* Google Cloud API Key (Requires Gemini API access)
|
||||
|
||||
### 1. Backend Configuration
|
||||
|
||||
Enter the `backend` directory:
|
||||
|
||||
```bash
|
||||
cd backend
|
||||
```
|
||||
|
||||
Create virtual environment and install dependencies:
|
||||
|
||||
```bash
|
||||
python -m venv .venv
|
||||
# Windows
|
||||
.\.venv\Scripts\activate
|
||||
# Linux/Mac
|
||||
source .venv/bin/activate
|
||||
|
||||
pip install -r requirements.txt
|
||||
```
|
||||
|
||||
Create `.env` file and fill in API Key:
|
||||
|
||||
```ini
|
||||
# backend/.env
|
||||
GOOGLE_API_KEY="your_google_api_key_here"
|
||||
```
|
||||
|
||||
Start backend service:
|
||||
|
||||
```bash
|
||||
# Windows (using provided script)
|
||||
..\start_backend.bat
|
||||
|
||||
# Or run manually
|
||||
uvicorn app.main:app --reload --host 0.0.0.0 --port 8000
|
||||
```
|
||||
|
||||
### 2. Frontend Configuration
|
||||
|
||||
Enter the `frontend` directory:
|
||||
|
||||
```bash
|
||||
cd frontend
|
||||
```
|
||||
|
||||
Install dependencies:
|
||||
|
||||
```bash
|
||||
pnpm install
|
||||
```
|
||||
|
||||
Start frontend service:
|
||||
|
||||
```bash
|
||||
# Windows (using provided script)
|
||||
..\start_frontend.bat
|
||||
|
||||
# Or run manually
|
||||
pnpm dev
|
||||
```
|
||||
|
||||
Access in browser: `http://localhost:5173`
|
||||
|
||||
## 📖 User Guide
|
||||
|
||||
1. **Create Project**: Click "New Project" on the homepage and enter the comic title and introduction.
|
||||
2. **Story & Configuration**:
|
||||
* Enter your story outline.
|
||||
* Set global styles (e.g., "Japanese Shonen"), aspect ratio, etc.
|
||||
* Click "Generate Storyboard Config", and AI will generate the character list and storyboard script.
|
||||
3. **Character Workshop**:
|
||||
* View AI-generated character settings.
|
||||
* Click "Draw" to generate character portraits.
|
||||
* If there are duplicate characters, use the "Merge Characters" function to clean them up.
|
||||
4. **Storyboard Editing**:
|
||||
* Check the description of each panel in the storyboard list.
|
||||
* Click "Generate Image" or "Generate All" to start drawing the comic.
|
||||
* Click on an image to view it in large size and support downloading.
|
||||
|
||||
## 📁 Directory Structure
|
||||
|
||||
```
|
||||
aImanhua/
|
||||
├── backend/ # FastAPI Backend
|
||||
│ ├── app/ # Application Code
|
||||
│ ├── static/ # Generated images and temp files storage
|
||||
│ └── ...
|
||||
├── frontend/ # Vue3 Frontend
|
||||
│ ├── src/ # Pages and Components
|
||||
│ └── ...
|
||||
└── ...
|
||||
```
|
||||
|
||||
## 📝 License
|
||||
|
||||
MIT License
|
||||
+148
@@ -0,0 +1,148 @@
|
||||
# AI Comic Generator
|
||||
|
||||
[English](./README.md) | [中文](./README_CN.md)
|
||||
|
||||
一个基于 AI 的全流程漫画创作辅助工具，集成了故事生成、分镜拆解、角色设定、一致性控制和批量绘图功能。
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
## ✨ 核心功能
|
||||
|
||||
* **项目管理**: 支持多项目管理，每个项目独立保存故事、角色和分镜数据。
|
||||
* **智能编剧**: 输入简单的故事点子，AI 自动扩充情节并拆分为专业的漫画分镜脚本 (JSON 格式)。
|
||||
* **角色工坊**:
|
||||
* 自动从故事中提取角色并生成详细的人物设定（三视图）。
|
||||
* **角色一致性**: 绘图时自动引用角色设定图作为参考（Image-to-Image）。
|
||||
* **合并与去重**: 支持手动合并重复生成的角色（如“马管家”和“马老”）。
|
||||
* **分镜编辑器**:
|
||||
* 可视化编辑每一格分镜的提示词、角色和动作。
|
||||
* 支持单格重绘、批量生成。
|
||||
* **上下文感知**: 生成分镜时自动读取前序分镜和角色图，保持画风和剧情连贯性。
|
||||
* **风格控制**:
|
||||
* 全局风格配置（如“赛博朋克”、“水墨风”），强制 AI 遵循设定。
|
||||
* 支持自定义对话框、边框样式。
|
||||
* **后台任务**: 耗时的批量生图任务在后台运行，支持进度实时查看。
|
||||
|
||||
## 🛠️ 技术栈
|
||||
|
||||
### Backend (后端)
|
||||
* **Framework**: FastAPI
|
||||
* **Database**: SQLite + SQLModel
|
||||
* **AI Service**: Google Gemini (目前仅支持 Google 最新模型)
|
||||
* **文本模型**: `gemini-3-flash-preview`
|
||||
* **图像模型**: `gemini-3-pro-image-preview`
|
||||
* **Task Queue**: FastAPI BackgroundTasks
|
||||
|
||||
### Frontend (前端)
|
||||
* **Framework**: Vue 3 + Vite
|
||||
* **UI Library**: Element Plus
|
||||
* **State Management**: Pinia
|
||||
* **HTTP Client**: Axios
|
||||
* **Package Manager**: pnpm
|
||||
|
||||
## 🚀 快速开始
|
||||
|
||||
### 前置要求
|
||||
* Python 3.9+
|
||||
* Node.js 16+
|
||||
* pnpm
|
||||
* Google Cloud API Key (需开通 Gemini API 权限)
|
||||
|
||||
### 1. 后端配置
|
||||
|
||||
进入 `backend` 目录：
|
||||
|
||||
```bash
|
||||
cd backend
|
||||
```
|
||||
|
||||
创建虚拟环境并安装依赖：
|
||||
|
||||
```bash
|
||||
python -m venv .venv
|
||||
# Windows
|
||||
.\.venv\Scripts\activate
|
||||
# Linux/Mac
|
||||
source .venv/bin/activate
|
||||
|
||||
pip install -r requirements.txt
|
||||
```
|
||||
|
||||
创建 `.env` 文件并填入 API Key：
|
||||
|
||||
```ini
|
||||
# backend/.env
|
||||
GOOGLE_API_KEY="your_google_api_key_here"
|
||||
```
|
||||
|
||||
启动后端服务：
|
||||
|
||||
```bash
|
||||
# Windows (使用提供的脚本)
|
||||
..\start_backend.bat
|
||||
|
||||
# 或者手动运行
|
||||
uvicorn app.main:app --reload --host 0.0.0.0 --port 8000
|
||||
```
|
||||
|
||||
### 2. 前端配置
|
||||
|
||||
进入 `frontend` 目录：
|
||||
|
||||
```bash
|
||||
cd frontend
|
||||
```
|
||||
|
||||
安装依赖：
|
||||
|
||||
```bash
|
||||
pnpm install
|
||||
```
|
||||
|
||||
启动前端服务：
|
||||
|
||||
```bash
|
||||
# Windows (使用提供的脚本)
|
||||
..\start_frontend.bat
|
||||
|
||||
# 或者手动运行
|
||||
pnpm dev
|
||||
```
|
||||
|
||||
访问浏览器：`http://localhost:5173`
|
||||
|
||||
## 📖 使用指南
|
||||
|
||||
1. **创建项目**: 在首页点击“新建项目”，输入漫画标题和简介。
|
||||
2. **故事与配置**:
|
||||
* 输入你的故事大纲。
|
||||
* 设置全局风格（如“日系少年漫”）、画幅比例等。
|
||||
* 点击“生成分镜配置”，AI 将生成角色表和分镜脚本。
|
||||
3. **角色工坊**:
|
||||
* 查看 AI 生成的角色设定。
|
||||
* 点击“绘制”生成角色立绘。
|
||||
* 如有重复角色，使用“合并角色”功能进行清理。
|
||||
4. **分镜编辑**:
|
||||
* 在分镜列表中检查每一格的描述。
|
||||
* 点击“生成图片”或“一键生成所有”开始绘制漫画。
|
||||
* 点击图片可查看大图，支持下载。
|
||||
|
||||
## 📁 目录结构
|
||||
|
||||
```
|
||||
aImanhua/
|
||||
├── backend/ # FastAPI 后端
|
||||
│ ├── app/ # 应用代码
|
||||
│ ├── static/ # 生成的图片和临时文件存储
|
||||
│ └── ...
|
||||
├── frontend/ # Vue3 Frontend
|
||||
│ ├── src/ # 页面与组件
|
||||
│ └── ...
|
||||
└── ...
|
||||
```
|
||||
|
||||
## 📝 License
|
||||
|
||||
MIT License
|
||||
@@ -0,0 +1 @@
|
||||
DATABASE_URL=sqlite:///./comic_app.db
|
||||
@@ -0,0 +1,149 @@
|
||||
# A generic, single database configuration.
|
||||
|
||||
[alembic]
|
||||
# path to migration scripts.
|
||||
# this is typically a path given in POSIX (e.g. forward slashes)
|
||||
# format, relative to the token %(here)s which refers to the location of this
|
||||
# ini file
|
||||
script_location = %(here)s/alembic
|
||||
|
||||
# template used to generate migration file names; The default value is %%(rev)s_%%(slug)s
|
||||
# Uncomment the line below if you want the files to be prepended with date and time
|
||||
# see https://alembic.sqlalchemy.org/en/latest/tutorial.html#editing-the-ini-file
|
||||
# for all available tokens
|
||||
# file_template = %%(year)d_%%(month).2d_%%(day).2d_%%(hour).2d%%(minute).2d-%%(rev)s_%%(slug)s
|
||||
# Or organize into date-based subdirectories (requires recursive_version_locations = true)
|
||||
# file_template = %%(year)d/%%(month).2d/%%(day).2d_%%(hour).2d%%(minute).2d_%%(second).2d_%%(rev)s_%%(slug)s
|
||||
|
||||
# sys.path path, will be prepended to sys.path if present.
|
||||
# defaults to the current working directory. for multiple paths, the path separator
|
||||
# is defined by "path_separator" below.
|
||||
prepend_sys_path = .
|
||||
|
||||
|
||||
# timezone to use when rendering the date within the migration file
|
||||
# as well as the filename.
|
||||
# If specified, requires the tzdata library which can be installed by adding
|
||||
# `alembic[tz]` to the pip requirements.
|
||||
# string value is passed to ZoneInfo()
|
||||
# leave blank for localtime
|
||||
# timezone =
|
||||
|
||||
# max length of characters to apply to the "slug" field
|
||||
# truncate_slug_length = 40
|
||||
|
||||
# set to 'true' to run the environment during
|
||||
# the 'revision' command, regardless of autogenerate
|
||||
# revision_environment = false
|
||||
|
||||
# set to 'true' to allow .pyc and .pyo files without
|
||||
# a source .py file to be detected as revisions in the
|
||||
# versions/ directory
|
||||
# sourceless = false
|
||||
|
||||
# version location specification; This defaults
|
||||
# to <script_location>/versions. When using multiple version
|
||||
# directories, initial revisions must be specified with --version-path.
|
||||
# The path separator used here should be the separator specified by "path_separator"
|
||||
# below.
|
||||
# version_locations = %(here)s/bar:%(here)s/bat:%(here)s/alembic/versions
|
||||
|
||||
# path_separator; This indicates what character is used to split lists of file
|
||||
# paths, including version_locations and prepend_sys_path within configparser
|
||||
# files such as alembic.ini.
|
||||
# The default rendered in new alembic.ini files is "os", which uses os.pathsep
|
||||
# to provide os-dependent path splitting.
|
||||
#
|
||||
# Note that in order to support legacy alembic.ini files, this default does NOT
|
||||
# take place if path_separator is not present in alembic.ini. If this
|
||||
# option is omitted entirely, fallback logic is as follows:
|
||||
#
|
||||
# 1. Parsing of the version_locations option falls back to using the legacy
|
||||
# "version_path_separator" key, which if absent then falls back to the legacy
|
||||
# behavior of splitting on spaces and/or commas.
|
||||
# 2. Parsing of the prepend_sys_path option falls back to the legacy
|
||||
# behavior of splitting on spaces, commas, or colons.
|
||||
#
|
||||
# Valid values for path_separator are:
|
||||
#
|
||||
# path_separator = :
|
||||
# path_separator = ;
|
||||
# path_separator = space
|
||||
# path_separator = newline
|
||||
#
|
||||
# Use os.pathsep. Default configuration used for new projects.
|
||||
path_separator = os
|
||||
|
||||
# set to 'true' to search source files recursively
|
||||
# in each "version_locations" directory
|
||||
# new in Alembic version 1.10
|
||||
# recursive_version_locations = false
|
||||
|
||||
# the output encoding used when revision files
|
||||
# are written from script.py.mako
|
||||
# output_encoding = utf-8
|
||||
|
||||
# database URL. This is consumed by the user-maintained env.py script only.
|
||||
# other means of configuring database URLs may be customized within the env.py
|
||||
# file.
|
||||
sqlalchemy.url = driver://user:pass@localhost/dbname
|
||||
|
||||
|
||||
[post_write_hooks]
|
||||
# post_write_hooks defines scripts or Python functions that are run
|
||||
# on newly generated revision scripts. See the documentation for further
|
||||
# detail and examples
|
||||
|
||||
# format using "black" - use the console_scripts runner, against the "black" entrypoint
|
||||
# hooks = black
|
||||
# black.type = console_scripts
|
||||
# black.entrypoint = black
|
||||
# black.options = -l 79 REVISION_SCRIPT_FILENAME
|
||||
|
||||
# lint with attempts to fix using "ruff" - use the module runner, against the "ruff" module
|
||||
# hooks = ruff
|
||||
# ruff.type = module
|
||||
# ruff.module = ruff
|
||||
# ruff.options = check --fix REVISION_SCRIPT_FILENAME
|
||||
|
||||
# Alternatively, use the exec runner to execute a binary found on your PATH
|
||||
# hooks = ruff
|
||||
# ruff.type = exec
|
||||
# ruff.executable = ruff
|
||||
# ruff.options = check --fix REVISION_SCRIPT_FILENAME
|
||||
|
||||
# Logging configuration. This is also consumed by the user-maintained
|
||||
# env.py script only.
|
||||
[loggers]
|
||||
keys = root,sqlalchemy,alembic
|
||||
|
||||
[handlers]
|
||||
keys = console
|
||||
|
||||
[formatters]
|
||||
keys = generic
|
||||
|
||||
[logger_root]
|
||||
level = WARNING
|
||||
handlers = console
|
||||
qualname =
|
||||
|
||||
[logger_sqlalchemy]
|
||||
level = WARNING
|
||||
handlers =
|
||||
qualname = sqlalchemy.engine
|
||||
|
||||
[logger_alembic]
|
||||
level = INFO
|
||||
handlers =
|
||||
qualname = alembic
|
||||
|
||||
[handler_console]
|
||||
class = StreamHandler
|
||||
args = (sys.stderr,)
|
||||
level = NOTSET
|
||||
formatter = generic
|
||||
|
||||
[formatter_generic]
|
||||
format = %(levelname)-5.5s [%(name)s] %(message)s
|
||||
datefmt = %H:%M:%S
|
||||
@@ -0,0 +1 @@
|
||||
Generic single-database configuration.
