安装可编辑Poetry包时Pip因依赖版本冲突失败（版本要求一致）

问题描述 #

使用pip安装带可编辑模式依赖的Poetry项目时，明明依赖版本完全一致，却触发版本解析冲突。错误核心提示：

The conflict is caused by:
The user requested second-library 0.1.0 (from libs/second_library)
first-library 0.1.0 depends on second-library 0.1.0 (from libs/second_library)

环境细节 #

• 文件夹结构：

  requirements.txt
  libs/first_library
  libs/second_library
  

• sim_requirements.txt内容：

  -e ./libs/first_library
  -e ./libs/second_library
  

• first-library的pyproject.toml依赖配置：

  [tool.poetry.dependencies]
  python = "^3.9"
  second-library = {path = "../second_library", develop = true}
  

• 完整错误日志：

  INFO: pip is looking at multiple versions of first-library to determine which version is compatible with other requirements. This could take a while.
  ERROR: Cannot install -r sim_requirements.txt (line 1) and second-library 0.1.0 (from libs/second_library) because these package versions have conflicting dependencies.
  
  The conflict is caused by:
      The user requested second-library 0.1.0 (from libs/second_library)
      first-library 0.1.0 depends on second-library 0.1.0 (from libs/second_library)
  
  To fix this you could try to:
  1. loosen the range of package versions you've specified
  2. remove package versions to allow pip attempt to solve the dependency conflict
  

问题根源 #

当同一个本地包同时通过两种方式被引用时：一是requirements类文件用-e可编辑模式声明，二是Poetry的pyproject.toml用path+develop=true声明，pip会把这两个引用判定为不同的包实例——哪怕它们指向同一个本地目录、版本完全一致，也会触发冲突检测。

解决方案 #

方案1：统一用Poetry管理所有依赖 #

放弃用pip安装sim_requirements.txt，全流程用Poetry处理：

1. 在sim_requirements.txt所在的项目根目录初始化Poetry（未初始化过的话）：

   poetry init
   

2. 添加本地可编辑依赖：

   poetry add -e ./libs/first_library -e ./libs/second_library
   

3. 安装所有依赖：

   poetry install
   

Poetry会统一解析本地依赖，不会出现pip的识别冲突问题。

方案2：统一本地包的引用路径 #

如果必须用pip安装，修改first-library的pyproject.toml，让second-library的路径和sim_requirements.txt中的路径完全一致：
原配置：

second-library = {path = "../second_library", develop = true}

修改为：

second-library = {path = "./libs/second_library", develop = true}

或者反过来调整sim_requirements.txt里的路径，核心是让两个地方引用second-library的相对路径完全相同，这样pip会识别为同一个包。

方案3：临时跳过依赖解析（应急用） #

如果只是临时安装，可通过--no-deps参数跳过依赖检查：

pip install -r sim_requirements.txt --no-deps

之后需要手动安装所有依赖，此方法容易遗漏依赖，仅建议临时使用。

内容的提问来源于stack exchange，提问作者gCoh