To connect to python remote interpreter inside docker, you have to make sure first, that Pycharm is aware of your docker.
Go to *Settings > Build, Execution, Deployment > Docker*. If you are on linux, you can use docker directly using its socket `unix:///var/run/docker.sock`, if you are on Windows or Mac, make sure that you have docker-machine installed, then you can simply *Import credentials from Docker Machine*.
..image:: images/1.png
Configure Remote Python Interpreter
-----------------------------------
This repository comes with already prepared "Run/Debug Configurations" for docker.
But as you can see, at the beggining there is something wrong with them. They have red X on django icon, and they cannot be used, without configuring remote python interpteter. To do that, you have to go to *Settings > Build, Execution, Deployment* first.
Next, you have to add new remote python interpreter, based on already tested deployment settings. Go to *Settings > Project > Project Interpreter*. Click on the cog icon, and click *Add Remote*.
..image:: images/3.png
Switch to *Docker Compose* and select `dev.yml` file from directory of your project, next set *Service name* to `django`
..image:: images/4.png
Because Pycharm restarts container every time you use Configuration Run, to not have server restarted during running tests, we defined second service in `dev.yml` file called pycharm. To use it, you have to add interpreter of second service as well.
..image:: images/5.png
The final result should be:
..image:: images/6.png
Having that, click *OK*. Close *Settings* panel, and wait few seconds...
..image:: images/7.png
After few seconds, all *Run/Debug Configurations* should be ready to use.
..image:: images/8.png
**Things you can do with provided configuration**:
Most of the files from `.idea/` were added to `.gitignore` with a few exceptions, which were made, to provide "ready to go" configuration. After adding remote interpreter some of these files are altered by PyCharm:
..image:: images/issue2.png
In theory you can remove them from repository, but then, other people will lose a ability to initialize a project from provided configurations as you did. To get rid of this annoying state, you can run command::