Problem
I have read this post, which provides a way to permanently avoid the sys.path hack when importing names between sibling directories. However, I followed the procedures listed in that post but found that I could not import installed package (i.e. test).
The following are things I have already done
Step1: create a project that looks like following. Both __init__.py are empty.
test
├── __init__.py
├── setup.py
├── subfolder1
│ ├── __init__.py
│ ├── program1.py
├── subfolder2
│ ├── __init__.py
│ └── program2.py
# setup.py
from setuptools import setup, find_packages
setup(name="test", version="0.1", packages=find_packages())
# program1
def func1():
print("I am from func1 in subfolder1/func1")
# program2
from test.subfolder1 import program1
program1.func1()
Step2. create virtual environment in project root directory (i.e. test directory)
conda create -n test --clone base
launch a new terminal and conda activate test
pip install -e .
conda list and I see the following, which means my test project is indeed installed in the virtual environment
...
test 0.1 dev_0 <develop>
...
Step3: go to the subfolder2 and python program2.py, but unexpectedly it returned
ModuleNotFoundError: No module named 'test.subfolder1'
The issue is I think test should be available as long as I am in virtual environment. However, it does not seem to be the case here.
Could some one help me? Thank you in advance!
You need to create an empty __init__.py file in subfolder1 to make it a package.
Edit:
You should change the import in program2.py:
from subfolder1 import program1
Or you can move setup.py a level up.
I am looking for a way to import modules so that I can run a script both from the sub-folder project/v0 and from the root folder project
My file structure in python 3.6 (which is why there are no init files)
project
├── api.py
├── v0
│ ├── SearchEngine.py => contains SearchEngine class
│ └── SearchEngineBE.py
My SearchEngineBE.py module contains
from SearchEngine import SearchEngine
My api.py module contains
from v0.SearchEngineBE import SearchEngineBE
step1: When from project/v0 I run python3 SearcheEngineBE.py my module is correctly imported and everything goes well.
step2: However, when from project I run python3 run api.py I get the error:
Traceback (most recent call last):
File "api.py", line 3, in <module>
from v0.SearchEngineBE import SearchEngineBE
File "/xxx/project/v0/SearchEngineBE.py", line 3, in <module>
from SearchEngine import SearchEngine
ModuleNotFoundError: No module named 'SearchEngine'
How can I fix this so that both step 1 and step 2 would work ?
Non-relative imports are searched by the interpreter in the current directory (and any additional search paths).
You could use relative imports in your SearchEngineBE.py file to let the interpreter know you want the relative module, and not a module off the import path:
# SearchEngineBE.py
from .SearchEngine import SearchEngine
The . lets the interpreter know that you are referencing a module relative to the current module. You'll need at least a blank __init__.py file in the same directory as SearchEngine.py for relative imports to work though:
The __init__.py files are required to make Python treat directories
containing the file as packages.
https://docs.python.org/3/tutorial/modules.html
See this guide for some more discussion: https://chrisyeh96.github.io/2017/08/08/definitive-guide-python-imports.html#absolute-vs-relative-import
Edit:
Without a root package, this won't work. See this post for an alternative approach: Importing modules from a neighbouring folder in Python
Add __init__.py to make it a package.
(it could be empty file : __init__.py)
project
├── __init__.py
├── api.py
├── v0
│ ├── SearchEngine.py => contains SearchEngine class
│ └── SearchEngineBE.py
Edit 1:
try :
from project.v0.SearchEngine import SearchEngine
I wonder why my simple example building a class is not properly working:
The tree-structure looks like the following
class_project/
├── class_project
│ ├── __init__.py
│ └── classtheclass.py
└── tests
├── __init__.py
└── test.py
classtheclass.py looks like this:
class ClassTheClass(object):
def __init__(self):
print "yeees"
test.py looks like this:
from class_project.classtheclass import ClassTheClass
i = ClassTheClass()
While init.py are empty
so if i execute on the shell
python test.py
its giving me
Traceback (most recent call last): File "test.py", line 1, in
from class_project.classtheclass import ClassTheClass ImportError: No module named class_project.classtheclass
Whats wrong with that. In Pycharm this even works...!