|
||||
@@ -0,0 +1,83 @@
|
||||
from logging.config import fileConfig
|
||||
|
||||
from sqlalchemy import engine_from_config
|
||||
from sqlalchemy import pool
|
||||
|
||||
from alembic import context
|
||||
|
||||
# Import SQLModel and your models
|
||||
from sqlmodel import SQLModel
|
||||
from app.models import models # Ensure models are imported to register metadata
|
||||
from app.core.config import settings
|
||||
|
||||
# this is the Alembic Config object, which provides
|
||||
# access to the values within the .ini file in use.
|
||||
config = context.config
|
||||
|
||||
# Interpret the config file for Python logging.
|
||||
# This line sets up loggers basically.
|
||||
if config.config_file_name is not None:
|
||||
fileConfig(config.config_file_name)
|
||||
|
||||
# Set the SQLAlchemy URL from settings
|
||||
config.set_main_option("sqlalchemy.url", settings.DATABASE_URL)
|
||||
|
||||
# target_metadata = None
|
||||
target_metadata = SQLModel.metadata
|
||||
|
||||
# other values from the config, defined by the needs of env.py,
|
||||
# can be acquired:
|
||||
# my_important_option = config.get_main_option("my_important_option")
|
||||
# ... etc.
|
||||
|
||||
|
||||
def run_migrations_offline() -> None:
|
||||
"""Run migrations in 'offline' mode.
|
||||
|
||||
This configures the context with just a URL
|
||||
and not an Engine, though an Engine is acceptable
|
||||
here as well. By skipping the Engine creation
|
||||
we don't even need a DBAPI to be available.
|
||||
|
||||
Calls to context.execute() here emit the given string to the
|
||||
script output.
|
||||
|
||||
"""
|
||||
url = config.get_main_option("sqlalchemy.url")
|
||||
context.configure(
|
||||
url=url,
|
||||
target_metadata=target_metadata,
|
||||
literal_binds=True,
|
||||
dialect_opts={"paramstyle": "named"},
|
||||
)
|
||||
|
||||
with context.begin_transaction():
|
||||
context.run_migrations()
|
||||
|
||||
|
||||
def run_migrations_online() -> None:
|
||||
"""Run migrations in 'online' mode.
|
||||
|
||||
In this scenario we need to create an Engine
|
||||
and associate a connection with the context.
|
||||
|
||||
"""
|
||||
connectable = engine_from_config(
|
||||
config.get_section(config.config_ini_section, {}),
|
||||
prefix="sqlalchemy.",
|
||||
poolclass=pool.NullPool,
|
||||
)
|
||||
|
||||
with connectable.connect() as connection:
|
||||
context.configure(
|
||||
connection=connection, target_metadata=target_metadata
|
||||
)
|
||||
|
||||
with context.begin_transaction():
|
||||
context.run_migrations()
|
||||
|
||||
|
||||
if context.is_offline_mode():
|
||||
run_migrations_offline()
|
||||
else:
|
||||
run_migrations_online()
|
||||
@@ -0,0 +1,28 @@
|
||||
"""${message}
|
||||
|
||||
Revision ID: ${up_revision}
|
||||
Revises: ${down_revision | comma,n}
|
||||
Create Date: ${create_date}
|
||||
|
||||
"""
|
||||
from typing import Sequence, Union
|
||||
|
||||
from alembic import op
|
||||
import sqlalchemy as sa
|
||||
${imports if imports else ""}
|
||||
|
||||
# revision identifiers, used by Alembic.
|
||||
revision: str = ${repr(up_revision)}
|
||||
down_revision: Union[str, Sequence[str], None] = ${repr(down_revision)}
|
||||
branch_labels: Union[str, Sequence[str], None] = ${repr(branch_labels)}
|
||||
depends_on: Union[str, Sequence[str], None] = ${repr(depends_on)}
|
||||
|
||||
|
||||
def upgrade() -> None:
|
||||
"""Upgrade schema."""
|
||||
${upgrades if upgrades else "pass"}
|
||||
|
||||
|
||||
def downgrade() -> None:
|
||||
"""Downgrade schema."""
|
||||
${downgrades if downgrades else "pass"}
|
||||
@@ -0,0 +1,110 @@
|
||||
"""Initial Full Schema
|
||||
|
||||
Revision ID: ca058f9ddb60
|
||||
Revises:
|
||||
Create Date: 2026-01-17 10:58:17.473117
|
||||
|
||||
"""
|
||||
from typing import Sequence, Union
|
||||
|
||||
from alembic import op
|
||||
import sqlalchemy as sa
|
||||
|
||||
|
||||
# revision identifiers, used by Alembic.
|
||||
revision: str = 'ca058f9ddb60'
|
||||
down_revision: Union[str, Sequence[str], None] = None
|
||||
branch_labels: Union[str, Sequence[str], None] = None
|
||||
depends_on: Union[str, Sequence[str], None] = None
|
||||
|
||||
|
||||
def upgrade() -> None:
|
||||
"""Upgrade schema."""
|
||||
# ### commands auto generated by Alembic - please adjust! ###
|
||||
op.create_table('modelconfig',
|
||||
sa.Column('provider', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('api_key', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('base_url', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('model_name', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('model_type', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('is_active', sa.Boolean(), nullable=False),
|
||||
sa.Column('id', sa.Integer(), nullable=False),
|
||||
sa.PrimaryKeyConstraint('id')
|
||||
)
|
||||
op.create_table('project',
|
||||
sa.Column('title', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('description', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('story_input', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('theme', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('language', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('panel_count', sa.Integer(), nullable=True),
|
||||
sa.Column('aspect_ratio', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('id', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('created_at', sa.DateTime(), nullable=False),
|
||||
sa.Column('updated_at', sa.DateTime(), nullable=False),
|
||||
sa.PrimaryKeyConstraint('id')
|
||||
)
|
||||
op.create_table('character',
|
||||
sa.Column('name', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('data', sa.JSON(), nullable=True),
|
||||
sa.Column('image_url', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('id', sa.Integer(), nullable=False),
|
||||
sa.Column('project_id', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.ForeignKeyConstraint(['project_id'], ['project.id'], ),
|
||||
sa.PrimaryKeyConstraint('id')
|
||||
)
|
||||
op.create_table('globalconfig',
|
||||
sa.Column('data', sa.JSON(), nullable=True),
|
||||
sa.Column('id', sa.Integer(), nullable=False),
|
||||
sa.Column('project_id', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.ForeignKeyConstraint(['project_id'], ['project.id'], ),
|
||||
sa.PrimaryKeyConstraint('id')
|
||||
)
|
||||
op.create_table('imagehistory',
|
||||
sa.Column('entity_type', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('entity_id', sa.Integer(), nullable=False),
|
||||
sa.Column('image_url', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('id', sa.Integer(), nullable=False),
|
||||
sa.Column('project_id', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('created_at', sa.DateTime(), nullable=False),
|
||||
sa.ForeignKeyConstraint(['project_id'], ['project.id'], ),
|
||||
sa.PrimaryKeyConstraint('id')
|
||||
)
|
||||
op.create_table('storyboarditem',
|
||||
sa.Column('sequence', sa.Integer(), nullable=False),
|
||||
sa.Column('data', sa.JSON(), nullable=True),
|
||||
sa.Column('image_url', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('id', sa.Integer(), nullable=False),
|
||||
sa.Column('project_id', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.ForeignKeyConstraint(['project_id'], ['project.id'], ),
|
||||
sa.PrimaryKeyConstraint('id')
|
||||
)
|
||||
op.create_table('task',
|
||||
sa.Column('type', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('status', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('progress', sa.Integer(), nullable=False),
|
||||
sa.Column('message', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('name', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('description', sqlmodel.sql.sqltypes.AutoString(), nullable=True),
|
||||
sa.Column('result', sa.JSON(), nullable=True),
|
||||
sa.Column('id', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('project_id', sqlmodel.sql.sqltypes.AutoString(), nullable=False),
|
||||
sa.Column('created_at', sa.DateTime(), nullable=False),
|
||||
sa.Column('updated_at', sa.DateTime(), nullable=False),
|
||||
sa.ForeignKeyConstraint(['project_id'], ['project.id'], ),
|
||||
sa.PrimaryKeyConstraint('id')
|
||||
)
|
||||
# ### end Alembic commands ###
|
||||
|
||||
|
||||
def downgrade() -> None:
|
||||
"""Downgrade schema."""
|
||||
# ### commands auto generated by Alembic - please adjust! ###
|
||||
op.drop_table('task')
|
||||
op.drop_table('storyboarditem')
|
||||
op.drop_table('imagehistory')
|
||||
op.drop_table('globalconfig')
|
||||
op.drop_table('character')
|
||||
op.drop_table('project')
|
||||
op.drop_table('modelconfig')
|
||||
# ### end Alembic commands ###
|
||||
@@ -0,0 +1,12 @@
|
||||
import os
|
||||
from pydantic_settings import BaseSettings
|
||||
|
||||
class Settings(BaseSettings):
|
||||
PROJECT_NAME: str = "AI Comic Generator"
|
||||
API_V1_STR: str = "/api/v1"
|
||||
DATABASE_URL: str = "sqlite:///./comic_app.db"
|
||||
|
||||
class Config:
|
||||
env_file = ".env"
|
||||
|
||||
settings = Settings()
|
||||
@@ -0,0 +1,15 @@
|
||||
from sqlmodel import SQLModel, create_engine, Session
|
||||
from app.core.config import settings
|
||||
|
||||
connect_args = {}
|
||||
if "sqlite" in settings.DATABASE_URL:
|
||||
connect_args["check_same_thread"] = False
|
||||
|
||||
engine = create_engine(settings.DATABASE_URL, echo=False, connect_args=connect_args)
|
||||
|
||||
def init_db():
|
||||
SQLModel.metadata.create_all(engine)
|
||||
|
||||
def get_session():
|
||||
with Session(engine) as session:
|
||||
yield session
|
||||
@@ -0,0 +1,102 @@
|
||||
# This file is used to centrally manage the core system prompt (System Prompt) of the comic generation system
|
||||
# Provided to the backend service in the form of Python variables to avoid path or encoding issues caused by directly reading text files
|
||||
|
||||
COMIC_GENERATION_SYSTEM_PROMPT = """As a 'Comic Split Generation' expert, you will play the dual role of a comic editing expert and a comic drawing expert. Your goal is to assist users in the entire process from story creation to storyboard design, and finally generate a serialized comic.
|
||||
|
||||
Purpose and Goals:
|
||||
* Expand or optimize the user-provided content into a story with a complete plot, ensuring logical self-consistency and engagement.
|
||||
* Clarify the overall comic style. You must strictly use the user-specified style (if any) and are prohibited from making decisions on your own.
|
||||
* Extract characters from the story and generate character setting cards containing core information about appearance, clothing, and personality.
|
||||
* Split the story into serialized comic storyboards, ensuring natural transitions between panels and appropriate narrative pacing.
|
||||
* Provide extremely detailed visual descriptions for each panel, precise to lighting, composition, character expression, and key actions.
|
||||
* Provide storyboard content in JSON file format, strictly stipulating that every four panels constitute an independent JSON object block for structured processing.
|
||||
|
||||
Behaviors and Rules:
|
||||
|
||||
1) Story Optimization & Character Setting (Story & Character):
|
||||
a) Receive the user's initial idea, enrich its background details, emotional ups and downs, and climax ending.
|
||||
b) Confirm Art Style: If the user provides specific "Theme" or "Style" requirements, you must follow them unconditionally. Do not modify the style based on the story content. For example, if the user requests "Cyberpunk", even if the story is set in a martial arts background, you must generate "Cyberpunk style martial arts".
|
||||
c) Before starting the storyboard, list the main characters and their physical characteristics (such as hair color, eye color, signature accessories, etc.) in detail to ensure visual consistency of characters in subsequent storyboards.
|
||||
|
||||
2) Storyboard Splitting & JSON Construction (Storyboard & JSON):
|
||||
a) Comic Global Configuration (Comic Configuration):
|
||||
- **Must be generated first**: Before starting to generate characters and storyboards, you must generate an independent 'comic_config' JSON block.
|
||||
- **Function**: Define the visual tone, typography standards, and border styles of the entire comic.
|
||||
- JSON structure should contain:
|
||||
- 'type': 'comic_config'
|
||||
- 'language': 'English',
|
||||
- 'style': '{User Specified Style}' (Must fill in the user-specified style, if not specified, default to 'Chibi/Fantasy Style')
|
||||
- 'bubble_style': { 'shape': 'Bubble Shape', 'color': 'Background Color', 'font_color': 'Font Color', 'stroke_width': 'Stroke Width' }
|
||||
- 'narration_style': { 'shape': 'Box/Rounded', 'color': 'Background Color', 'font_color': 'Font Color', 'opacity': 'Opacity' }
|
||||
- 'border_style': { 'width': 'Line Width', 'color': 'Color', 'type': 'Solid/Hand-drawn' }
|
||||
- 'gutter_style': { 'type': 'Standard Cross Split', 'color': 'White', 'width': '10px' }
|
||||
- 'layout_settings': {
|
||||
'show_panel_numbers': false, // [Switch] Whether to show panel numbers
|
||||
'panel_number_style': { 'position': 'top-left', 'bg_color': '#000000', 'text_color': '#00FF41', 'font_size': '14px' },
|
||||
'force_uniform_borders': true,
|
||||
'composition_mode': 'grid' // grid=Grid Splicing, cinematic=Cinematic Widescreen
|
||||
}
|
||||
- 'aspect_ratio': '16:9'
|
||||
|
||||
b) Character Sheet Generation (Character Sheets):
|
||||
- **All characters must have setting cards**: Before generating story storyboards, you must generate independent JSON setting blocks for all named characters appearing in the story, including protagonists, frequent supporting characters, and villains.
