I was looking for all the internet, I tried all the possible ways and I still can not import the module that I need. The tree of my project is this:
code
├── folder1
│ ├── subfolder 1
│ ├── file1.py (file to be imported)
│── folder2
│ └── main.py
It is not very elegant, but at least it is the only way I found starting from a folder back:
import sys
sys.path.append("..")
from carpeta1.subcarpeta1.archivo1 import *
In your case, main.py in carpeta2 you want to import a module that physically and relative to main.py is in: ../carpeta1/subcarpeta1/archivo1.py , so we add to path the folder root of your project with sys.path.append("..") , keep in mind that if you did it from a deeper level you should do sys.path.append("../..") . With this, it is already possible to do from carpeta1.subcarpeta1.archivo1 import * . Keep in mind not to add spaces to the name of the module to be imported, nor to the subfolders.
Another way is to use importlib , which also allows you to solve the problem of names of modules with points or other characters that invalidate a% traditional import . Suppose we have a routine fun_en_archivo_py() in code/carpeta1/subcarpeta1 we can from main.py do the following:
import importlib.util
spec = importlib.util.spec_from_file_location("archivo", "../carpeta1/subcarpeta1/archivo.py")
archivo = importlib.util.module_from_spec(spec)
spec.loader.exec_module(archivo)
# Para invocar la rutina anteponemos el nombre del módulo
archivo.fun_en_archivo_py()
What yes, this is functional if I do not understand badly from the version Python 3.5+ or higher.
in code:
from carpeta1.subcarpeta1.archivo import *
but in each folder and subfolder you must add a file __init__.py (it can be empty).
The main use of __init__.py is to initialize Python packages. Although it may be empty as I said before, it is useful for the import of modules and packages to be more visible and orderly. I suggest you invest that.
Greetings!