When you run python test.py, the interpreter will look for Python code in standard library locations (e.g. /usr/local/lib/python2.7/site-packages and friends) and in the tests folder where you invoked the interpreter (this set of locations is known as the "Python Path", and you can view it with: import sys; print sys.path).
None of those locations include class_project.classtheclass.
There are a variety of ways to solve this problem.
You can set the PYTHONPATH environment variable to include class_project:
export PYTHONPATH="/path/to/class_project:$PYTHONPATH" # Note: this has to be the top-level class_project directory, you have two with this name here.
python test.py # This will now work
You could probably also use a relative import, but I'd argue that's working around the problem not solving the problem.
Your file tree structure is wrong, if you are not exporting PYTHONPATH with the new lib path, you have to put the test.py as below structure to make the test.py to access classtheclass from class_project.
class_project/
├── class_project
│ ├── classtheclass.py
│ └── __init__.py
└── test.py
Python search imported module as follow, for example import foo
Firstly, search the built-in module, If foo isn't a built-in module, go to next
Interpreter will search the sys.path, sys.path will include: current directory, PYTHONPATH and installation dependent path. If foo is in sys.path, it will be imported. Otherwise, it will be ImportError.
Here, you can execute at project parent directory or explicit add /path/to/project to sys.path.
I've tried reading through questions about sibling imports and even the
package documentation, but I've yet to find an answer.
With the following structure:
├── LICENSE.md
├── README.md
├── api
│ ├── __init__.py
│ ├── api.py
│ └── api_key.py
├── examples
│ ├── __init__.py
│ ├── example_one.py
│ └── example_two.py
└── tests
│ ├── __init__.py
│ └── test_one.py
How can the scripts in the examples and tests directories import from the
api module and be run from the commandline?
Also, I'd like to avoid the ugly sys.path.insert hack for every file. Surely
this can be done in Python, right?
Tired of sys.path hacks?
There are plenty of sys.path.append -hacks available, but I found an alternative way of solving the problem in hand.
Summary
Wrap the code into one folder (e.g. packaged_stuff)
Create setup.py script where you use setuptools.setup(). (see minimal setup.py below)
Pip install the package in editable state with pip install -e <myproject_folder>
Import using from packaged_stuff.modulename import function_name
Setup
The starting point is the file structure you have provided, wrapped in a folder called myproject.
.
└── myproject
├── api
│ ├── api_key.py
│ ├── api.py
│ └── __init__.py
├── examples
│ ├── example_one.py
│ ├── example_two.py
│ └── __init__.py
├── LICENCE.md
├── README.md
└── tests
├── __init__.py
└── test_one.py
I will call the . the root folder, and in my example case it is located at C:\tmp\test_imports\.
api.py
As a test case, let's use the following ./api/api.py
def function_from_api():
return 'I am the return value from api.api!'
test_one.py
from api.api import function_from_api
def test_function():
print(function_from_api())
if __name__ == '__main__':
test_function()
Try to run test_one:
PS C:\tmp\test_imports> python .\myproject\tests\test_one.py
Traceback (most recent call last):
File ".\myproject\tests\test_one.py", line 1, in <module>
from api.api import function_from_api
ModuleNotFoundError: No module named 'api'
Also trying relative imports wont work:
Using from ..api.api import function_from_api would result into
PS C:\tmp\test_imports> python .\myproject\tests\test_one.py
Traceback (most recent call last):
File ".\tests\test_one.py", line 1, in <module>
from ..api.api import function_from_api
ValueError: attempted relative import beyond top-level package
Steps
Make a setup.py file to the root level directory
The contents for the setup.py would be*
from setuptools import setup, find_packages
setup(name='myproject', version='1.0', packages=find_packages())
Use a virtual environment
If you are familiar with virtual environments, activate one, and skip to the next step. Usage of virtual environments are not absolutely required, but they will really help you out in the long run (when you have more than 1 project ongoing..). The most basic steps are (run in the root folder)
Create virtual env
python -m venv venv
Activate virtual env
source ./venv/bin/activate (Linux, macOS) or ./venv/Scripts/activate (Win)
To learn more about this, just Google out "python virtual env tutorial" or similar. You probably never need any other commands than creating, activating and deactivating.