|
||||
- **Protagonist Design**: The protagonist's image must be designed to be extremely attractive, with distinct physical features, meeting "high aesthetic" standards.
|
||||
- **Character Deduplication**: When generating the character list, carefully identify different names for the same character (e.g., "Butler Ma" and "Old Ma" are the same person). If found to be the same person, generate only one character setting card and use the most formal or common name in the name field. Strictly prohibit generating multiple duplicate setting cards for the same character.
|
||||
- If there are multiple main characters, please generate multiple independent 'character_sheet' JSON blocks respectively, or output them in a JSON array.
|
||||
- JSON structure should contain:
|
||||
- 'type': 'character_sheet'
|
||||
- 'name': 'Character Name'
|
||||
- 'meta_info':{
|
||||
- 'language': Language based on user input
|
||||
- 'role': 'Protagonist' | 'Supporting' | 'Extra' (Must indicate character type)
|
||||
- 'personality': 'Character personality traits, e.g., Cheerful, Cold, Hot-blooded, etc., which will affect expressions and poses'
|
||||
- 'age': 'Age description, approximate range'
|
||||
- 'relationships': 'Description of relationship with protagonist or other characters' (Must indicate interpersonal relationships)
|
||||
- 'style': '{User Specified Style}',
|
||||
- 'feature': 'Explicit character features, e.g., Youthful, Plump, etc.'
|
||||
- 'aspect_ratio': '16:9'
|
||||
}
|
||||
- 'design_panels': [
|
||||
{'view': 'Front View', 'description': 'Detailed front full-body description...'},
|
||||
{'view': 'Side View', 'description': 'Detailed side view description...'},
|
||||
{'view': 'Clothing', 'description': 'Detailed clothing details...'},
|
||||
{'view': 'Accessories', 'description': 'Detailed accessories/weapon details...'}
|
||||
]
|
||||
|
||||
c) Story Storyboard Generation (Story Storyboard):
|
||||
- Decompose the optimized story into concrete, visualizable storyboard frames.
|
||||
- **Storyboard Count Mandatory Requirement**: The total number of generated panels must be determined based on the story content. The richer the story, the more panels, unless the user input requires a minimum number of panels (e.g., "at least 36 panels").
|
||||
- The total number of generated panels must be >= the minimum number required by the user.
|
||||
- The total number of generated panels must be an integer multiple of 4 (e.g., 36, 40, 44...), rounding up if not satisfied.
|
||||
|
||||
- Construct every four panels as an independent JSON code block output.
|
||||
- JSON structure should contain:
|
||||
- 'type': 'storyboard'
|
||||
- 'meta_info': {
|
||||
'style': '{User Specified Style}',
|
||||
'language': 'English',
|
||||
'volume': 'Current Volume/Total Volumes',
|
||||
'aspect_ratio': '16:9',
|
||||
}
|
||||
- 'characters': ['List of characters appearing in this group of panels']
|
||||
- 'plot_breakdown': [
|
||||
{'panel': 1, 'scene': '...', 'action': '...', 'dialogue': '...'},
|
||||
{'panel': 2, 'scene': '...', 'action': '...', 'dialogue': '...'},
|
||||
{'panel': 3, 'scene': '...', 'action': '...', 'dialogue': '...'},
|
||||
{'panel': 4, 'scene': '...', 'action': '...', 'dialogue': '...'}
|
||||
]
|
||||
|
||||
3) Quality Control (Quality Control):
|
||||
a) Ensure visual logic consistency between panels, avoiding sudden changes in characters or environment.
|
||||
b) Visual prompts should include elements such as environment, weather, shot type (e.g., close-up, panoramic), etc.
|
||||
|
||||
4) Language & Format Requirements (Language & Format):
|
||||
a) Use the same language for dialogue and narration as the user input. English is the standard.
|
||||
b) All JSON outputs must maintain a strict, parsable code block format.
|
||||
c) All JSON keys must use lowercase English (e.g., 'language', 'style').
|
||||
|
||||
|
||||
Overall Tone:
|
||||
* Professional and highly creative, demonstrating the rigor and aesthetics of a senior industry practitioner.
|
||||
* Descriptions of visual details should be precise and evocative.
|
||||
"""
|
||||
@@ -0,0 +1,35 @@
|
||||
from sqlmodel import Session, select
|
||||
from app.models.models import ModelConfig
|
||||
from app.schemas.schemas import ModelConfigCreate, ModelConfigUpdate
|
||||
from typing import List, Optional
|
||||
|
||||
def create_model_config(session: Session, config_in: ModelConfigCreate) -> ModelConfig:
|
||||
db_config = ModelConfig.model_validate(config_in)
|
||||
session.add(db_config)
|
||||
session.commit()
|
||||
session.refresh(db_config)
|
||||
return db_config
|
||||
|
||||
def get_model_configs(session: Session, skip: int = 0, limit: int = 100) -> List[ModelConfig]:
|
||||
statement = select(ModelConfig).offset(skip).limit(limit)
|
||||
return session.exec(statement).all()
|
||||
|
||||
def get_model_config(session: Session, config_id: int) -> Optional[ModelConfig]:
|
||||
return session.get(ModelConfig, config_id)
|
||||
|
||||
def update_model_config(session: Session, db_config: ModelConfig, config_in: ModelConfigUpdate) -> ModelConfig:
|
||||
config_data = config_in.model_dump(exclude_unset=True)
|
||||
for key, value in config_data.items():
|
||||
setattr(db_config, key, value)
|
||||
session.add(db_config)
|
||||
session.commit()
|
||||
session.refresh(db_config)
|
||||
return db_config
|
||||
|
||||
def delete_model_config(session: Session, db_config: ModelConfig):
|
||||
session.delete(db_config)
|
||||
session.commit()
|
||||
|
||||
def get_active_config(session: Session, model_type: str) -> Optional[ModelConfig]:
|
||||
statement = select(ModelConfig).where(ModelConfig.model_type == model_type, ModelConfig.is_active == True)
|
||||
return session.exec(statement).first()
|
||||
@@ -0,0 +1,99 @@
|
||||
from sqlmodel import Session, select
|
||||
from app.models.models import Project, Character, StoryboardItem, GlobalConfig
|
||||
from app.schemas.schemas import ProjectCreate, ProjectUpdate
|
||||
from typing import List, Optional
|
||||
|
||||
def create_project(session: Session, project_in: ProjectCreate) -> Project:
|
||||
db_project = Project.model_validate(project_in)
|
||||
session.add(db_project)
|
||||
session.commit()
|
||||
session.refresh(db_project)
|
||||
return db_project
|
||||
|
||||
def get_projects(session: Session, skip: int = 0, limit: int = 100) -> List[Project]:
|
||||
statement = select(Project).offset(skip).limit(limit).order_by(Project.updated_at.desc())
|
||||
return session.exec(statement).all()
|
||||
|
||||
def get_project(session: Session, project_id: str) -> Optional[Project]:
|
||||
return session.get(Project, project_id)
|
||||
|
||||
def update_project(session: Session, db_project: Project, project_in: ProjectUpdate) -> Project:
|
||||
project_data = project_in.model_dump(exclude_unset=True)
|
||||
for key, value in project_data.items():
|
||||
setattr(db_project, key, value)
|
||||
session.add(db_project)
|
||||
session.commit()
|
||||
session.refresh(db_project)
|
||||
return db_project
|
||||
|
||||
def delete_project(session: Session, db_project: Project):
|
||||
session.delete(db_project)
|
||||
session.commit()
|
||||
|
||||
# Helpers for sub-entities
|
||||
def create_global_config(session: Session, project_id: str, data: dict) -> GlobalConfig:
|
||||
# Check if exists
|
||||
statement = select(GlobalConfig).where(GlobalConfig.project_id == project_id)
|
||||
existing = session.exec(statement).first()
|
||||
if existing:
|
||||
existing.data = data
|
||||
session.add(existing)
|
||||
session.commit()
|
||||
session.refresh(existing)
|
||||
return existing
|
||||
|
||||
db_config = GlobalConfig(project_id=project_id, data=data)
|
||||
session.add(db_config)
|
||||
session.commit()
|
||||
session.refresh(db_config)
|
||||
return db_config
|
||||
|
||||
def save_characters(session: Session, project_id: str, characters_data: List[dict]) -> List[Character]:
|
||||
# Strategy: Clear existing or update?
|
||||
# For simplicity in this flow: Clear and Re-insert is easier if we regenerate all.
|
||||
# But user might want to edit specific ones.
|
||||
# Better: Update by name match, create if new.
|
||||
|
||||
results = []
|
||||
for char_data in characters_data:
|
||||
name = char_data.get("name")
|
||||
if not name: continue
|
||||
|
||||
statement = select(Character).where(Character.project_id == project_id, Character.name == name)
|
||||
existing = session.exec(statement).first()
|
||||
|
||||
if existing:
|
||||
existing.data = char_data
|
||||
session.add(existing)
|
||||
results.append(existing)
|
||||
else:
|
||||
new_char = Character(project_id=project_id, name=name, data=char_data)
|
||||
session.add(new_char)
|
||||
results.append(new_char)
|
||||
|
||||
session.commit()
|
||||
return results
|
||||
|
||||
def save_storyboard(session: Session, project_id: str, storyboard_data: List[dict]) -> List[StoryboardItem]:
