Python编译的字节码：如果存在非ASCII字符，则无法读取文件

Question

First, a bit of background: I work in japan, so folders with japanese characters are common (something like 作業フォルダ, "working folder" ).

I have a python program that I then compile to bytecode. If I execute this compiled file from the windows command line, it runs without trouble:

for example:

python.exe C:\\作業フォルダ\\myProgram.py

however, once compiled with the following command:

python.exe -OO -m py_compile myProgram.py

I get the following error:

python.exe C:\\作業フォルダ\\myCompiledProgram.pyc
python: Can't reopen pyc file

if I go into the 作業フォルダ folder and execute like this:

python.exe myCompiledProgram.pyc

It runs without problems. However, I would like it to run in the case i give the Absolute path, and moreover i want to pass file paths as parameters, which also generates errors like:

RuntimeError: Unable to open C:\\作業フォルダ\\somefile.txt

Why does this happens only after compiling? Is there a way to fix it?

---

• If relevant, I am using windows 8.1, but happens in windows 10 computers as well, Python version is 3.6, installed with Anaconda3 5.10

Answer 1

A related answer suggests the fix is to move the script out of directories with non-ASCII characters. Junctions/symlinks can also be used to 'fake' non-ASCII directory names, but that is considerably more complicated.