Once you have made and activated a virtual environment, your console should give the name of the virtual environment in parenthesis
PS C:\tmp\test_imports> python -m venv venv
PS C:\tmp\test_imports> .\venv\Scripts\activate
(venv) PS C:\tmp\test_imports>
and your folder tree should look like this**
.
├── myproject
│ ├── api
│ │ ├── api_key.py
│ │ ├── api.py
│ │ └── __init__.py
│ ├── examples
│ │ ├── example_one.py
│ │ ├── example_two.py
│ │ └── __init__.py
│ ├── LICENCE.md
│ ├── README.md
│ └── tests
│ ├── __init__.py
│ └── test_one.py
├── setup.py
└── venv
├── Include
├── Lib
├── pyvenv.cfg
└── Scripts [87 entries exceeds filelimit, not opening dir]
pip install your project in editable state
Install your top level package myproject using pip. The trick is to use the -e flag when doing the install. This way it is installed in an editable state, and all the edits made to the .py files will be automatically included in the installed package.
In the root directory, run
pip install -e . (note the dot, it stands for "current directory")
You can also see that it is installed by using pip freeze
(venv) PS C:\tmp\test_imports> pip install -e .
Obtaining file:///C:/tmp/test_imports
Installing collected packages: myproject
Running setup.py develop for myproject
Successfully installed myproject
(venv) PS C:\tmp\test_imports> pip freeze
myproject==1.0
Add myproject. into your imports
Note that you will have to add myproject. only into imports that would not work otherwise. Imports that worked without the setup.py & pip install will work still work fine. See an example below.
Test the solution
Now, let's test the solution using api.py defined above, and test_one.py defined below.
test_one.py
from myproject.api.api import function_from_api
def test_function():
print(function_from_api())
if __name__ == '__main__':
test_function()
running the test
(venv) PS C:\tmp\test_imports> python .\myproject\tests\test_one.py
I am the return value from api.api!
* See the setuptools docs for more verbose setup.py examples.
** In reality, you could put your virtual environment anywhere on your hard disk.
Seven years after
Since I wrote the answer below, modifying sys.path is still a quick-and-dirty trick that works well for private scripts, but there has been several improvements
Installing the package (in a virtualenv or not) will give you what you want, though I would suggest using pip to do it rather than using setuptools directly (and using setup.cfg to store the metadata)
Using the -m flag and running as a package works too (but will turn out a bit awkward if you want to convert your working directory into an installable package).
For the tests, specifically, pytest is able to find the api package in this situation and takes care of the sys.path hacks for you
So it really depends on what you want to do. In your case, though, since it seems that your goal is to make a proper package at some point, installing through pip -e is probably your best bet, even if it is not perfect yet.
Old answer
As already stated elsewhere, the awful truth is that you have to do ugly hacks to allow imports from siblings modules or parents package from a __main__ module. The issue is detailed in PEP 366. PEP 3122 attempted to handle imports in a more rational way but Guido has rejected it one the account of
The only use case seems to be running scripts that happen
to be living inside a module's directory, which I've always seen as an
antipattern.
(here)
Though, I use this pattern on a regular basis with
# Ugly hack to allow absolute import from the root folder
# whatever its name is. Please forgive the heresy.
if __name__ == "__main__" and __package__ is None:
from sys import path
from os.path import dirname as dir
path.append(dir(path[0]))
__package__ = "examples"
import api
Here path[0] is your running script's parent folder and dir(path[0]) your top level folder.
I have still not been able to use relative imports with this, though, but it does allow absolute imports from the top level (in your example api's parent folder).
Here is another alternative that I insert at top of the Python files in tests folder:
# Path hack.
import sys, os
sys.path.insert(0, os.path.abspath('..'))
You don't need and shouldn't hack sys.path unless it is necessary and in this case it is not. Use:
import api.api_key # in tests, examples
Run from the project directory: python -m tests.test_one.