|
||||
# Similar strategy: Clear and Re-insert is risky if we have images.
|
||||
# But storyboard is sequential.
|
||||
# Let's delete all and re-insert for now as "Regenerate JSON" usually means fresh start.
|
||||
# IF the user is just editing JSON text, we replace everything.
|
||||
|
||||
# Check if there are existing items with images we want to preserve?
|
||||
# Ideally, we should try to map them back, but it's hard if sequence changes.
|
||||
# For now: delete all items for this project and insert new.
|
||||
|
||||
statement = select(StoryboardItem).where(StoryboardItem.project_id == project_id)
|
||||
existing_items = session.exec(statement).all()
|
||||
for item in existing_items:
|
||||
session.delete(item)
|
||||
|
||||
results = []
|
||||
for i, item_data in enumerate(storyboard_data):
|
||||
new_item = StoryboardItem(project_id=project_id, sequence=i+1, data=item_data)
|
||||
session.add(new_item)
|
||||
results.append(new_item)
|
||||
|
||||
session.commit()
|
||||
return results
|
||||
@@ -0,0 +1,40 @@
|
||||
from fastapi import FastAPI
|
||||
from fastapi.staticfiles import StaticFiles
|
||||
from fastapi.middleware.cors import CORSMiddleware
|
||||
from app.core.config import settings
|
||||
from app.core.database import init_db
|
||||
from app.routers import configs, projects, generation, export, tasks, history
|
||||
import os
|
||||
|
||||
app = FastAPI(title=settings.PROJECT_NAME)
|
||||
|
||||
# CORS
|
||||
app.add_middleware(
|
||||
CORSMiddleware,
|
||||
allow_origins=["*"],
|
||||
allow_credentials=True,
|
||||
allow_methods=["*"],
|
||||
allow_headers=["*"],
|
||||
)
|
||||
|
||||
# Mount static files
|
||||
static_dir = os.path.join(os.path.dirname(os.path.dirname(__file__)), "static")
|
||||
if not os.path.exists(static_dir):
|
||||
os.makedirs(static_dir)
|
||||
app.mount("/static", StaticFiles(directory=static_dir), name="static")
|
||||
|
||||
# Include routers
|
||||
app.include_router(configs.router, prefix=f"{settings.API_V1_STR}/configs", tags=["configs"])
|
||||
app.include_router(projects.router, prefix=f"{settings.API_V1_STR}/projects", tags=["projects"])
|
||||
app.include_router(generation.router, prefix=f"{settings.API_V1_STR}/generate", tags=["generation"])
|
||||
app.include_router(export.router, prefix=f"{settings.API_V1_STR}/export", tags=["export"])
|
||||
app.include_router(tasks.router, prefix=f"{settings.API_V1_STR}/tasks", tags=["tasks"])
|
||||
app.include_router(history.router, prefix=f"{settings.API_V1_STR}/history", tags=["history"])
|
||||
|
||||
@app.on_event("startup")
|
||||
def on_startup():
|
||||
init_db()
|
||||
|
||||
@app.get("/")
|
||||
def read_root():
|
||||
return {"message": "Welcome to AI Comic Generator API"}
|
||||
@@ -0,0 +1,97 @@
|
||||
from typing import Optional, List, Dict, Any
|
||||
from sqlmodel import SQLModel, Field, Relationship, Column, JSON
|
||||
from datetime import datetime
|
||||
import uuid
|
||||
|
||||
# --- Base Models ---
|
||||
|
||||
class ModelConfigBase(SQLModel):
|
||||
provider: str
|
||||
api_key: str
|
||||
base_url: Optional[str] = None
|
||||
model_name: str
|
||||
model_type: str
|
||||
is_active: bool = True
|
||||
|
||||
class ProjectBase(SQLModel):
|
||||
title: str
|
||||
description: Optional[str] = None
|
||||
story_input: Optional[str] = None
|
||||
# Generation Preferences
|
||||
theme: Optional[str] = None
|
||||
language: Optional[str] = "zh-CN"
|
||||
panel_count: Optional[int] = 16
|
||||
aspect_ratio: Optional[str] = "16:9"
|
||||
|
||||
class CharacterBase(SQLModel):
|
||||
name: str
|
||||
data: Dict = Field(default={}, sa_column=Column(JSON))
|
||||
image_url: Optional[str] = None
|
||||
|
||||
class StoryboardItemBase(SQLModel):
|
||||
sequence: int
|
||||
data: Dict = Field(default={}, sa_column=Column(JSON))
|
||||
image_url: Optional[str] = None
|
||||
|
||||
class GlobalConfigBase(SQLModel):
|
||||
data: Dict = Field(default={}, sa_column=Column(JSON))
|
||||
|
||||
class TaskBase(SQLModel):
|
||||
type: str # 'storyboard', 'image_generation', 'export'
|
||||
status: str # 'pending', 'processing', 'completed', 'failed'
|
||||
progress: int = 0 # 0-100
|
||||
message: Optional[str] = None
|
||||
name: Optional[str] = None
|
||||
description: Optional[str] = None
|
||||
result: Dict = Field(default={}, sa_column=Column(JSON))
|
||||
|
||||
class ImageHistoryBase(SQLModel):
|
||||
entity_type: str # 'character' or 'storyboard_item'
|
||||
entity_id: int
|
||||
image_url: str
|
||||
|
||||
# --- Table Models ---
|
||||
|
||||
class ModelConfig(ModelConfigBase, table=True):
|
||||
id: Optional[int] = Field(default=None, primary_key=True)
|
||||
|
||||
class Project(ProjectBase, table=True):
|
||||
id: str = Field(default_factory=lambda: str(uuid.uuid4()), primary_key=True)
|
||||
created_at: datetime = Field(default_factory=datetime.utcnow)
|
||||
updated_at: datetime = Field(default_factory=datetime.utcnow)
|
||||
|
||||
characters: List["Character"] = Relationship(back_populates="project", sa_relationship_kwargs={"cascade": "all, delete"})
|
||||
storyboard_items: List["StoryboardItem"] = Relationship(back_populates="project", sa_relationship_kwargs={"cascade": "all, delete"})
|
||||
global_config: Optional["GlobalConfig"] = Relationship(back_populates="project", sa_relationship_kwargs={"cascade": "all, delete"})
|
||||
tasks: List["Task"] = Relationship(back_populates="project", sa_relationship_kwargs={"cascade": "all, delete"})
|
||||
image_history: List["ImageHistory"] = Relationship(back_populates="project", sa_relationship_kwargs={"cascade": "all, delete"})
|
||||
|
||||
class Character(CharacterBase, table=True):
|
||||
id: Optional[int] = Field(default=None, primary_key=True)
|
||||
project_id: str = Field(foreign_key="project.id")
|
||||
project: Project = Relationship(back_populates="characters")
|
||||
|
||||
class StoryboardItem(StoryboardItemBase, table=True):
|
||||
id: Optional[int] = Field(default=None, primary_key=True)
|
||||
project_id: str = Field(foreign_key="project.id")
|
||||
project: Project = Relationship(back_populates="storyboard_items")
|
||||
|
||||
class GlobalConfig(GlobalConfigBase, table=True):
|
||||
id: Optional[int] = Field(default=None, primary_key=True)
|
||||
project_id: str = Field(foreign_key="project.id")
|
||||
project: Project = Relationship(back_populates="global_config")
|
||||
|
||||
class Task(TaskBase, table=True):
|
||||
id: str = Field(default_factory=lambda: str(uuid.uuid4()), primary_key=True)
|
||||
project_id: str = Field(foreign_key="project.id")
|
||||
created_at: datetime = Field(default_factory=datetime.utcnow)
|
||||
updated_at: datetime = Field(default_factory=datetime.utcnow)
|
||||
|
||||
project: Project = Relationship(back_populates="tasks")
|
||||
|
||||
class ImageHistory(ImageHistoryBase, table=True):
|
||||
id: Optional[int] = Field(default=None, primary_key=True)
|
||||
project_id: str = Field(foreign_key="project.id")
|
||||
created_at: datetime = Field(default_factory=datetime.utcnow)
|
||||
|
||||
project: Project = Relationship(back_populates="image_history")
|
||||
@@ -0,0 +1,39 @@
|
||||
from fastapi import APIRouter, Depends, HTTPException
|
||||
from sqlmodel import Session
|
||||
from typing import List
|
||||
from app.core.database import get_session
|
||||
from app.models.models import ModelConfig
|
||||
from app.schemas.schemas import ModelConfigCreate, ModelConfigUpdate
|
||||
from app.cruds import crud_config
|
||||
|
||||
router = APIRouter()
|
||||
|
||||
@router.post("/", response_model=ModelConfig)
|
||||
def create_config(config_in: ModelConfigCreate, session: Session = Depends(get_session)):
|
||||
return crud_config.create_model_config(session, config_in)
|
||||
|
||||
@router.get("/", response_model=List[ModelConfig])
|
||||
def read_configs(skip: int = 0, limit: int = 100, session: Session = Depends(get_session)):
|
||||
return crud_config.get_model_configs(session, skip, limit)
|
||||
|
||||
@router.get("/{config_id}", response_model=ModelConfig)
|
||||
def read_config(config_id: int, session: Session = Depends(get_session)):
|
||||
config = crud_config.get_model_config(session, config_id)
|
||||
if not config:
|
||||
raise HTTPException(status_code=404, detail="Config not found")
|
||||
return config
|
||||
|
||||
@router.put("/{config_id}", response_model=ModelConfig)
|
||||
def update_config(config_id: int, config_in: ModelConfigUpdate, session: Session = Depends(get_session)):
|
||||
config = crud_config.get_model_config(session, config_id)
|
||||
if not config:
|
||||
raise HTTPException(status_code=404, detail="Config not found")
|
||||
return crud_config.update_model_config(session, config, config_in)
|
||||
|
||||
@router.delete("/{config_id}")
|
||||
def delete_config(config_id: int, session: Session = Depends(get_session)):
|
||||
config = crud_config.get_model_config(session, config_id)
|
||||
if not config:
|
||||
raise HTTPException(status_code=404, detail="Config not found")
|
||||
crud_config.delete_model_config(session, config)
|
||||
return {"ok": True}
|
||||
@@ -0,0 +1,80 @@
|
||||
import shutil
|
||||
import os
|
||||
from fastapi import APIRouter, Depends, HTTPException
|
||||
from sqlmodel import Session
|
||||
from app.core.database import get_session
|
||||
from app.models.models import Project
|
||||
from app.services.image_service import split_comic_page
|
||||
|
||||
router = APIRouter()
|
||||
|
||||
@router.get("/{project_id}")
|
||||
def export_project(
|
||||
project_id: str,
|
||||
split_images: bool = False,
|
||||
session: Session = Depends(get_session)
|
||||
):
|
||||
project = session.get(Project, project_id)
|
||||
if not project:
|
||||
raise HTTPException(status_code=404, detail="Project not found")
|
||||
|
||||
# Check if any images generated
|
||||
has_images = any(item.image_url for item in project.storyboard_items) or any(c.image_url for c in project.characters)
|
||||
if not has_images:
|
||||
raise HTTPException(status_code=400, detail="No images generated yet. Cannot export.")
|
||||
|
||||
base_dir = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||
project_static_dir = os.path.join(base_dir, "static", project_id)
|
||||
export_dir = os.path.join(project_static_dir, "export")
|
||||
|
||||
if os.path.exists(export_dir):
|
||||
shutil.rmtree(export_dir)
|
||||
os.makedirs(export_dir)
|
||||
|
||||
# Export Characters
|
||||
chars_dir = os.path.join(export_dir, "characters")
|
||||
os.makedirs(chars_dir)
|
||||
for char in project.characters:
|
||||
if char.image_url:
|
||||
# Resolve absolute path from relative URL
|
||||
# URL: /static/{project_id}/characters/xxx.png
|
||||
# Path: backend/static/{project_id}/characters/xxx.png
|
||||
# We can construct it directly if we know the structure, but let's parse url
|
||||
rel_path = char.image_url.lstrip("/") # static/...
|
||||
local_path = os.path.join(base_dir, rel_path.replace("/", os.sep))
|
||||
|
||||
if os.path.exists(local_path):
|
||||
shutil.copy(local_path, os.path.join(chars_dir, f"{char.name}.png"))
|
||||
|
||||
panels_dir = os.path.join(export_dir, "panels")
|
||||
if split_images:
|
||||
os.makedirs(panels_dir)
|
||||
|
||||
# Sort items
|
||||
items = sorted(project.storyboard_items, key=lambda x: x.sequence)
|
||||
|
||||
for item in items:
|
||||
if item.image_url:
|
||||
rel_path = item.image_url.lstrip("/")
|
||||
local_path = os.path.join(base_dir, rel_path.replace("/", os.sep))
|
||||
|
||||
if os.path.exists(local_path):
|
||||
shutil.copy(local_path, os.path.join(export_dir, f"comic_part_{item.sequence}.png"))
|
||||
|
||||
if split_images:
|
||||
with open(local_path, "rb") as f:
|
||||
img_bytes = f.read()
|
||||
|
||||
try:
|
||||
panels = split_comic_page(img_bytes)
|
||||
for idx, panel_bytes in enumerate(panels):
|
||||
p_name = f"panel_{item.sequence}_{idx+1}.png"
|
||||
with open(os.path.join(panels_dir, p_name), "wb") as f:
|
||||
f.write(panel_bytes)
|
||||
except Exception as e:
|
||||
print(f"Failed to split panel {item.id}: {e}")
|
||||
|
||||
zip_path_base = os.path.join(project_static_dir, "export_archive")
|
||||
shutil.make_archive(zip_path_base, 'zip', export_dir)
|
||||
|
||||
return {"download_url": f"/static/{project_id}/export_archive.zip"}
|
||||
@@ -0,0 +1,824 @@
|
||||
from fastapi import APIRouter, Depends, HTTPException, BackgroundTasks
|
||||
from sqlmodel import Session
|
||||
from app.core.database import get_session
|
||||
from app.models.models import Project, Character, StoryboardItem, Task, ImageHistory
|
||||
from app.services.ai_service import AIService
|
||||
from app.services.consistency_service import ConsistencyService
|
||||
from app.utils.json_utils import extract_json_blocks
|
||||
from app.cruds import crud_project
|
||||
import os
|
||||
import uuid
|
||||
import json
|
||||
import traceback
|
||||
|
||||
import logging
|
||||
import sys
|
||||
|
||||
# Configure logging
|
||||
logging.basicConfig(level=logging.INFO, stream=sys.stdout)
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
def save_generated_image(session, project_id, entity_type, entity_id, image_bytes):
|
||||
base_dir = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||
project_static_dir = os.path.join(base_dir, "static", project_id)
|
||||
|
||||
sub_dir = "characters" if entity_type == "character" else "panels"
|
||||
target_dir = os.path.join(project_static_dir, sub_dir)
|
||||
|
||||
if not os.path.exists(target_dir):
|
||||
os.makedirs(target_dir)
|
||||
|
||||
filename = f"{entity_type}_{entity_id}_{uuid.uuid4().hex[:8]}.png"
|
||||
filepath = os.path.join(target_dir, filename)
|
||||
|
||||
with open(filepath, "wb") as f:
|
||||
f.write(image_bytes)
|
||||
|
||||
relative_url = f"/static/{project_id}/{sub_dir}/{filename}"
|
||||
|
||||
# Save History
|
||||
history = ImageHistory(
|
||||
project_id=project_id,
|
||||
entity_type=entity_type,
|
||||
entity_id=entity_id,
|
||||
image_url=relative_url
|
||||
)
|
||||
session.add(history)
|
||||
|
||||
return relative_url
|
||||
|
||||
router = APIRouter()
|
||||
|
||||
from app.core.prompts import COMIC_GENERATION_SYSTEM_PROMPT
|
||||
|
||||
def get_system_prompt():
|
||||
return COMIC_GENERATION_SYSTEM_PROMPT
|
||||
|
||||
# --- Background Task Functions ---
|
||||
|
||||
def generate_storyboard_task(task_id: str, project_id: str, user_input: str):
|
||||
logger.info(f"Starting storyboard generation task: {task_id} for project: {project_id}")
|
||||
# We need a fresh session for the background task
|
||||
from app.core.database import engine
|
||||
with Session(engine) as session:
|
||||
task = session.get(Task, task_id)
|
||||
if not task:
|
||||
logger.error(f"Task {task_id} not found")
|
||||
return
|
||||
|
||||
task.status = "processing"
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
try:
|
||||
project = crud_project.get_project(session, project_id)
|
||||
logger.info(f"Project found: {project.title}")
|
||||
|
||||
# Save User Input (Persist it)
|
||||
project.story_input = user_input
|
||||
session.add(project)
|
||||
session.commit()
|
||||
|
||||
ai = AIService(session)
|
||||
system_prompt = get_system_prompt()
|
||||
|
||||
# Construct Final Prompt with Preferences
|
||||
final_prompt = user_input
|
||||
|
||||
# --- Replace Placeholders in System Prompt ---
|
||||
system_prompt = get_system_prompt()
|
||||
|
||||
# Defaults
|
||||
style = "Standard"
|
||||
if project.theme: style = project.theme
|
||||
|
||||
lang = "English"
|
||||
if project.language:
|
||||
lang_map = {"zh-CN": "Simplified Chinese", "en-US": "English", "ja-JP": "Japanese"}
|
||||
lang = lang_map.get(project.language, project.language)
|
||||
|
||||
# Inject into System Prompt
|
||||
system_prompt = system_prompt.replace("{User Specified Style}", style)
|
||||
# We could also inject language if we had a placeholder, but style is the main one failing.
|
||||
# Let's add language instruction to system prompt dynamically if needed,
|
||||
# or rely on the "Language & Format" section in prompt which says "Use user input language".
|
||||
|
||||
# --- Construct User Prompt ---
|
||||
final_prompt = user_input
|
||||
|
||||
# Append preferences as normal requirements
|
||||
prefs = []
|
||||
if project.theme: prefs.append(f"Theme: {project.theme}")
|
||||
if project.language: prefs.append(f"Language: {project.language}")
|
||||
if project.panel_count: prefs.append(f"Estimated Panel Count: {project.panel_count}")
|
||||
if project.aspect_ratio: prefs.append(f"Aspect Ratio: {project.aspect_ratio}")
|
||||
|
||||
if prefs:
|
||||
final_prompt += "\n\nRequirements:\n" + "\n".join(prefs)
|
||||
|
||||
logger.info("Calling AI service for storyboard generation...")
|
||||
generated_text = ai.generate_storyboard(system_prompt, final_prompt)
|
||||
|
||||
# --- Save Generated Text to Temp File ---
|
||||
import time
|
||||
base_dir = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||
temp_dir = os.path.join(base_dir, "static", project_id, "temp")
|
||||
if not os.path.exists(temp_dir):
|
||||
os.makedirs(temp_dir)
|
||||
|
||||
timestamp = int(time.time())
|
||||
temp_file = os.path.join(temp_dir, f"ai_output_{timestamp}.txt")
|
||||
try:
|
||||
with open(temp_file, "w", encoding="utf-8") as f:
|
||||
f.write(generated_text)
|
||||
logger.info(f"Saved raw AI output to {temp_file}")
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to save temp AI output: {e}")
|
||||
# ----------------------------------------
|
||||
|
||||
logger.info("AI generation complete. Extracting JSON blocks...")
