Skip to content

Latest commit

 

History

History
80 lines (50 loc) · 2.89 KB

python.md

File metadata and controls

80 lines (50 loc) · 2.89 KB

Hello Python

Next we introduce some Python code into our workflow. First we will run pylint to check our code for possible syntax errors, then we run some python code to produce an output which we will save as an artifact. As a test case i prepared a small python programm generating Mandelbrot Sets. It generates an ASCII version by default, printing to stdout, but also a higher resolution .pdf when --pdf is specified as an argument..

In many production cases one would run for example some sort of unit test or build a python package and deploy it to pip.

To get the sample python script run

curl -L https://derse-ci.beuscher.dev/mandelbrot.py -o mandelbrot.py

or download it manually.

Setup Python and pylint

Starting with a copy of the hello-ci.yml workflow, we choose suitable names the workflow(-run) and name the job plot

{{#include .github/workflows/python-ci.yml:1:10}}

To setup our python environment we use an action provided by GitHub and specify a python version, in this case 3.10

{{#include .github/workflows/python-ci.yml:11:14}}

To install the numpy and matplotlib dependency used by this programm as well as pylint we provide pip (the python package manager) with a requirements.txt file which will be located in the top level directory of our repository.

Our requirements.txt:

{{#include requirements.txt}}

In the workflow file we add the following two commands:

{{#include .github/workflows/python-ci.yml:15:18}}

Next, we run pylint on mandelbrot.py to check for syntax error and if it obeys code conventions.

{{#include .github/workflows/python-ci.yml:19:21}}

Note: Here we need the argument --disable=import-outside-toplevel since the matplotlib dependency is only imported when --pdf is specified.

When a syntax error is detected pylint will return an error and the job will be marked as failed.

Finally we execute the python program with the --pdf argument:

{{#include .github/workflows/python-ci.yml:22:24}}

Save an Artifact

By default we do not retain any build artifacts or other files generated during our workflow run. When files should be made available outside a job one needs to upload an artifact explicitly. This Artifact is then available in another job to "download" or can be viewed and downloaded from GitHub. If no retention days are specified GitHub will use the account default (initially set to 30 Days)

{{#include .github/workflows/python-ci.yml:25:}}

The full action should now be similar to

{{#include .github/workflows/python-ci.yml}}

Again one could check the worklfow using actionlint.

Commit and view the Artifacts

We again commit our changes to the remote repository.

After the run is finished we can access the generated artifacts when selecting a specific workflow run.