Pyinstaller打包pikepdf失败的问题排查

问题

最近在项目里用到了pikepdf这个库，用于实现pdf水印插入的一个小功能，源码调试阶段运行一切OK但是在出包时报了如下异常。

Traceback (most recent call last):  File "pikepdf\__init__.py", line 19, in <module>  File "PyInstaller\loader\pyimod03_importers.py", line 495, in exec_module  File "pikepdf\_version.py", line 13, in <module>  File "importlib\metadata.py", line 530, in version  File "importlib\metadata.py", line 503, in distribution  File "importlib\metadata.py", line 177, in from_nameimportlib.metadata.PackageNotFoundError: pikepdfThe above exception was the direct cause of the following exception:Traceback (most recent call last):  File "main.py", line 1, in <module>  File "PyInstaller\loader\pyimod03_importers.py", line 495, in exec_module  File "pikepdf\__init__.py", line 21, in <module>ImportError: Failed to determine version[29708] Failed to execute script 'main' due to unhandled exception!

异常定位

打印了两份堆栈的信息，翻翻炸出来点 init.py 的21行，代码如下

try:    from ._version import __version__except ImportError as _e:  # pragma: no cover    raise ImportError("Failed to determine version") from _e

从.version.py文件导入__version__失败？看看_version.py

try:    from importlib_metadata import version as _package_version  # type: ignoreexcept ImportError:    from importlib.metadata import version as _package_version__version__ = _package_version('pikepdf')__all__ = ['__version__']

再看看上面的异常，也就再_package_version这个函数了。这边可以先写个简单的demo.py验证下，使用pyinstgaller编译后运行。

# demo.pyimport pikepdfif __name__ == '__main__':    print("Hello World")

# 输出Traceback (most recent call last):  File "pikepdf\__init__.py", line 19, in <module>  File "PyInstaller\loader\pyimod03_importers.py", line 495, in exec_module  File "pikepdf\_version.py", line 13, in <module>  File "importlib\metadata.py", line 530, in version  File "importlib\metadata.py", line 503, in distribution  File "importlib\metadata.py", line 177, in from_nameimportlib.metadata.PackageNotFoundError: pikepdfThe above exception was the direct cause of the following exception:Traceback (most recent call last):  File "main.py", line 1, in <module>  File "PyInstaller\loader\pyimod03_importers.py", line 495, in exec_module  File "pikepdf\__init__.py", line 21, in <module>ImportError: Failed to determine version[29708] Failed to execute script 'main' due to unhandled exception!

符合预期。所以说importlib.metadata.version 无法在pyinstaller打包后运行？

问题原因

对于pkgs_to_check_at_runtime中列出的每个包，需要通过在spec文件中使用copy_metadata(name)来收集元数据。说白了就是pyinstaller打包后缺少对应的metadata信息。

修复方案

1. 降低版本pikepdf的版本

这个库以前用过，没有出这个幺蛾子。看了下旧版本的源代码，一下是5.1.3版本之前的get_version实现，没有使用importlib库，自然也不会有问题。

from pkg_resources import DistributionNotFoundfrom pkg_resources import get_distribution as _get_distributiontry:    __version__ = _get_distribution(__package__).versionexcept DistributionNotFound:  # pragma: no cover    __version__ = "Not installed"__all__ = ['__version__']

2. 在pyinstaller打包时指定copy-metadata

像这样

pyinstaller -F --copy-metadata pikepdf main.py

碎碎念

我觉得完全可以像版本5.1.3之前一样，获取不到时赋值为"Not installed"一样就行。然后提了个pr[https://github.com/pikepdf/pikepdf/pull/358]给作者，可惜作者认为这是importlib的锅（不改），倒也时说得过去。