|
||||
json_blocks = extract_json_blocks(generated_text)
|
||||
|
||||
char_blocks = [b for b in json_blocks if b.get("type") == "character_sheet"]
|
||||
story_blocks = [b for b in json_blocks if b.get("type") == "storyboard"]
|
||||
|
||||
if not story_blocks:
|
||||
story_blocks = [b for b in json_blocks if b.get("type") not in ["character_sheet", "comic_config"]]
|
||||
|
||||
# --- Missing Character Check & Fix ---
|
||||
story_char_names = set()
|
||||
for block in story_blocks:
|
||||
chars = block.get("characters", [])
|
||||
if isinstance(chars, str):
|
||||
story_char_names.add(chars)
|
||||
elif isinstance(chars, list):
|
||||
for c in chars:
|
||||
if isinstance(c, str): story_char_names.add(c)
|
||||
elif isinstance(c, dict): story_char_names.add(c.get("name", ""))
|
||||
|
||||
generated_char_names = set(b.get("name") for b in char_blocks if b.get("name"))
|
||||
|
||||
# Simple fuzzy matching or direct check
|
||||
missing_chars = []
|
||||
for name in story_char_names:
|
||||
# Check if name is contained in any generated char name (e.g. "Xiao Ming" vs "Ming")
|
||||
found = False
|
||||
for g_name in generated_char_names:
|
||||
if name in g_name or g_name in name:
|
||||
found = True
|
||||
break
|
||||
if not found and name and len(name) > 1: # Ignore single chars or empty
|
||||
missing_chars.append(name)
|
||||
|
||||
if missing_chars:
|
||||
logger.info(f"Detected missing characters: {missing_chars}. Requesting AI to generate them...")
|
||||
fix_prompt = f"You missed generating character sheets for the following characters that appeared in the storyboard: {', '.join(missing_chars)}. Please generate 'character_sheet' JSON blocks for them now. Do not generate anything else."
|
||||
|
||||
try:
|
||||
fix_response = ai.generate_storyboard(system_prompt, fix_prompt) # Re-use generate method
|
||||
fix_blocks = extract_json_blocks(fix_response)
|
||||
new_chars = [b for b in fix_blocks if b.get("type") == "character_sheet"]
|
||||
if new_chars:
|
||||
logger.info(f"Successfully generated {len(new_chars)} missing characters.")
|
||||
char_blocks.extend(new_chars)
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to generate missing characters: {e}")
|
||||
|
||||
config_block = next((b for b in json_blocks if b.get("type") == "comic_config"), None)
|
||||
|
||||
# If AI didn't return config, create one from project prefs
|
||||
if not config_block and (project.aspect_ratio or project.language):
|
||||
config_block = {
|
||||
"type": "comic_config",
|
||||
"style": "Standard", # Default
|
||||
"aspect_ratio": project.aspect_ratio or "16:9",
|
||||
"language": project.language or "en-US"
|
||||
}
|
||||
|
||||
if config_block:
|
||||
crud_project.create_global_config(session, project_id, config_block)
|
||||
|
||||
# --- Enforce Consistency: Update meta_info for all blocks ---
|
||||
# Re-read global config if we just created/updated it
|
||||
# Or use the config_block we have
|
||||
|
||||
if not config_block:
|
||||
# Try to fetch existing
|
||||
# But we just generated it. If None, we create a default one above.
|
||||
# Let's use the one we have.
|
||||
pass
|
||||
|
||||
if config_block:
|
||||
global_style = config_block.get("style", "")
|
||||
global_aspect = config_block.get("aspect_ratio", "16:9")
|
||||
global_lang = config_block.get("language", "en-US")
|
||||
|
||||
# Update Character Sheets
|
||||
for char in char_blocks:
|
||||
char["meta_info"] = char.get("meta_info", {})
|
||||
char["meta_info"]["language"] = global_lang
|
||||
char["meta_info"]["style"] = global_style
|
||||
|
||||
# Remove top-level redundant keys if they exist to avoid confusion
|
||||
char.pop("language", None)
|
||||
char.pop("style", None)
|
||||
|
||||
# Update Storyboard Items
|
||||
for block in story_blocks:
|
||||
meta = block.get("meta_info", {})
|
||||
meta["style"] = global_style
|
||||
meta["language"] = global_lang
|
||||
meta["aspect_ratio"] = global_aspect
|
||||
|
||||
# Also inject specific style configs if present
|
||||
if "bubble_style" in config_block: meta["bubble_style"] = config_block["bubble_style"]
|
||||
if "narration_style" in config_block: meta["narration_style"] = config_block["narration_style"]
|
||||
if "border_style" in config_block: meta["border_style"] = config_block["border_style"]
|
||||
if "gutter_style" in config_block: meta["gutter_style"] = config_block["gutter_style"]
|
||||
if "layout_settings" in config_block: meta["layout_settings"] = config_block["layout_settings"]
|
||||
|
||||
block["meta_info"] = meta
|
||||
|
||||
# Save to DB
|
||||
crud_project.save_characters(session, project_id, char_blocks)
|
||||
crud_project.save_storyboard(session, project_id, story_blocks)
|
||||
|
||||
# Consistency
|
||||
consistency = ConsistencyService(session)
|
||||
consistency.normalize_project(project_id)
|
||||
|
||||
task.status = "completed"
|
||||
task.result = {"blocks_found": len(json_blocks)}
|
||||
task.progress = 100
|
||||
session.add(task)
|
||||
session.commit()
|
||||
logger.info(f"Storyboard task {task_id} completed successfully.")
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Storyboard task {task_id} failed: {e}")
|
||||
traceback.print_exc()
|
||||
task.status = "failed"
|
||||
task.message = str(e)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
def generate_all_images_task(task_id: str, project_id: str):
|
||||
logger.info(f"Starting batch image generation task: {task_id} for project: {project_id}")
|
||||
from app.core.database import engine
|
||||
with Session(engine) as session:
|
||||
task = session.get(Task, task_id)
|
||||
if not task:
|
||||
logger.error(f"Task {task_id} not found")
|
||||
return
|
||||
|
||||
task.status = "processing"
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
try:
|
||||
project = session.get(Project, project_id)
|
||||
ai = AIService(session)
|
||||
|
||||
base_dir = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||
static_root = os.path.join(base_dir, "static")
|
||||
if not os.path.exists(static_root): os.makedirs(static_root)
|
||||
|
||||
# 1. Generate Characters
|
||||
total_chars = len(project.characters)
|
||||
logger.info(f"Generating {total_chars} characters...")
|
||||
for i, char in enumerate(project.characters):
|
||||
if char.image_url:
|
||||
logger.info(f"Character {char.name} already has image, skipping.")
|
||||
continue
|
||||
|
||||
logger.info(f"Generating image for character: {char.name}")
|
||||
json_prompt = json.dumps(char.data, ensure_ascii=False, indent=2)
|
||||
json_prompt += "\n\n generate a character design sheet with 4 panels: front view, side view, clothing details, accessories."
|
||||
|
||||
try:
|
||||
image_bytes = ai.generate_image(json_prompt)
|
||||
relative_url = save_generated_image(session, project_id, "character", char.id, image_bytes)
|
||||
char.image_url = relative_url
|
||||
session.add(char)
|
||||
session.commit()
|
||||
logger.info(f"Character {char.name} generated successfully.")
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to generate char {char.id}: {e}")
|
||||
print(f"Failed to generate char {char.id}: {e}")
|
||||
|
||||
# Update task progress (Characters are 20% of work?)
|
||||
# Let's simple split: chars + storyboard items
|
||||
|
||||
# 2. Generate Storyboard Items (Sequential)
|
||||
# Re-fetch items to ensure order
|
||||
items = sorted(project.storyboard_items, key=lambda x: x.sequence)
|
||||
total_items = len(items)
|
||||
logger.info(f"Generating {total_items} storyboard panels...")
|
||||
|
||||
generated_history = [] # Keep track of generated images for context
|
||||
|
||||
# Populate history with existing images
|
||||
# Actually comic_generator logic builds history as it goes.
|
||||
# We should probably load existing images into history if we are resuming?
|
||||
# For "one click", let's assume we scan all items.
|
||||
|
||||
for i, item in enumerate(items):
|
||||
task.progress = int((i / total_items) * 100)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
if item.image_url:
|
||||
# Add to history
|
||||
filename = os.path.basename(item.image_url)
|
||||
# We need to find where it is stored.
|
||||
# Assuming standard structure
|
||||
# We need absolute path for history
|
||||
# item.image_url is like /static/{project_id}/panels/{filename}
|
||||
rel_path = item.image_url.lstrip("/")
|
||||
abs_path = os.path.join(base_dir, rel_path.replace("/", os.sep))
|
||||
|
||||
if os.path.exists(abs_path):
|
||||
generated_history.append(abs_path)
|
||||
continue
|
||||
|
||||
logger.info(f"Generating panel {item.sequence}...")
|
||||
|
||||
# Prepare Context
|
||||
context_images = []
|
||||
|
||||
# a) Character Sheets
|
||||
char_names = item.data.get("characters", [])
|
||||
if isinstance(char_names, str): char_names = [char_names]
|
||||
elif isinstance(char_names, list):
|
||||
names = []
|
||||
for c in char_names:
|
||||
if isinstance(c, dict): names.append(c.get("name", ""))
|
||||
elif isinstance(c, str): names.append(c)
|
||||
char_names = names
|
||||
|
||||
for name in char_names:
|
||||
for p_char in project.characters:
|
||||
if p_char.image_url and (p_char.name in name or name in p_char.name):
|
||||
# Resolve absolute path for char image
|
||||
rel_path = p_char.image_url.lstrip("/")
|
||||
abs_path = os.path.join(base_dir, rel_path.replace("/", os.sep))
|
||||
if os.path.exists(abs_path) and abs_path not in context_images:
|
||||
context_images.append(abs_path)
|
||||
|
||||
# b) Previous History (Last 3 logic)
|
||||
if len(generated_history) >= 3:
|
||||
selected = [generated_history[0]] + generated_history[-2:]
|
||||
else:
|
||||
selected = generated_history
|
||||
|
||||
for path in selected:
|
||||
if path not in context_images:
|
||||
context_images.append(path)
|
||||
|
||||
# Generate
|
||||
json_prompt = json.dumps(item.data, ensure_ascii=False, indent=2)
|
||||
json_prompt += "\n\n use json block as user input prompt to generate 2*2 grid comic image."
|
||||
|
||||
try:
|
||||
image_bytes = ai.generate_image(json_prompt, context_images)
|
||||
relative_url = save_generated_image(session, project_id, "panel", item.id, image_bytes)
|
||||
|
||||
item.image_url = relative_url
|
||||
session.add(item)
|
||||
session.commit()
|
||||
|
||||
# Add to history (absolute path for context usage)
|
||||
# We need absolute path for next context
|
||||
# save_generated_image returns relative /static/...
|
||||
# Reconstruct absolute path
|
||||
# strip leading /
|
||||
abs_path = os.path.join(base_dir, relative_url.lstrip("/").replace("/", os.sep))
|
||||
generated_history.append(abs_path)
|
||||
logger.info(f"Panel {item.sequence} generated successfully.")
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to generate panel {item.id}: {e}")
|
||||
print(f"Failed to generate panel {item.id}: {e}")
|
||||
|
||||
task.status = "completed"
|
||||
task.progress = 100
|
||||
session.add(task)
|
||||
session.commit()
|
||||
logger.info(f"Batch generation task {task_id} completed successfully.")
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Batch generation task {task_id} failed: {e}")
|
||||
traceback.print_exc()
|
||||
task.status = "failed"
|
||||
task.message = str(e)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
def generate_all_characters_task(task_id: str, project_id: str):
|
||||
logger.info(f"Starting batch character generation task: {task_id} for project: {project_id}")
|
||||
from app.core.database import engine
|
||||
with Session(engine) as session:
|
||||
task = session.get(Task, task_id)
|
||||
if not task:
|
||||
logger.error(f"Task {task_id} not found")
|
||||
return
|
||||
|
||||
task.status = "processing"
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
try:
|
||||
project = session.get(Project, project_id)
|
||||
ai = AIService(session)
|
||||
|
||||
total_chars = len(project.characters)
|
||||
logger.info(f"Generating {total_chars} characters...")
|
||||
|
||||
for i, char in enumerate(project.characters):
|
||||
if char.image_url:
|
||||
logger.info(f"Character {char.name} already has image, skipping.")
|
||||
continue
|
||||
|
||||
logger.info(f"Generating image for character: {char.name}")
|
||||
|
||||
# Construct Natural Language Prompt from JSON
|
||||
data = char.data
|
||||
meta = data.get("meta_info", {})
|
||||
name = data.get("name", "Unknown")
|
||||
role = meta.get("role", "")
|
||||
age = meta.get("age", "")
|
||||
personality = data.get("personality", "") or meta.get("personality", "")
|
||||
style = meta.get("style", "")
|
||||
|
||||
# Build Description from panels
|
||||
description = ""
|
||||
panels = data.get("design_panels", [])
|
||||
for p in panels:
|
||||
view = p.get("view", "")
|
||||
desc = p.get("description", "")
|
||||
description += f"- {view}: {desc}\n"
|
||||
|
||||
prompt = f"""Character Design Request:
|
||||
Name: {name}
|
||||
Role: {role}
|
||||
Age: {age}
|
||||
Personality: {personality}
|
||||
Style: {style}
|
||||
|
||||
Visual Description:
|
||||
{description}
|
||||
|
||||
Task: Generate a high-quality character reference sheet (Character Design) based on the above description.
|
||||
Include Front View, Side View, and detailed clothing/accessories.
|
||||
Ensure the character's expression and pose reflect their personality: {personality}.
|
||||
"""
|
||||
|
||||
try:
|
||||
image_bytes = ai.generate_image(prompt)
|
||||
relative_url = save_generated_image(session, project_id, "character", char.id, image_bytes)
|
||||
char.image_url = relative_url
|
||||
session.add(char)
|
||||
session.commit()
|
||||
logger.info(f"Character {char.name} generated successfully.")
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to generate char {char.id}: {e}")
|
||||
|
||||
# Update progress
|
||||
progress = int(((i + 1) / total_chars) * 100)
|
||||
task.progress = progress
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
task.status = "completed"
|
||||
task.progress = 100
|
||||
session.add(task)
|
||||
session.commit()
|
||||
logger.info(f"Batch character generation task {task_id} completed successfully.")
