[PYTHON] Edit and debug the code in the Raspberry Pi with VS Code's SSH connection feature

What is VS Code?

Abbreviation for "Visual Studio Code", a source code editor developed by Microsoft. It is said that it is rapidly gaining popularity as a free, lightweight, and multifunctional editor.

Since the 2019/10 version, the SSH connection function to RaspberryPi (ARM v8) is officially supported, so I wrote an article about the result of trying Python code remote editing in Raspberry Pi.

** In conclusion, we were able to realize a comfortable remote development environment, such as debugging with breakpoints! ** **

Things necessary

・ RaspberryPi (RaspberryPi3 Model B in this example) -PC connected to the same network as above (Windows 10 in this example) -Visual Studio Code 1.39.2 or later (installed on the above PC, 1.44.2 is used in this example)

procedure

I referred to the following article https://qiita.com/hukatama024e/items/9f96730381f0a34c6da6

(1) SSH connection settings and saving of config file

[This article "For Public Key Authentication"](https://qiita.com/c60evaporator/items/2384416f1122ae124f50#%E5%85%AC%E9%96%8B%E9%8D%B5%E8%AA%8D % E8% A8% BC% E3% 81% AE% E5% A0% B4% E5% 90% 88% E6% 89% 8B% E9% A0% 86) to establish SSH public key authentication and config file Please save.

In addition, please save the config file in the following folder in the PC C: \ Users \ [username] \ .ssh

② Install VS Code

Please install VS Code on ** PC side **

Reference article https://qiita.com/psychoroid/items/7d85ae6bade4a67aedb1

③ Installation of Remote Development

Start VSCode and follow the steps below to install Remote Development

• Remote --Ssh and other necessary related tools are also installed at the same time you install Remote Development.

④ Connect to Raspberry Pi

Follow the steps below to select the SSH server to connect to = RaspberryPi ("raspi" in the figure below)

A new window will open and you will be asked to select a platform, so choose Linux

You will be asked for a password for public key authentication, enter it and press Enter

The first connection takes time, so wait for a while. If successful, you can access the files in the Raspberry Pi by clicking the red frame in the figure below.

⑤ Create code in Raspberry Pi

From here onward, you can use the console, but we will proceed with the description on the assumption that you will operate with the GUI.

**-Create an empty file ** Create an empty file by selecting "File"-> "New File"

** ・ Description of code contents ** For example, write the following Python code (calculate and display the standard deviation with numpy)

**-Save code ** "File"-> "Save As"-> specify save path-> "OK"

**-Try running from the console ** Open the console by selecting "Terminal" → "New Terminal" Go to the folder with the code and

python3 test.py

Run python code in You can see that the result is output to the console normally.

⑥ Debug execution of the created code

With VS Code GUI, you can even debug the code. You can also stop at a breakpoint, so there is no doubt that development efficiency will increase!

**-Install Python linter ** If you have already installed it locally on your PC, click "install in SSH: [hostname]" to install the Python linter on your Raspberry Pi as well.

When the installation is complete, "Reload Required" will appear, so click it to restart.

**-Open a working folder (usually the folder with the code) **

**-Debugging ** After opening the code you want to debug, execute the following operation

The console will open and debug will be performed

** ・ Breakpoint operation check ** As shown in the figure below, you can stop processing at a breakpoint and check the variable contents.

Now you can remotely debug the code in your Raspberry Pi with VS Code. ** The environment for quick development with GUI is ready, and I am personally satisfied with the result! ** **

Addendum: When an error occurs when importing a library installed with pip

When building an environment with pyenv on the Raspberry Pi side, I get the following error when importing a library installed with pip.

Error details: Unable to import'bluepy' pylint (import-error)

Cause

As described in here, It seems that the cause is that the location of Python referenced when debugging with VS Code and Python of pyenv are different.

approach

/home/[username]/.vscode-server/extensions/ms-python.python-‥/pythonFiles/.vscode It is in settings.json Open and add the last line below

settings.json

{
    //
    //Original description
    //
    "python.pythonPath": "[Python location for pyenv]"
}

• By default, the Python location of pyenv is /home/[username]/.pyenv/shims/python Should be
• If python.pythonPath is already listed, comment it out and add a new line.

If you still get an error

Create a ".vscode" folder in the work folder opened in ⑥, Create the following settings.json file in the folder

settings.json

{
    "python.pythonPath": "[Python location for pyenv]"
}

This should force pyenv's Python to run