You should probably move tests (if they are api's unittests) inside api and run python -m api.test to run all tests (assuming there is __main__.py) or python -m api.test.test_one to run test_one instead.
You could also remove __init__.py from examples (it is not a Python package) and run the examples in a virtualenv where api is installed e.g., pip install -e . in a virtualenv would install inplace api package if you have proper setup.py.
I don't yet have the comprehension of Pythonology necessary to see the intended way of sharing code amongst unrelated projects without a sibling/relative import hack. Until that day, this is my solution. For examples or tests to import stuff from ..\api, it would look like:
import sys.path
import os.path
# Import from sibling directory ..\api
sys.path.append(os.path.dirname(os.path.abspath(__file__)) + "/..")
import api.api
import api.api_key
For siblings package imports, you can use either the insert or the append method of the [sys.path][2] module:
if __name__ == '__main__' and if __package__ is None:
import sys
from os import path
sys.path.append( path.dirname( path.dirname( path.abspath(__file__) ) ) )
import api
This will work if you are launching your scripts as follows:
python examples/example_one.py
python tests/test_one.py
On the other hand, you can also use the relative import:
if __name__ == '__main__' and if __package__ is not None:
import ..api.api
In this case you will have to launch your script with the '-m' argument (note that, in this case, you must not give the '.py' extension):
python -m packageName.examples.example_one
python -m packageName.tests.test_one
Of course, you can mix the two approaches, so that your script will work no matter how it is called:
if __name__ == '__main__':
if __package__ is None:
import sys
from os import path
sys.path.append( path.dirname( path.dirname( path.abspath(__file__) ) ) )
import api
else:
import ..api.api
For readers in 2021: If you're not confident with pip install -e :
Consider this hierarchy, as recommended by an answer from Relative imports in Python 3:
MyProject
├── src
│ ├── bot
│ │ ├── __init__.py
│ │ ├── main.py
│ │ └── sib1.py
│ └── mod
│ ├── __init__.py
│ └── module1.py
└── main.py
The content of main.py, which is the starting point and we use absolute import (no leading dots) here:
from src.bot import main
if __name__ == '__main__':
main.magic_tricks()
The content of bot/main.py, which takes advantage of explicit relative imports:
from .sib1 import my_drink # Both are explicit-relative-imports.
from ..mod.module1 import relative_magic
def magic_tricks():
# Using sub-magic
relative_magic(in=["newbie", "pain"], advice="cheer_up")
my_drink()
# Do your work
...
Now here comes the reasoning:
When executing python MyProject/main.py, the path/to/MyProject is added into the sys.path.
The absolute import import src.bot will read it.
The from ..mod part means it will go up one level to MyProject/src.
Can we see it? YES, since path/to/MyProject is added into the sys.path.
So the point is:
We should put the main script next to MyProject/src, since that when doing relative-referencing, we won't go out of the src, and the absolute import import src. provides the just-fit scope for us: the src/ scope.
See also: ModuleNotFoundError: No module named 'sib1'
TLDR
This method does not require setuptools, path hacks, additional command line arguments, or specifying the top level of the package in every single file of your project.
Just make a script in the parent directory of whatever your are calling to be your __main__ and run everything from there. For further explanation continue reading.
Explanation
This can be accomplished without hacking a new path together, extra command line args, or adding code to each of your programs to recognize its siblings.
The reason this fails as I believe was mentioned before is the programs being called have their __name__ set as __main__. When this occurs the script being called accepts itself to be on the top level of the package and refuses to recognize scripts in sibling directories.
However, everything under the top level of the directory will still recognize ANYTHING ELSE under the top level. This means the ONLY thing you have to do to get files in sibling directories to recognize/utilize each other is to call them from a script in their parent directory.