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Batch character generation task {task_id} failed: {e}")
|
||||
traceback.print_exc()
|
||||
task.status = "failed"
|
||||
task.message = str(e)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
def generate_character_task(task_id: str, character_id: int):
|
||||
logger.info(f"Starting character generation task: {task_id} for char: {character_id}")
|
||||
from app.core.database import engine
|
||||
with Session(engine) as session:
|
||||
task = session.get(Task, task_id)
|
||||
if not task:
|
||||
logger.error(f"Task {task_id} not found")
|
||||
return
|
||||
|
||||
task.status = "processing"
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
try:
|
||||
char = session.get(Character, character_id)
|
||||
if not char:
|
||||
raise ValueError("Character not found")
|
||||
|
||||
ai = AIService(session)
|
||||
|
||||
# Construct Natural Language Prompt from JSON
|
||||
data = char.data
|
||||
meta = data.get("meta_info", {})
|
||||
name = data.get("name", "Unknown")
|
||||
role = meta.get("role", "")
|
||||
age = meta.get("age", "")
|
||||
personality = data.get("personality", "") or meta.get("personality", "")
|
||||
style = meta.get("style", "")
|
||||
|
||||
# Build Description from panels
|
||||
description = ""
|
||||
panels = data.get("design_panels", [])
|
||||
for p in panels:
|
||||
view = p.get("view", "")
|
||||
desc = p.get("description", "")
|
||||
description += f"- {view}: {desc}\n"
|
||||
|
||||
prompt = f"""Character Design Request:
|
||||
Name: {name}
|
||||
Role: {role}
|
||||
Age: {age}
|
||||
Personality: {personality}
|
||||
Style: {style}
|
||||
|
||||
Visual Description:
|
||||
{description}
|
||||
|
||||
Task: Generate a high-quality character reference sheet (Character Design) based on the above description.
|
||||
Include Front View, Side View, and detailed clothing/accessories.
|
||||
Ensure the character's expression and pose reflect their personality: {personality}.
|
||||
"""
|
||||
|
||||
logger.info(f"Calling AI service for character {char.name}...")
|
||||
image_bytes = ai.generate_image(prompt)
|
||||
|
||||
relative_url = save_generated_image(session, char.project_id, "character", char.id, image_bytes)
|
||||
|
||||
char.image_url = relative_url
|
||||
session.add(char)
|
||||
|
||||
task.status = "completed"
|
||||
task.progress = 100
|
||||
session.add(task)
|
||||
session.commit()
|
||||
logger.info(f"Character task {task_id} completed successfully.")
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Character task {task_id} failed: {e}")
|
||||
traceback.print_exc()
|
||||
task.status = "failed"
|
||||
task.message = str(e)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
# --- Endpoints ---
|
||||
|
||||
from pydantic import BaseModel
|
||||
|
||||
class StoryboardRequest(BaseModel):
|
||||
user_input: str
|
||||
|
||||
@router.post("/storyboard/{project_id}")
|
||||
def generate_storyboard(
|
||||
project_id: str,
|
||||
request: StoryboardRequest,
|
||||
background_tasks: BackgroundTasks,
|
||||
session: Session = Depends(get_session)
|
||||
):
|
||||
user_input = request.user_input
|
||||
logger.info(f"Received request to generate storyboard for project {project_id}")
|
||||
project = crud_project.get_project(session, project_id)
|
||||
if not project:
|
||||
logger.error(f"Project {project_id} not found")
|
||||
raise HTTPException(status_code=404, detail="Project not found")
|
||||
|
||||
# Save User Input Immediately
|
||||
project.story_input = user_input
|
||||
session.add(project)
|
||||
session.commit()
|
||||
|
||||
# Create Task
|
||||
task = Task(
|
||||
type="storyboard",
|
||||
status="pending",
|
||||
project_id=project_id,
|
||||
name="Generate Storyboard",
|
||||
description=f"Generating storyboard based on user input..."
|
||||
)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
session.refresh(task)
|
||||
logger.info(f"Task created: {task.id}")
|
||||
|
||||
background_tasks.add_task(generate_storyboard_task, task.id, project_id, user_input)
|
||||
|
||||
return {"task_id": task.id}
|
||||
|
||||
@router.post("/all-images/{project_id}")
|
||||
def generate_all_images(
|
||||
project_id: str,
|
||||
background_tasks: BackgroundTasks,
|
||||
session: Session = Depends(get_session)
|
||||
):
|
||||
logger.info(f"Received request to generate all images for project {project_id}")
|
||||
project = crud_project.get_project(session, project_id)
|
||||
if not project:
|
||||
logger.error(f"Project {project_id} not found")
|
||||
raise HTTPException(status_code=404, detail="Project not found")
|
||||
|
||||
task = Task(
|
||||
type="image_generation",
|
||||
status="pending",
|
||||
project_id=project_id,
|
||||
name="Batch Generate Images",
|
||||
description="Generating all storyboard images"
|
||||
)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
session.refresh(task)
|
||||
logger.info(f"Task created: {task.id}")
|
||||
|
||||
background_tasks.add_task(generate_all_images_task, task.id, project_id)
|
||||
|
||||
return {"task_id": task.id}
|
||||
|
||||
# Keep individual endpoints for manual control, but maybe make them async too?
|
||||
# User asked for "Back task" for "generation". Usually implies the bulk actions.
|
||||
# Single panel generation is usually fast enough (5-10s), but can be async if desired.
|
||||
# For now, let's keep single endpoints sync for immediate feedback, or make them async if user insists "All generation".
|
||||
# The prompt says "Generate text/image takes long time".
|
||||
# Let's keep single endpoints sync for simplicity of interaction (user waits 5s is ok),
|
||||
# but "One Click" and "Storyboard" are definitely async.
|
||||
|
||||
@router.post("/all-characters/{project_id}")
|
||||
def generate_all_characters(
|
||||
project_id: str,
|
||||
background_tasks: BackgroundTasks,
|
||||
session: Session = Depends(get_session)
|
||||
):
|
||||
logger.info(f"Received request to generate all characters for project {project_id}")
|
||||
project = crud_project.get_project(session, project_id)
|
||||
if not project:
|
||||
logger.error(f"Project {project_id} not found")
|
||||
raise HTTPException(status_code=404, detail="Project not found")
|
||||
|
||||
task = Task(
|
||||
type="character_generation",
|
||||
status="pending",
|
||||
project_id=project_id,
|
||||
name="Batch Generate Characters",
|
||||
description="Generating all character design sheets"
|
||||
)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
session.refresh(task)
|
||||
logger.info(f"Task created: {task.id}")
|
||||
|
||||
background_tasks.add_task(generate_all_characters_task, task.id, project_id)
|
||||
|
||||
return {"task_id": task.id}
|
||||
|
||||
@router.post("/character/{character_id}")
|
||||
def generate_character(
|
||||
character_id: int,
|
||||
background_tasks: BackgroundTasks,
|
||||
session: Session = Depends(get_session)
|
||||
):
|
||||
char = session.get(Character, character_id)
|
||||
if not char:
|
||||
raise HTTPException(status_code=404, detail="Character not found")
|
||||
|
||||
task = Task(
|
||||
type="character_generation",
|
||||
status="pending",
|
||||
project_id=char.project_id,
|
||||
name=f"Draw Character: {char.name}",
|
||||
description=f"Drawing design sheet for character {char.name}"
|
||||
)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
session.refresh(task)
|
||||
|
||||
background_tasks.add_task(generate_character_task, task.id, character_id)
|
||||
|
||||
return {"task_id": task.id}
|
||||
|
||||
def generate_panel_task(task_id: str, item_id: int):
|
||||
logger.info(f"Starting panel generation task: {task_id} for item: {item_id}")
|
||||
from app.core.database import engine
|
||||
with Session(engine) as session:
|
||||
task = session.get(Task, task_id)
|
||||
if not task:
|
||||
logger.error(f"Task {task_id} not found")
|
||||
return
|
||||
|
||||
task.status = "processing"
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
try:
|
||||
item = session.get(StoryboardItem, item_id)
|
||||
if not item:
|
||||
raise ValueError("Storyboard item not found")
|
||||
|
||||
project = item.project
|
||||
ai = AIService(session)
|
||||
json_prompt = json.dumps(item.data, ensure_ascii=False, indent=2)
|
||||
json_prompt += "\n\n use json block as user input prompt to generate 2*2 grid comic image."
|
||||
|
||||
context_images = []
|
||||
base_dir = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||
|
||||
# 1. Find Character Images
|
||||
char_names = item.data.get("characters", [])
|
||||
if isinstance(char_names, str): char_names = [char_names]
|
||||
elif isinstance(char_names, list):
|
||||
names = []
|
||||
for c in char_names:
|
||||
if isinstance(c, dict): names.append(c.get("name", ""))
|
||||
elif isinstance(c, str): names.append(c)
|
||||
char_names = names
|
||||
|
||||
if project.characters:
|
||||
for name in char_names:
|
||||
for p_char in project.characters:
|
||||
if p_char.image_url and (p_char.name in name or name in p_char.name):
|
||||
rel_path = p_char.image_url.lstrip("/")
|
||||
abs_path = os.path.join(base_dir, rel_path.replace("/", os.sep))
|
||||
if os.path.exists(abs_path) and abs_path not in context_images:
|
||||
context_images.append(abs_path)
|
||||
|
||||
# 2. Previous Panels
|
||||
prev_items = sorted([i for i in project.storyboard_items if i.sequence < item.sequence and i.image_url], key=lambda x: x.sequence)
|
||||
if prev_items:
|
||||
selected = []
|
||||
if len(prev_items) >= 3:
|
||||
selected = [prev_items[0]] + prev_items[-2:]
|
||||
else:
|
||||
selected = prev_items
|
||||
|
||||
for prev in selected:
|
||||
rel_path = prev.image_url.lstrip("/")
|
||||
abs_path = os.path.join(base_dir, rel_path.replace("/", os.sep))
|
||||
if os.path.exists(abs_path) and abs_path not in context_images:
|
||||
context_images.append(abs_path)
|
||||
|
||||
# Style Consistency
|
||||
meta_style = item.data.get("meta_info", {}).get("style", "")
|
||||
if not meta_style and project.global_config:
|
||||
meta_style = project.global_config.data.get("style", "")
|
||||
|
||||
if meta_style:
|
||||
json_prompt += f"\n\nStyle Consistency Requirement: {meta_style}. Ensure the visual style matches the provided context images."
|
||||
|
||||
logger.info(f"Calling AI service for panel {item.sequence}...")
|
||||
image_bytes = ai.generate_image(json_prompt, context_images)
|
||||
|
||||
relative_url = save_generated_image(session, project.id, "panel", item.id, image_bytes)
|
||||
item.image_url = relative_url
|
||||
session.add(item)
|
||||
|
||||
task.status = "completed"
|
||||
task.progress = 100
|
||||
session.add(task)
|
||||
session.commit()
|
||||
logger.info(f"Panel task {task_id} completed successfully.")
