Back to Details

uv-python-workflow

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

uv Python Workflow

uv Python 工作流

All Python dependency and environment management in this codebase uses 
uv
 exclusively.
本代码库中所有Python依赖与环境管理均仅使用
uv
。

Package management

包管理

  • Add dependencies: 
    uv add <package>
  • Remove dependencies: 
    uv remove <package>
  • Install/sync the current project: 
    uv sync
  • Run a command inside the project environment: 
    uv run <command>
  • Never use 
    pip install
    , 
    pip-compile
    , or bare 
    python -m venv
    .
  • 添加依赖：
    uv add <package>
  • 移除依赖：
    uv remove <package>
  • 安装/同步当前项目：
    uv sync
  • 在项目环境中运行命令：
    uv run <command>
  • 切勿使用
    pip install
    、
    pip-compile
    或直接使用
    python -m venv
    。

Workspace layout

工作区布局

Projects may be organized as a 
uv
workspace. Code locations managed by 
dg
do not use 
uv
workspaces — they rely on the 
dg
workspace instead.
项目可组织为
uv
工作区。由
dg
管理的代码位置不使用
uv
工作区——它们依赖于
dg
工作区。

Docker / container images

Docker / 容器镜像

When building a Python app inside a Docker image, create a relocatable virtual environment so it works after the image layers are assembled:
dockerfile
RUN uv venv --relocatable /app/.venv
RUN uv sync --frozen --no-dev
Do not mount 
ol-orchestrate-lib
(or any shared library) as a volume at runtime. Shared libraries are installed as build-time dependencies only.
在Docker镜像中构建Python应用时，创建一个可重定位的虚拟环境，以便在镜像层组装后仍能正常工作：
dockerfile
RUN uv venv --relocatable /app/.venv
RUN uv sync --frozen --no-dev
运行时不要将
ol-orchestrate-lib
（或任何共享库）作为卷挂载。共享库仅作为构建时依赖安装。

Running pre-commit and type checks

运行预提交与类型检查

bash
uv run pre-commit run --all-files
uv run mypy <package>
bash
uv run pre-commit run --all-files
uv run mypy <package>