Proof of Concept
In a dir with the following structure:
.
|__Main.py
|
|__Siblings
|
|___sib1
| |
| |__call.py
|
|___sib2
|
|__callsib.py
Main.py contains the following code:
import sib1.call as call
def main():
call.Call()
if __name__ == '__main__':
main()
sib1/call.py contains:
import sib2.callsib as callsib
def Call():
callsib.CallSib()
if __name__ == '__main__':
Call()
and sib2/callsib.py contains:
def CallSib():
print("Got Called")
if __name__ == '__main__':
CallSib()
If you reproduce this example you will notice that calling Main.py will result in "Got Called" being printed as is defined in sib2/callsib.py even though sib2/callsib.py got called through sib1/call.py. However if one were to directly call sib1/call.py (after making appropriate changes to the imports) it throws an exception. Even though it worked when called by the script in its parent directory, it will not work if it believes itself to be on the top level of the package.
You need to look to see how the import statements are written in the related code. If examples/example_one.py uses the following import statement:
import api.api
...then it expects the root directory of the project to be in the system path.
The easiest way to support this without any hacks (as you put it) would be to run the examples from the top level directory, like this:
PYTHONPATH=$PYTHONPATH:. python examples/example_one.py
Just in case someone using Pydev on Eclipse end up here: you can add the sibling's parent path (and thus the calling module's parent) as an external library folder using Project->Properties and setting External Libraries under the left menu Pydev-PYTHONPATH. Then you can import from your sibling, e. g. from sibling import some_class.
I wanted to comment on the solution provided by np8 but I don't have enough reputation so I'll just mention that you can create a setup.py file exactly as they suggested, and then do pipenv install --dev -e . from the project root directory to turn it into an editable dependency. Then your absolute imports will work e.g. from api.api import foo and you don't have to mess around with system-wide installations.
Documentation
If you're using pytest then the pytest docs describe a method of how to reference source packages from a separate test package.
The suggested project directory structure is:
setup.py
src/
mypkg/
__init__.py
app.py
view.py
tests/
__init__.py
foo/
__init__.py
test_view.py
bar/
__init__.py
test_view.py
Contents of the setup.py file:
from setuptools import setup, find_packages
setup(name="PACKAGENAME", packages=find_packages())
Install the packages in editable mode:
pip install -e .
The pytest article references this blog post by Ionel Cristian Mărieș.
I made a sample project to demonstrate how I handled this, which is indeed another sys.path hack as indicated above. Python Sibling Import Example, which relies on:
if __name__ == '__main__': import os import sys sys.path.append(os.getcwd())
This seems to be pretty effective so long as your working directory remains at the root of the Python project.
in your main file add this:
import sys
import os
sys.path.append(os.path.abspath(os.path.join(__file__,mainScriptDepth)))
mainScriptDepth = the depth of the main file from the root of the project.
Here is your case mainScriptDepth = "../../". Then you can import by specifying the path (from api.api import * ) from the root of your project.
The problem:
You simply can not get import mypackage to work in test.py. You will need either an editable install, change to path, or changes to __name__ and path
demo
├── dev
│ └── test.py
└── src
└── mypackage
├── __init__.py
└── module_of_mypackage.py
--------------------------------------------------------------
ValueError: attempted relative import beyond top-level package
The solution:
import sys; sys.path += [sys.path[0][:-3]+"src"]
Put the above before attempting imports in test.py. Thats it. You can now import mypackage.
This will work both on Windows and Linux. It will also not care from which path you run your script. It is short enough to slap it anywhere you might need it.
Why it works:
The sys.path contains the places, in order, where to look for packages when attempting imports if they are not found in installed site packages. When you run test.py the first item in sys.path will be something like /mnt/c/Users/username/Desktop/demo/dev i.e.: where you ran your file. The oneliner will simply add the sibling folder to path and everything works. You will not have to worry about Windows vs Linux file paths since we are only editing the last folder name and nothing else. If you project structure is already set in stone for your repository we can also reasonably just use the magic number 3 to slice away dev and substitute src
for the main question:
call sibling folder as module:
from .. import siblingfolder
call a_file.py from sibling folder as module:
from ..siblingfolder import a_file
call a_function inside a file in sibling folder as module:
from..siblingmodule.a_file import func_name_exists_in_a_file
The easiest way.
go to lib/site-packages folder.
if exists 'easy_install.pth' file, just edit it and add your directory that you have script that you want make it as module.
if not exists, just make it one...and put your folder that you want there
after you add it..., python will be automatically perceive that folder as similar like site-packages and you can call every script from that folder or subfolder as a module.
i wrote this by my phone, and hard to set it to make everyone comfortable to read.