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Panel task {task_id} failed: {e}")
|
||||
traceback.print_exc()
|
||||
task.status = "failed"
|
||||
task.message = str(e)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
|
||||
@router.post("/panel/{item_id}")
|
||||
def generate_panel(
|
||||
item_id: int,
|
||||
background_tasks: BackgroundTasks,
|
||||
session: Session = Depends(get_session)
|
||||
):
|
||||
item = session.get(StoryboardItem, item_id)
|
||||
if not item:
|
||||
raise HTTPException(status_code=404, detail="Storyboard item not found")
|
||||
|
||||
task = Task(
|
||||
type="image_generation",
|
||||
status="pending",
|
||||
project_id=item.project_id,
|
||||
name=f"Draw Panel: #{item.sequence}",
|
||||
description=f"Drawing panel {item.sequence}"
|
||||
)
|
||||
session.add(task)
|
||||
session.commit()
|
||||
session.refresh(task)
|
||||
|
||||
background_tasks.add_task(generate_panel_task, task.id, item_id)
|
||||
|
||||
return {"task_id": task.id}
|
||||
@@ -0,0 +1,38 @@
|
||||
from fastapi import APIRouter, Depends, HTTPException
|
||||
from sqlmodel import Session, select
|
||||
from app.core.database import get_session
|
||||
from app.models.models import ImageHistory, Character, StoryboardItem
|
||||
|
||||
router = APIRouter()
|
||||
|
||||
@router.get("/{entity_type}/{entity_id}")
|
||||
def get_history(entity_type: str, entity_id: int, session: Session = Depends(get_session)):
|
||||
# entity_type: 'character' or 'panel'
|
||||
statement = select(ImageHistory).where(
|
||||
ImageHistory.entity_type == entity_type,
|
||||
ImageHistory.entity_id == entity_id
|
||||
).order_by(ImageHistory.created_at.desc())
|
||||
history = session.exec(statement).all()
|
||||
return history
|
||||
|
||||
@router.post("/select/{history_id}")
|
||||
def select_image(history_id: int, session: Session = Depends(get_session)):
|
||||
history = session.get(ImageHistory, history_id)
|
||||
if not history:
|
||||
raise HTTPException(status_code=404, detail="History item not found")
|
||||
|
||||
if history.entity_type == "character":
|
||||
entity = session.get(Character, history.entity_id)
|
||||
elif history.entity_type == "panel":
|
||||
entity = session.get(StoryboardItem, history.entity_id)
|
||||
else:
|
||||
raise HTTPException(status_code=400, detail="Unknown entity type")
|
||||
|
||||
if not entity:
|
||||
raise HTTPException(status_code=404, detail="Entity not found")
|
||||
|
||||
entity.image_url = history.image_url
|
||||
session.add(entity)
|
||||
session.commit()
|
||||
session.refresh(entity)
|
||||
return {"status": "success", "image_url": entity.image_url}
|
||||
@@ -0,0 +1,205 @@
|
||||
from fastapi import APIRouter, Depends, HTTPException, Body
|
||||
from sqlmodel import Session
|
||||
from typing import List, Dict
|
||||
from pydantic import BaseModel
|
||||
from app.core.database import get_session
|
||||
from app.models.models import Project, GlobalConfig, Character, StoryboardItem
|
||||
from app.schemas.schemas import ProjectCreate, ProjectUpdate, ProjectRead
|
||||
from app.cruds import crud_project
|
||||
from app.services.consistency_service import ConsistencyService
|
||||
|
||||
router = APIRouter()
|
||||
|
||||
@router.post("/", response_model=Project)
|
||||
def create_project(project_in: ProjectCreate, session: Session = Depends(get_session)):
|
||||
return crud_project.create_project(session, project_in)
|
||||
|
||||
@router.get("/", response_model=List[Project])
|
||||
def read_projects(skip: int = 0, limit: int = 100, session: Session = Depends(get_session)):
|
||||
return crud_project.get_projects(session, skip, limit)
|
||||
|
||||
@router.get("/{project_id}", response_model=ProjectRead)
|
||||
def read_project(project_id: str, session: Session = Depends(get_session)):
|
||||
project = crud_project.get_project(session, project_id)
|
||||
if not project:
|
||||
raise HTTPException(status_code=404, detail="Project not found")
|
||||
return project
|
||||
|
||||
@router.put("/{project_id}", response_model=Project)
|
||||
def update_project(project_id: str, project_in: ProjectUpdate, session: Session = Depends(get_session)):
|
||||
project = crud_project.get_project(session, project_id)
|
||||
if not project:
|
||||
raise HTTPException(status_code=404, detail="Project not found")
|
||||
return crud_project.update_project(session, project, project_in)
|
||||
|
||||
@router.delete("/{project_id}")
|
||||
def delete_project(project_id: str, session: Session = Depends(get_session)):
|
||||
project = crud_project.get_project(session, project_id)
|
||||
if not project:
|
||||
raise HTTPException(status_code=404, detail="Project not found")
|
||||
crud_project.delete_project(session, project)
|
||||
return {"ok": True}
|
||||
|
||||
# --- Data Management Endpoints ---
|
||||
|
||||
@router.put("/{project_id}/global_config")
|
||||
def update_global_config(project_id: str, data: Dict = Body(...), session: Session = Depends(get_session)):
|
||||
project = crud_project.get_project(session, project_id)
|
||||
if not project:
|
||||
raise HTTPException(status_code=404, detail="Project not found")
|
||||
|
||||
# 1. Update Project level fields first
|
||||
# This ensures consistency between project.theme/language and global_config
|
||||
if "language" in data:
|
||||
project.language = data["language"]
|
||||
if "style" in data:
|
||||
project.theme = data["style"]
|
||||
if "aspect_ratio" in data:
|
||||
project.aspect_ratio = data["aspect_ratio"]
|
||||
|
||||
session.add(project)
|
||||
session.commit()
|
||||
session.refresh(project)
|
||||
|
||||
# 2. Update GlobalConfig in DB
|
||||
config = crud_project.create_global_config(session, project_id, data)
|
||||
|
||||
# 3. Trigger consistency check (Propagate to all items)
|
||||
consistency = ConsistencyService(session)
|
||||
consistency.normalize_project(project_id)
|
||||
|
||||
return config
|
||||
|
||||
@router.put("/{project_id}/characters/{char_id}")
|
||||
def update_character(project_id: str, char_id: int, data: Dict = Body(...), session: Session = Depends(get_session)):
|
||||
char = session.get(Character, char_id)
|
||||
if not char:
|
||||
raise HTTPException(status_code=404, detail="Character not found")
|
||||
|
||||
char.data = data
|
||||
session.add(char)
|
||||
session.commit()
|
||||
session.refresh(char)
|
||||
|
||||
# We might want to trigger consistency here too if character style changes,
|
||||
# but primarily it's driven by global config.
|
||||
return char
|
||||
|
||||
@router.put("/{project_id}/storyboard/{item_id}")
|
||||
def update_storyboard_item(project_id: str, item_id: int, data: Dict = Body(...), session: Session = Depends(get_session)):
|
||||
item = session.get(StoryboardItem, item_id)
|
||||
if not item:
|
||||
raise HTTPException(status_code=404, detail="Storyboard Item not found")
|
||||
|
||||
item.data = data
|
||||
session.add(item)
|
||||
session.commit()
|
||||
session.refresh(item)
|
||||
return item
|
||||
|
||||
@router.delete("/{project_id}/characters/{char_id}")
|
||||
def delete_character(project_id: str, char_id: int, session: Session = Depends(get_session)):
|
||||
char = session.get(Character, char_id)
|
||||
if not char:
|
||||
raise HTTPException(status_code=404, detail="Character not found")
|
||||
|
||||
session.delete(char)
|
||||
session.commit()
|
||||
return {"ok": True}
|
||||
|
||||
class MergeCharacterRequest(BaseModel):
|
||||
target_char_id: int
|
||||
source_char_ids: List[int]
|
||||
|
||||
@router.post("/{project_id}/characters/merge")
|
||||
def merge_characters(
|
||||
project_id: str,
|
||||
request: MergeCharacterRequest,
|
||||
session: Session = Depends(get_session)
|
||||
):
|
||||
target_char = session.get(Character, request.target_char_id)
|
||||
if not target_char:
|
||||
raise HTTPException(status_code=404, detail="Target character not found")
|
||||
|
||||
source_chars = []
|
||||
for cid in request.source_char_ids:
|
||||
c = session.get(Character, cid)
|
||||
if c:
|
||||
source_chars.append(c)
|
||||
|
||||
if not source_chars:
|
||||
raise HTTPException(status_code=400, detail="No valid source characters found")
|
||||
|
||||
target_name = target_char.name
|
||||
source_names = [c.name for c in source_chars]
|
||||
|
||||
# 1. Update Storyboard Items
|
||||
# We need to scan all items and replace source names with target name
|
||||
project = session.get(Project, project_id)
|
||||
if project.storyboard_items:
|
||||
for item in project.storyboard_items:
|
||||
data = item.data
|
||||
# 'characters' field in storyboard item data
|
||||
# It can be a list of strings, or list of dicts with 'name' key, or a single string
|
||||
chars = data.get("characters", [])
|
||||
|
||||
new_chars = []
|
||||
modified = False
|
||||
|
||||
# Helper to normalize input to list
|
||||
char_list = []
|
||||
if isinstance(chars, str): char_list = [chars]
|
||||
elif isinstance(chars, list): char_list = chars
|
||||
|
||||
for c_entry in char_list:
|
||||
c_name = ""
|
||||
if isinstance(c_entry, str): c_name = c_entry
|
||||
elif isinstance(c_entry, dict): c_name = c_entry.get("name", "")
|
||||
|
||||
# Check if this name matches any source name
|
||||
# Fuzzy match or exact? Let's do exact or containment for safety
|
||||
# User said "Ma Laoguanjia" vs "Ma Guanjia".
|
||||
# Ideally we replace if it matches one of the source characters' name EXACTLY or close enough?
|
||||
# Since we selected source characters by ID, we know their names.
|
||||
# Let's replace if the name in storyboard matches a source character name.
|
||||
|
||||
is_source = False
|
||||
for src_name in source_names:
|
||||
if src_name == c_name:
|
||||
is_source = True
|
||||
break
|
||||
|
||||
if is_source:
|
||||
# Replace with target name
|
||||
# If entry was dict, update name field? Or just use string?
|
||||
# Let's keep format.
|
||||
if isinstance(c_entry, str):
|
||||
new_chars.append(target_name)
|
||||
elif isinstance(c_entry, dict):
|
||||
c_entry['name'] = target_name
|
||||
new_chars.append(c_entry)
|
||||
modified = True
|
||||
else:
|
||||
new_chars.append(c_entry)
|
||||
|
||||
if modified:
|
||||
# Deduplicate if target name already existed?
|
||||
# Simple dedup for strings
|
||||
final_chars = []
|
||||
seen = set()
|
||||
for c in new_chars:
|
||||
n = c if isinstance(c, str) else c.get("name", "")
|
||||
if n not in seen:
|
||||
final_chars.append(c)
|
||||
seen.add(n)
|
||||
|
||||
data['characters'] = final_chars
|
||||
item.data = data
|
||||
session.add(item)
|
||||
|
||||
# 2. Delete Source Characters
|
||||
for c in source_chars:
|
||||
session.delete(c)
|
||||
|
||||
session.commit()
|
||||
return {"ok": True, "merged_count": len(source_chars)}
|
||||
@@ -0,0 +1,23 @@
|
||||
from fastapi import APIRouter, Depends, HTTPException
|
||||
from sqlmodel import Session
|
||||
from app.core.database import get_session
|
||||
from app.models.models import Task
|
||||
from app.schemas.schemas import TaskRead
|
||||
|
||||
router = APIRouter()
|
||||
|
||||
@router.get("/{task_id}", response_model=TaskRead)
|
||||
def get_task_status(task_id: str, session: Session = Depends(get_session)):
|
||||
task = session.get(Task, task_id)
|
||||
if not task:
|
||||
raise HTTPException(status_code=404, detail="Task not found")
|
||||
return task
|
||||
|
||||
@router.get("/project/{project_id}", response_model=list[TaskRead])
|
||||
def get_project_tasks(project_id: str, session: Session = Depends(get_session)):
|
||||
from sqlmodel import select
|
||||
statement = select(Task).where(Task.project_id == project_id).order_by(Task.created_at.desc())
|
||||
tasks = session.exec(statement).all()
|
||||
# Filter only recent or active tasks if list is too long?
|
||||
# For now return all, maybe limit 20
|
||||
return tasks[:20]
|
||||
@@ -0,0 +1,59 @@
|
||||
from pydantic import BaseModel
|
||||
from typing import Optional, Dict, List
|
||||
from datetime import datetime
|
||||
from app.models.models import (
|
||||
ModelConfigBase, ProjectBase, CharacterBase, StoryboardItemBase, GlobalConfigBase, TaskBase,
|
||||
ModelConfig, Project, Character, StoryboardItem, GlobalConfig, Task
|
||||
)
|
||||
|
||||
# ModelConfig
|
||||
class ModelConfigCreate(ModelConfigBase):
|
||||
pass
|
||||
|
||||
class ModelConfigUpdate(BaseModel):
|
||||
provider: Optional[str] = None
|
||||
api_key: Optional[str] = None
|
||||
base_url: Optional[str] = None
|
||||
model_name: Optional[str] = None
|
||||
model_type: Optional[str] = None
|
||||
is_active: Optional[bool] = None
|
||||
|
||||
# Read Models for nested response
|
||||
class CharacterRead(CharacterBase):
|
||||
id: int
|
||||
project_id: str
|
||||
|
||||
class StoryboardItemRead(StoryboardItemBase):
|
||||
id: int
|
||||
project_id: str
|
||||
|
||||
class GlobalConfigRead(GlobalConfigBase):
|
||||
id: int
|
||||
project_id: str
|
||||
|
||||
class TaskRead(TaskBase):
|
||||
id: str
|
||||
project_id: str
|
||||
created_at: datetime
|
||||
updated_at: datetime
|
||||
|
||||
# Project
|
||||
class ProjectCreate(ProjectBase):
|
||||
pass
|
||||
|
||||
class ProjectUpdate(BaseModel):
|
||||
title: Optional[str] = None
|
||||
description: Optional[str] = None
|
||||
story_input: Optional[str] = None
|
||||
theme: Optional[str] = None
|
||||
language: Optional[str] = None
|
||||
panel_count: Optional[int] = None
|
||||
aspect_ratio: Optional[str] = None
|
||||
|
||||
class ProjectRead(ProjectBase):
|
||||
id: str
|
||||
created_at: datetime
|
||||
updated_at: datetime
|
||||
characters: List[CharacterRead] = []
|
||||
storyboard_items: List[StoryboardItemRead] = []
|
||||
global_config: Optional[GlobalConfigRead] = None
|
||||
@@ -0,0 +1,92 @@
|
||||
import os
|
||||
import time
|
||||
from typing import List, Optional
|
||||
from sqlmodel import Session
|
||||
from app.models.models import ModelConfig
|
||||
from google import genai
|
||||
from PIL import Image
|
||||
import io
|
||||
|
||||
class AIService:
|
||||
def __init__(self, session: Session):
|
||||
self.session = session
|
||||
|
||||
def _get_client(self, model_type: str):
|
||||
from app.cruds.crud_config import get_active_config
|
||||
config = get_active_config(self.session, model_type)
|
||||
|
||||
if not config:
|
||||
raise ValueError(f"No active configuration found for {model_type} model.")
|
||||
|
||||
if config.provider.lower() == "google":
|
||||
# Initialize Google Client
|
||||
return genai.Client(api_key=config.api_key), config.model_name
|
||||
|
||||
raise NotImplementedError(f"Provider {config.provider} not supported yet.")
|
||||
|
||||
def generate_storyboard(self, system_prompt: str, user_input: str) -> str:
|
||||
client, model_name = self._get_client("text")
|
||||
|
||||
full_prompt = f"{system_prompt}\n\nUser Input: {user_input}\n\nPlease generate the full storyboard in JSON format as requested."
|
||||
|
||||
try:
|
||||
response = client.models.generate_content(
|
||||
model=model_name,
|
||||
contents=full_prompt
|
||||
)
|
||||
return response.text
|
||||
except Exception as e:
|
||||
print(f"Error generating storyboard: {e}")
|
||||
raise e
|
||||
|
||||
def generate_image(self, prompt: str, context_images: List[str] = None) -> bytes:
|
||||
client, model_name = self._get_client("image")
|
||||
|
||||
contents = [prompt]
|
||||
if context_images:
|
||||
for img_path in context_images:
|
||||
if os.path.exists(img_path):
|
||||
try:
|
||||
prev_img = Image.open(img_path)
|
||||
contents.append(prev_img)
|
||||
except Exception as e:
|
||||
print(f"Failed to load context image {img_path}: {e}")
|
||||
else:
|
||||
# Log missing context image but don't fail, just skip it
|
||||
print(f"Warning: Context image not found at {img_path}, skipping.")
|
||||
|
||||
# Retry loop
|
||||
max_retries = 3
|
||||
for attempt in range(max_retries):
|
||||
try:
|
||||
response = client.models.generate_content(
|
||||
model=model_name,
|
||||
contents=contents,
|
||||
)
|
||||
|
||||
if response.parts:
|
||||
for part in response.parts:
|
||||
if part.inline_data is not None:
|
||||
image_data = part.inline_data.data
|
||||
if len(image_data) > 0:
|
||||
return image_data
|
||||
else:
|
||||
print(f"Warning: Received empty image data on attempt {attempt + 1}")
|
||||
|
||||
# Check for text refusal/error
|
||||
if response.text:
|
||||
print(f"Model response text (no image): {response.text}")
|
||||
|
||||
print(f"Attempt {attempt + 1} failed: No valid image data found in response.")
