"fatal error: sql.h" with unixODBC and pyodbc as a vendor app dependency

Question

我正在尝试将 unixODBC 添加为供应商应用程序依赖项，按照此处概述的注释：https://github.com/cloudfoundry/python-buildpack/issues/34，但我仍然遇到致命错误：

sql.h: No such file or directory.

我在下面创建了一个最小的、可验证的、完整的脚本，以方便读者重现我的问题：

mkdir new_app
cd new_app

cat <<EOF>requirements.txt
pyodbc==4.0.17
EOF

cat <<EOF>manifest.yml
applications:
 - name: pyodbc-test
   memory: 128M
EOF

cat <<EOF>.profile
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/app/vendor/unixodbc/lib
EOF

sudo docker run -it --rm -v ${PWD}:/app cloudfoundry/cflinuxfs2 bash -c \
   "wget ftp://ftp.unixodbc.org/pub/unixODBC/unixODBC-2.3.4.tar.gz;  tar xzf unixODBC-2.3.4.tar.gz; cd unixODBC-2.3.4;  ./configure --prefix=/app/vendor/unixodbc; make; make install;"

其次是：

sudo chown -R ${USER}:${USER} vendor/
pip3 download -r requirements.txt -d vendor/
cf push

cf推送输出：

...
-------> Buildpack version 1.5.15
 !     Warning: Your application is missing a Procfile. This file tells Cloud Foundry how to run your application.
 !     Learn more: https://docs.cloudfoundry.org/buildpacks/prod-server.html#procfile
-----> Installing python-2.7.13
Downloaded [file:///tmp/buildpack/dependencies/https___buildpacks.cloudfoundry.org_dependencies_python_python-2.7.13-linux-x64.tgz]
     $ pip install -r requirements.txt
       Collecting pyodbc==4.0.17 (from -r requirements.txt (line 1))
       Installing collected packages: pyodbc
         Running setup.py install for pyodbc: started
           Running setup.py install for pyodbc: finished with status 'error'
           Complete output from command /app/.heroku/python/bin/python -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-sxxn23/pyodbc/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" install --record /tmp/pip-lRnuMu-record/install-record.txt --single-version-externally-managed --compile:
           running build
           running build_ext
           building 'pyodbc' extension
           creating build/temp.linux-x86_64-2.7
           creating build/temp.linux-x86_64-2.7/src
           gcc -pthread -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -DPYODBC_VERSION=4.0.17 -I/app/.heroku/python/include/python2.7 -c src/errors.cpp -o build/temp.linux-x86_64-2.7/src/errors.o -Wno-write-strings
           cc1plus: warning: command line option ‘-Wstrict-prototypes’ is valid for C/ObjC but not for C++ [enabled by default]
           In file included from src/errors.cpp:2:0:
           src/pyodbc.h:56:17: fatal error: sql.h: No such file or directory
            #include <sql.h>
                            ^
           compilation terminated.

           ----------------------------------------
           running install
           creating build
           error: command 'gcc' failed with exit status 1
Command "/app/.heroku/python/bin/python -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-sxxn23/pyodbc/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" install --record /tmp/pip-lRnuMu-record/install-record.txt --single-version-externally-managed --compile" failed with error code 1 in /tmp/pip-build-sxxn23/pyodbc/
Failed to compile droplet
Exit status 223
Staging failed: Exited with status 223
Destroying container
Successfully destroyed container

这里有一个类似的问题：，但是这个问题与这个问题的不同之处在于它非常笼统，而这个问题是关于同一主题但涉及一组非常具体的构建步骤。

更新： 这是基于已接受答案的更新脚本：

mkdir new_app
cd new_app

cat <<EOF>environment.yml
name: pyodbc-test
dependencies:
- pyodbc
EOF

cat <<EOF>manifest.yml
applications:
 - name: pyodbc-test
   memory: 128M
EOF

cf push

Answer 1

src/pyodbc.h:56:17: fatal error: sql.h: No such file or directory #include

您的设置总体上看起来不错，但是在暂存期间 pip 运行秒时，构建工具仍然无法找到库 headers。 headers 在那里，因为你将它们包含在 vendor/ 目录中，但它不知道它们在那里，因为 python 模块的构建脚本不会寻找他们在那个位置。

我不知道有什么方法可以让它在 non-standard 位置查找 headers 而无需修改构建包，因为 you need to send additional arguments 到 pip。

一些人建议设置 CPATH and/or C_INCLUDE_PATH 环境变量，因为这些应该被 gcc 获取，但在我的测试中，这些是忽略。

最后，我认为最好的办法是使用 Conda 安装 pyodbc。以下是您的操作方式。

将 environment.yml 添加到项目的根目录。
在其中添加以下内容：

name: pyodbc-test
dependencies:
- pyodbc

运行 cf push.

构建包应该运行在暂存期间安装 miniconda，然后使用 conda 安装依赖项，特别是 pyodbc。

希望对您有所帮助！

Answer 2

根据@Daniel Mikusa 的回答...

you need to send additional arguments to pip.

Pip 支持 --global-options 传递此类内容。不幸的是，选项从这里传递给 setup.py.

setup.py 有它自己的全局选项，即 build_ext...

Options for 'build_ext' command:
  --build-lib (-b)     directory for compiled extension modules
  --build-temp (-t)    directory for temporary files (build by-products)
  --plat-name (-p)     platform name to cross-compile for, if supported
                       (default: linux-x86_64)
  --inplace (-i)       ignore build-lib and put compiled extensions into the
                       source directory alongside your pure Python modules
  --include-dirs (-I)  list of directories to search for header files
  ...

Setup.py 需要 build_ext --include-dirs <path to unixodbc/include>

据我所知，要将这些参数提供给 setup.py，每个参数必须单独添加 --global-option=.

我刚刚经历了同样的头痛，以下对我有用。

pip install pyodbc --global-option="build_ext" --global-option="--include-dirs" --global-option="<path to unixodbc/include>"

"fatal error: sql.h" with unixODBC and pyodbc as a vendor app dependency

"fatal error: sql.h" with unixODBC and pyodbc as a vendor app dependency

cloud-foundry