First, you should avoid having files with the same name as the module itself. It may break other imports.
When you import a file, first the interpreter checks the current directory and then searchs global directories.
Inside examples or tests you can call:
from ..api import api
Project
1.1 User
1.1.1 about.py
1.1.2 init.py
1.2 Tech
1.2.1 info.py
1.1.2 init.py
Now, if you want to access about.py module in the User package, from the info.py module in Tech package then you have to bring the cmd (in windows) path to Project i.e.
**C:\Users\Personal\Desktop\Project>**as per the above Package example. And from this path you have to enter, python -m Package_name.module_name
For example for the above Package we have to do,
C:\Users\Personal\Desktop\Project>python -m Tech.info
Imp Points
Don't use .py extension after info module i.e. python -m Tech.info.py
Enter this, where the siblings packages are in the same level.
-m is the flag, to check about it you can type from the cmd python --help
I have a question regarding one single module that is distributed over multiple directories.
Let's say I have these two file and directories:
~/lib/python
xxx
__init__.py
util
__init__.py
module1.py
module2.py
~/graphics/python
xxx
__init__.py
misc
__init__.py
module3.py
module4.py
So then in my Python modules, I did this:
import sys
pythonlibpath = '~/lib/python'
if pythonlibpath not in sys.path: sys.path.append(pythonlibpath)
import xxx.util.module1
which works.
Now, the problem is that I need xxx.misc.module3, so I did this:
import sys
graphicslibpath = '~/graphics/python'
if graphicslibpath not in sys.path: sys.path.append(graphicslibpath)
import xxx.misc.module3
but I get this error:
ImportError: No module named misc.module3
It seems like it somehow still remembers that there was a xxx package in ~/lib/python and then tries to find misc.module3 from there.
How do I get around this issue?
You can't without an extreme amount of trickery that pulls one package structure into the other. Python requires that all modules in a package be under a single subdirectory. See the os source to learn how it handles os.path.
Python does indeed remember that there was a xxx package. This is pretty much necessary to achieve acceptable performance, once modules and packages are loaded they are cached. You can see which modules are loaded by looking the the dictionary sys.modules.
sys.modules is a normal dictionary so you can remove a package from it to force it to be reloaded like below:
import sys
print sys.modules
import xml
print sys.modules
del sys.modules['xml']
print sys.modules
Notice that after importing the xml package it is the dictionary, however it is possible to remove it from that dictionary too. This is a point I make for pedagogical purposes only, I would not recommend this approach in a real application. Also if you need to use your misc and util packages together this would not work so great. If at all possible rearrange your source code structure to better fit the normal Python module loading mechanism.
This is addressed by Implicit Namespace Packages in Python 3.3. See PEP-420.
This is an adaptation of an answer to a similar question.
Following up on #Gary's answer, the PEP 420 page says to use the following code on shared __init__.py packages.
__init__.py:
from pkgutil import extend_path
__path__ = extend_path(__path__, __name__)
This code should be placed inside the xxx directory's __init__.py.
See the *s below
someroot/
├── graphics
│ └── python
│ └── xxx
│ ├── ****__init__.py****
│ └── misc
│ ├── __init__.py
│ ├── module3.py
│ └── module4.py
└── lib
└── python
└── xxx
├── ****__init__.py****
└── util
├── __init__.py
├── module1.py
└── module2.py
Some setup.sh file to add to the Python Path:
libPath=someroot/lib/python/
graphicsPath=someroot/graphics/python/
export PYTHONPATH=$PYTHONPATH:$libPath:$graphicsPath
Python test code (tested on Python versions 2.7.14 and 3.6.4 using pyenv):
import xxx.util.module1
import xxx.misc.module3 # No errors