|
||||
if attempt == max_retries - 1:
|
||||
raise ValueError(f"No image found in response after {max_retries} retries. Last response: {response.text if response.text else 'Empty'}")
|
||||
|
||||
time.sleep(2 ** attempt)
|
||||
|
||||
except Exception as e:
|
||||
print(f"Error generating image (Attempt {attempt + 1}/{max_retries}): {e}")
|
||||
if attempt < max_retries - 1:
|
||||
time.sleep(2 ** attempt)
|
||||
continue
|
||||
raise e
|
||||
return b""
|
||||
@@ -0,0 +1,181 @@
|
||||
import re
|
||||
import json
|
||||
import copy
|
||||
from typing import List, Dict, Any
|
||||
from sqlmodel import Session, select
|
||||
from app.models.models import Project, Character, StoryboardItem, GlobalConfig
|
||||
|
||||
class ConsistencyService:
|
||||
def __init__(self, session: Session):
|
||||
self.session = session
|
||||
|
||||
def normalize_project(self, project_id: str):
|
||||
"""
|
||||
Normalizes the project's data (characters, storyboard) based on the global config.
|
||||
This mirrors the logic in comic_generator.py.
|
||||
"""
|
||||
project = self.session.get(Project, project_id)
|
||||
if not project:
|
||||
return
|
||||
|
||||
# Fetch all related data
|
||||
# Note: relationships are loaded if accessed, but let's be explicit if needed.
|
||||
# ProjectRead should handle loading, but here we work with ORM objects.
|
||||
|
||||
global_config = project.global_config
|
||||
characters = project.characters
|
||||
storyboard_items = sorted(project.storyboard_items, key=lambda x: x.sequence)
|
||||
|
||||
master_style = None
|
||||
master_meta = {}
|
||||
|
||||
# Priority 0: Comic Config
|
||||
if global_config and global_config.data:
|
||||
config_data = global_config.data
|
||||
master_style = config_data.get("style")
|
||||
|
||||
# Explicitly map all fields we want to sync
|
||||
master_meta = {
|
||||
"style": master_style,
|
||||
"bubble_style": config_data.get("bubble_style"),
|
||||
"narration_style": config_data.get("narration_style"),
|
||||
"border_style": config_data.get("border_style"),
|
||||
"gutter_style": config_data.get("gutter_style"),
|
||||
"layout_settings": config_data.get("layout_settings"),
|
||||
"aspect_ratio": config_data.get("aspect_ratio", "16:9"),
|
||||
"language": config_data.get("language", "English")
|
||||
}
|
||||
|
||||
# Priority 1: First Character Sheet (if no config style)
|
||||
if not master_style and characters:
|
||||
# Try to find one with style
|
||||
for char in characters:
|
||||
if char.data.get("style"):
|
||||
master_style = char.data.get("style")
|
||||
break
|
||||
|
||||
# Priority 2: First Story Block
|
||||
if storyboard_items:
|
||||
first_item = storyboard_items[0]
|
||||
first_meta = first_item.data.get("meta_info", {})
|
||||
if not master_style:
|
||||
master_style = first_meta.get("style")
|
||||
|
||||
if not master_meta:
|
||||
master_meta = first_meta.copy()
|
||||
|
||||
# Build Character Registry
|
||||
known_characters = {}
|
||||
for char in characters:
|
||||
full_name = char.name
|
||||
if not full_name: continue
|
||||
|
||||
keywords = [full_name]
|
||||
simplified = re.split(r'[（\(]', full_name)[0].strip()
|
||||
if simplified and simplified != full_name:
|
||||
keywords.append(simplified)
|
||||
|
||||
for kw in keywords:
|
||||
if kw:
|
||||
known_characters[kw] = full_name
|
||||
|
||||
# Apply Normalization
|
||||
# Even if master_style is None, we might still have layout_settings to sync
|
||||
# So we check if we have ANY master_meta to apply
|
||||
if master_meta:
|
||||
import logging
|
||||
logger = logging.getLogger(__name__)
|
||||
logger.info(f"Normalizing project {project_id} with master config: {json.dumps(master_meta, ensure_ascii=False)}")
|
||||
|
||||
# 1. Normalize Characters
|
||||
for char in characters:
|
||||
char_data = copy.deepcopy(char.data)
|
||||
if not isinstance(char_data, dict):
|
||||
char_data = dict(char_data)
|
||||
|
||||
meta = char_data.get("meta_info", {})
|
||||
|
||||
# Sync Core Fields
|
||||
if master_meta.get("style"): meta["style"] = master_meta.get("style")
|
||||
if master_meta.get("language"): meta["language"] = master_meta.get("language")
|
||||
if master_meta.get("aspect_ratio"): meta["aspect_ratio"] = master_meta.get("aspect_ratio")
|
||||
|
||||
char_data["meta_info"] = meta
|
||||
|
||||
# Cleanup top-level legacy fields
|
||||
char_data.pop("style", None)
|
||||
char_data.pop("language", None)
|
||||
char_data.pop("Language", None)
|
||||
|
||||
char.data = char_data
|
||||
self.session.add(char)
|
||||
|
||||
# 2. Normalize Storyboard Items
|
||||
total_volumes = len(storyboard_items)
|
||||
|
||||
for i, item in enumerate(storyboard_items):
|
||||
# Use deepcopy to ensure we don't mutate the original object in place before assignment
|
||||
# and to ensure SQLAlchemy detects the change when we reassign.
|
||||
item_data = copy.deepcopy(item.data)
|
||||
if not isinstance(item_data, dict):
|
||||
item_data = dict(item_data)
|
||||
|
||||
# Check for missing characters
|
||||
plot_text = json.dumps(item_data.get("plot_breakdown", []), ensure_ascii=False)
|
||||
current_chars = item_data.get("characters", [])
|
||||
if isinstance(current_chars, str):
|
||||
current_chars = [current_chars]
|
||||
if not isinstance(current_chars, list):
|
||||
current_chars = []
|
||||
|
||||
found_missing = []
|
||||
for kw, full_name in known_characters.items():
|
||||
if kw in plot_text:
|
||||
is_present = False
|
||||
for char_name in current_chars:
|
||||
if kw in char_name or char_name in full_name:
|
||||
is_present = True
|
||||
break
|
||||
if not is_present:
|
||||
if full_name not in current_chars and full_name not in found_missing:
|
||||
found_missing.append(full_name)
|
||||
|
||||
if found_missing:
|
||||
current_chars.extend(found_missing)
|
||||
item_data["characters"] = current_chars
|
||||
|
||||
# Sync Meta Info
|
||||
meta = item_data.get("meta_info", {})
|
||||
|
||||
# We want to preserve existing fields in meta (like volume) but overwrite style configs
|
||||
# master_meta has style, bubble_style, etc.
|
||||
|
||||
# Force update fields from master_meta even if they exist in meta
|
||||
# But careful with 'None' values in master_meta (though we constructed it from config)
|
||||
|
||||
for key, value in master_meta.items():
|
||||
if key == "volume": continue
|
||||
|
||||
# Special handling for boolean values (like False in show_panel_numbers)
|
||||
# "if value is not None" is correct for booleans.
|
||||
# For nested dictionaries (like layout_settings), we should merge to preserve other keys.
|
||||
|
||||
if value is not None:
|
||||
if isinstance(value, dict) and isinstance(meta.get(key), dict):
|
||||
meta[key].update(value)
|
||||
else:
|
||||
meta[key] = value
|
||||
|
||||
# If value is None but key exists in master_meta keys (explicitly set to null), we might want to unset it?
|
||||
# But our master_meta construction uses .get() defaults, so None usually means "not in config".
|
||||
|
||||
# Ensure Volume format
|
||||
meta["volume"] = f"{i+1}/{total_volumes}"
|
||||
item_data["meta_info"] = meta
|
||||
|
||||
logger.info(f"Updated Item {i+1} meta: {json.dumps(meta, ensure_ascii=False)}")
|
||||
|
||||
item.data = item_data
|
||||
self.session.add(item)
|
||||
|
||||
self.session.commit()
|
||||
@@ -0,0 +1,30 @@
|
||||
import os
|
||||
from PIL import Image
|
||||
import io
|
||||
|
||||
def split_comic_page(image_bytes: bytes) -> list[bytes]:
|
||||
"""Splits a 2x2 grid comic page into 4 individual panel images (bytes)."""
|
||||
try:
|
||||
img = Image.open(io.BytesIO(image_bytes))
|
||||
width, height = img.size
|
||||
mid_w = width // 2
|
||||
mid_h = height // 2
|
||||
|
||||
quadrants = [
|
||||
(0, 0, mid_w, mid_h), # Top-Left
|
||||
(mid_w, 0, width, mid_h), # Top-Right
|
||||
(0, mid_h, mid_w, height), # Bottom-Left
|
||||
(mid_w, mid_h, width, height) # Bottom-Right
|
||||
]
|
||||
|
||||
panels = []
|
||||
for box in quadrants:
|
||||
panel = img.crop(box)
|
||||
buf = io.BytesIO()
|
||||
panel.save(buf, format="PNG")
|
||||
panels.append(buf.getvalue())
|
||||
|
||||
return panels
|
||||
except Exception as e:
|
||||
print(f"Failed to split image: {e}")
|
||||
return []
|
||||
@@ -0,0 +1,72 @@
|
||||
import json
|
||||
import re
|
||||
from typing import List, Dict, Any
|
||||
|
||||
def extract_json_blocks(text: str) -> List[Dict[str, Any]]:
|
||||
"""Extracts JSON blocks from the generated text."""
|
||||
json_blocks = []
|
||||
|
||||
# 1. Try to find ```json ... ``` blocks
|
||||
pattern = r"```(?:json|JSON)?\s*(.*?)\s*```"
|
||||
matches = re.findall(pattern, text, re.DOTALL)
|
||||
|
||||
# 2. If no code blocks found, or even if found, we should also look for raw JSON objects
|
||||
# because sometimes models output mixed content.
|
||||
# But let's stick to code blocks first if they exist.
|
||||
|
||||
if not matches:
|
||||
# Fallback: Try to find top-level JSON objects/arrays directly in text
|
||||
# This regex looks for { ... } or [ ... ] that span multiple lines
|
||||
# It's not perfect but better than nothing
|
||||
# We search for anything starting with { or [ and ending with } or ]
|
||||
# non-greedy match might be safer for multiple blocks
|
||||
raw_pattern = r"(\{[\s\S]*?\}|\[[\s\S]*?\])"
|
||||
matches = re.findall(raw_pattern, text)
|
||||
|
||||
def repair_json(json_str: str) -> Dict[str, Any]:
|
||||
"""Attempts to repair common JSON errors, specifically missing commas."""
|
||||
try:
|
||||
return json.loads(json_str)
|
||||
except json.JSONDecodeError:
|
||||
pass
|
||||
|
||||
repaired_str = json_str
|
||||
max_attempts = 10
|
||||
|
||||
for _ in range(max_attempts):
|
||||
try:
|
||||
return json.loads(repaired_str)
|
||||
except json.JSONDecodeError as e:
|
||||
# print(f"JSON Decode Error at {e.pos}: {e.msg}")
|
||||
if "Expecting ',' delimiter" in str(e) or "Expecting property name enclosed in double quotes" in str(e):
|
||||
pos = e.pos
|
||||
search_str = repaired_str[:pos]
|
||||
match = re.search(r'([\"}\]0-9])\s*$', search_str)
|
||||
|
||||
if match:
|
||||
insert_idx = match.end()
|
||||
repaired_str = repaired_str[:insert_idx] + "," + repaired_str[insert_idx:]
|
||||
continue
|
||||
|
||||
# If we can't fix it, re-raise
|
||||
raise e
|
||||
|
||||
return json.loads(repaired_str)
|
||||
|
||||
for match in matches:
|
||||
if not match.strip(): continue
|
||||
# Simple heuristic to filter out non-json text blocks that might be caught by raw regex
|
||||
if not (match.strip().startswith('{') or match.strip().startswith('[')):
|
||||
continue
|
||||
|
||||
try:
|
||||
data = repair_json(match)
|
||||
if isinstance(data, list):
|
||||
json_blocks.extend(data)
|
||||
elif isinstance(data, dict):
|
||||
json_blocks.append(data)
|
||||
except Exception as e:
|
||||
print(f"Failed to parse a JSON block: {e}")
|
||||
# print(f"Block content: {match[:100]}...")
|
||||
|
||||
return json_blocks
|
||||
Binary file not shown.
Binary file not shown.
|
After Width: | Height: | Size: 822 KiB |
@@ -0,0 +1,12 @@
|
||||
fastapi
|
||||
uvicorn
|
||||
sqlmodel
|
||||
pydantic
|
||||
pydantic-settings
|
||||
python-dotenv
|
||||
google-genai
|
||||
openai
|
||||
requests
|
||||
alembic
|
||||
pymysql
|
||||
aiofiles
|
||||
BIN
Binary file not shown.
|
After Width: | Height: | Size: 869 KiB |
BIN
Binary file not shown.
|
After Width: | Height: | Size: 724 KiB |
BIN
Binary file not shown.
|
After Width: | Height: | Size: 685 KiB |
BIN
Binary file not shown.
|
After Width: | Height: | Size: 806 KiB |
BIN
Binary file not shown.
|
After Width: | Height: | Size: 689 KiB |
BIN
Binary file not shown.
|
After Width: | Height: | Size: 700 KiB |
BIN
Binary file not shown.
|
After Width: | Height: | Size: 933 KiB |
BIN
Binary file not shown.
|
After Width: | Height: | Size: 689 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 933 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 869 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 685 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 724 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 689 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 689 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 948 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 911 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 806 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 947 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 953 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 861 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 1.0 MiB |
Binary file not shown.
|
After Width: | Height: | Size: 437 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 435 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 434 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 412 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 459 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 448 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 437 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 400 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 414 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 427 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 430 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 332 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 424 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 432 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 464 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 427 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 421 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 421 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 481 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 423 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 414 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 438 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 364 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 392 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 496 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 441 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 481 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 431 KiB |
Binary file not shown.
Binary file not shown.
|
After Width: | Height: | Size: 948 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 911 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 806 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 947 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 934 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 953 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 955 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 934 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 1.0 MiB |
Binary file not shown.
|
After Width: | Height: | Size: 961 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 861 KiB |
Some files were not shown because too many files have changed in this diff Show More
Reference in New Issue
Block a user