Skip to content

Tests

The rests are triggered as part of an GitHub Action in any Pull Request you open.

Run the tests locally

In the tests/ folder, run make selenium to start a selenium container in podman. In another terminal run make test to fire up the tests.

HAProxy test

test_haproxy.sh queries the HAProxy stats endpoint, it will fail if there is any DOWN backens. I use this like below to wait for the different components to start. It usually takes around a minute or two for the state to get ready.

wait:
    ./test_haproxy.sh

The HAProxy tests are configured like this:

backend be_server
    option httpchk
    http-check send meth GET uri /server-info/ping
    server immich-server 127.0.0.1:3001 maxconn 64 check inter 5s fall 2 rise 6

To summarize, HAProxy do a few basic backend checks to make sure that the services runs correctly. I use this information to detect if everyting is started and behaves correctly. This is exposed with the test_haproxy.sh script.

Selenium tests

These tests are triggered when the above tests succeeds. If you have started the selenium container, you can inspect the selenium browser process at localhost:7900 (password: secret).

The selenium tests are inside tests_selenium.py. The tests is a mix of browser tests, API calls and cli commands. At the moment the test contains these steps:

  1. Register a user
  2. Log in with the above user and check for JS and 404 errors
  3. Make sure that we have an empty timeline
  4. Make an API key
  5. Verify that we can execute the CLI
  6. Upload all assets in tests/assets to Immich via the CLI using the API key
  7. Use the API to list all assets, this test will check that:
    • Immich contains all expected assets
    • Our videos are detected as videos
    • Exif extraction works on both videos and pictures
    • Our VP9 video is transcoded to MPEG4
    • Live/Motion photos are detected
  8. Verify that our six people are detected and processed
Hard to match elements in tests_selenium.py

The generated HTML is a mess, it looks like the sample below. id attributes and unique classes are rare.

<div class="w-[500px] max-w-[95vw] rounded-3xl border bg-immich-bg p-4 py-8 shadow-sm dark:border-immich-dark-gray dark:bg-immich-dark-gray dark:text-immich-dark-fg"><div class="flex flex-col place-content-center place-items-center gap-4 px-4 text-immich-primary dark:text-immich-dark-primary">...

So I'm forced to use CSS selectors on content, or Beautiful Soup + code to process the DOM to generate unique paths to select against.

self.assert_element("p:contains('CLICK TO UPLOAD YOUR FIRST PHOTO')")

soup = self.get_beautiful_soup()
api_keys_div = soup.find(string="API Keys").parent.parent.parent
api_keys_button = api_keys_div.find("button")
self.click(css_selector_path(api_keys_button))

I hope to add more test over time, like more filetypes. If you like to contribute a file with an appropriate license, open an issue.