32位Python虚拟环境安装scikit-image报错求助

问题描述 #

64位Windows系统中同时安装了32位和64位Python 3.9.13，在VSCode创建的32位虚拟环境中执行python -m pip install -U scikit-image时，Meson构建系统报错：Need python for x86_64, but found x86，最终提示Python dependency not found。已确认当前环境为32位且路径配置正常，64位Python环境可正常安装该库，依赖列表也已核对无误。

完整报错日志：

(.venv) C:\Users\osimsek\Desktop\Python\vscode_venv_x86>python -m pip install -U scikit-image
Collecting scikit-image
  Using cached scikit_image-0.22.0.tar.gz (22.7 MB)
  Installing build dependencies ... done
  Getting requirements to build wheel ... done
  Installing backend dependencies ... done
  Preparing metadata (pyproject.toml) ... error
  error: subprocess-exited-with-error

  × Preparing metadata (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [28 lines of output]
      + meson setup C:\Users\osimsek\AppData\Local\Temp\pip-install-9pouxues\scikit-image_959a707540ba4f3d9c6bef5cbffc6d12 C:\Users\osimsek\AppData\Local\Temp\pip-install-9pouxues\scikit-image_959a707540ba4f3d9c6bef5cbffc6d12\.mesonpy-agpxo0eg -Dbuildtype=release -Db_ndebug=if-release -Db_vscrt=md --native-file=C:\Users\osimsek\AppData\Local\Temp\pip-install-9pouxues\scikit-image_959a707540ba4f3d9c6bef5cbffc6d12\.mesonpy-agpxo0eg\meson-python-native-file.ini
      The Meson build system
      Version: 1.2.2
      Source dir: C:\Users\osimsek\AppData\Local\Temp\pip-install-9pouxues\scikit-image_959a707540ba4f3d9c6bef5cbffc6d12
      Build dir: C:\Users\osimsek\AppData\Local\Temp\pip-install-9pouxues\scikit-image_959a707540ba4f3d9c6bef5cbffc6d12\.mesonpy-agpxo0eg
      Build type: native build
      Project name: scikit-image
      Project version: 0.22.0
      Activating VS 17.7.1
      C compiler for the host machine: cl (msvc 19.37.32822 "x64 için Microsoft (R) C/C++ İyileştirmeli Derleyici Sürüm 19.37.32822")
      C linker for the host machine: link link 14.37.32822.0
      C++ compiler for the host machine: cl (msvc 19.37.32822 "x64 için Microsoft (R) C/C++ İyileştirmeli Derleyici Sürüm 19.37.32822")
      C++ linker for the host machine: link link 14.37.32822.0
      Host machine cpu family: x86_64
      Host machine cpu: x86_64
      Compiler for C supports arguments -Wno-unused-function: NO
      Library m found: NO
      Checking if "-Wl,--version-script" : links: NO
      Program cython found: YES (C:\Users\osimsek\AppData\Local\Temp\pip-build-env-egn3taya\overlay\Scripts\cython.EXE)
      Program pythran found: YES (C:\Users\osimsek\AppData\Local\Temp\pip-build-env-egn3taya\overlay\Scripts\pythran.EXE)
      Program cp found: YES (C:\cygwin\bin\cp.EXE)
      Program python found: YES (C:\Users\osimsek\Desktop\Python\vscode_venv_x86\.venv\Scripts\python.exe)
      Need python for x86_64, but found x86
      Run-time dependency python found: NO (tried sysconfig)
     
      ..\meson.build:69:14: ERROR: Python dependency not found
     
      A full log can be found at C:\Users\osimsek\AppData\Local\Temp\pip-install-9pouxues\scikit-image_959a707540ba4f3d9c6bef5cbffc6d12\.mesonpy-agpxo0eg\meson-logs\meson-log.txt
      [end of output]

  note: This error originates from a subprocess, and is likely not a problem with pip.
error: metadata-generation-failed

× Encountered error while generating package metadata.
╰─> See above for output.

note: This is an issue with the package mentioned above, not pip.
hint: See above for details.

解决方案 #

1. 切换到32位MSVC编译环境 #

报错日志显示当前使用的是x64版本的MSVC编译器，32位Python环境需要对应x86的编译工具链：

• 打开Visual Studio开发者命令提示符（x86）（根据你的VS版本选择对应工具，比如VS 2022对应"Developer Command Prompt for VS 2022 (x86)"）
• 在该命令行中激活32位虚拟环境：C:\Users\osimsek\Desktop\Python\vscode_venv_x86\.venv\Scripts\activate.bat
• 再次执行安装命令：python -m pip install -U scikit-image

如果使用VSCode终端，可先运行VS的32位环境配置脚本（路径根据你的VS安装位置调整）：

"C:\Program Files (x86)\Microsoft Visual Studio\2022\Community\VC\Auxiliary\Build\vcvars32.bat"

激活虚拟环境后再执行pip安装。

2. 强制指定Meson构建架构 #

在pip安装时传递Meson参数，强制指定x86架构：

python -m pip install -U scikit-image --config-settings=setup-args=-Dhost_cpu=x86

或者先设置环境变量再安装：

SET MESON_ARGS=-Dhost_cpu=x86
python -m pip install -U scikit-image

3. 尝试安装预编译wheel包 #

升级pip和wheel后，优先使用预编译包避免源码编译：

python -m pip install -U pip wheel
python -m pip install -U scikit-image --only-binary :all:

注：需确认scikit-image提供了Python 3.9的32位预编译wheel，若没有则需回到前两种方案。

4. 再次确认虚拟环境架构 #

执行以下命令验证当前环境是否为32位：

python -c "import platform; print(platform.architecture())"

输出应为('32bit', 'WindowsPE')，确保虚拟环境确实基于32位Python创建。

内容的提问来源于stack exchange，提问作者oguzhantr