Upgrade to Django 6 and switch to uv

Author: rogargon

.github/workflows/cicd.yml

@@ -15,14 +15,12 @@ jobs:
       uses: actions/setup-python@v2
       with:
         python-version: 3.13
-    - name: Set up Chromedriver for e2e
-      uses: nanasess/setup-chromedriver@v2
     - name: Install Dependencies
       run: |
-        pip install pipenv
-        pipenv install --dev
+        pip install uv
+        uv pip install --system -r pyproject.toml
     - name: Run Tests
       run: |
-        pipenv run python manage.py test
-        pipenv run python manage.py collectstatic --noinput
-        pipenv run python manage.py behave
+        python manage.py test
+        python manage.py collectstatic --noinput
+        python manage.py behave

Dockerfile

@@ -3,9 +3,9 @@ ENV PYTHONUNBUFFERED 1
 RUN mkdir /code
 WORKDIR /code
 
-COPY Pipfile Pipfile.lock ./
-RUN pip install pipenv
-RUN pipenv install --system --deploy
+COPY pyproject.toml ./
+RUN pip install uv
+RUN uv pip install --system -r pyproject.toml
 
 COPY myrecommendations ./myrecommendations
 COPY myrestaurants ./myrestaurants

Pipfile

@@ -28,7 +28,7 @@ Starting the MyRecommendations Project from Scratch
 ===================================================
 
 First of all, install the latest version of Python from [downloads](https://www.python.org/downloads/) and 
-[pipenv](https://pipenv.readthedocs.io/en/latest/install/) to help you manage dependencies and virtual environments.
+[uv](https://docs.astral.sh/uv/getting-started/installation/) to help you manage dependencies and virtual environments.
 
 Then, create the folder for the new project, in our case the project is called 'myrecommendations':
 
@@ -38,13 +38,15 @@ $ mkdir myrecommendations
 $ cd myrecommendations
 ```
 
-Once in the `myrecommendations` folder, activate the pipenv virtual environment to keep the Python packages
+Once in the `myrecommendations` folder, create a uv virtual environment to keep the Python packages
 for your project organised and start by installing Django. Then, create a new Django project:
 
 ```shell script
-$ pipenv shell
+$ uv venv
 
-$ pipenv install Django
+$ source .venv/bin/activate  # On Windows use: .venv\Scripts\activate
+
+$ uv pip install Django
 
 $ django-admin startproject myrecommendations .
 ```
@@ -174,8 +176,8 @@ To facilitate the description of the feature scenarios, while connecting them to
 To get Behave and integrate it with Django, install:
 
 ```shell script
-$ pipenv install behave
-$ pipenv install behave-django
+$ uv pip install behave
+$ uv pip install behave-django
 ```
 
 And add the 'behave_django' application at the end of the INSTALLED_APPS list in *myrecommendations/settings.py*:
@@ -196,52 +198,10 @@ INSTALLED_APPS = [
 Moreover, to make it possible to guide a browser from the test, and thus check if the application follows the expected behaviour from a end-user perspective, we will also use Splinter. It can be installed with the following command:
 
 ```shell script
-$ pipenv install splinter
-```
-
-Finally, for end-to-end test, it is necessary to have a browser to test from client side. With Splinter, different browsers can be configured for testing, for instance Chrome, the most common one. 
-
-Assuming Chrome is already installed in your computer, the only requirement to use it for automated testing is the ChromeDriver, available for Windows, Linux and Mac from [https://sites.google.com/a/chromium.org/chromedriver/downloads](https://sites.google.com/a/chromium.org/chromedriver/downloads)
-
-You can also install it using different package managers. For instance with **apt** on Linux:
-
-```shell script
-apt-get update
-apt-get install chromedriver
+$ uv pip install splinter selenium
 ```
 
-Or **[Brew](https://brew.sh)** on OSX:
-
-```shell script
-brew update
-brew install chromedriver
-xattr -d com.apple.quarantine $(which chromedriver)
-```
-
-Or **[Chocolatey](https://chocolatey.org/docs/installation)** on Windows:
-
-```shell script
-choco install chromedriver
-```
-
-### Testing with Firefox instead of Chrome ###
-
-Alternatively, for Firefox, install the browser:
-
-``` 
-sudo apt-get update
-sudo apt-get install firefox
-``` 
-
-Then, download Geckodriver and unpack it from [https://github.com/mozilla/geckodriver/releases/](https://github.com/mozilla/geckodriver/releases/)
-
-Finally, add the geckodriver to your path:
-
-```
-export PATH=$PATH:/your/path/to/geckodriver
-```
-
-And configure the test environment in `environment.py`, as detailed next, to use `firefox` instead of `chrome`.
+Finally, for end-to-end test, it is necessary to have a browser to test from client side. With Splinter and Selenium, different browsers can be configured for testing, for instance Chrome or Firefox, which should be installed in the machine where the tests are run.
 
 ## Environment ##
 
@@ -254,6 +214,7 @@ from splinter.browser import Browser
 
 def before_all(context):
     context.browser = Browser('chrome', headless=True)
+    # Alternatively, use `firefox` and headless=False to see the browser while testing
 
 def after_all(context):
     context.browser.quit()
@@ -362,6 +323,15 @@ def step_impl(context, username, password):
     context.browser.fill('password', password)
     form.find_by_value('login').first.click()
     assert context.browser.is_text_present('User: ' + username)
+
+@then("I'm redirected to the login form")
+def step_impl(context):
+    wait = WebDriverWait(context.browser.driver, 3)
+    login_form = wait.until(
+        visibility_of_element_located((By.ID, "login-form"))
+    )
+    # context.browser.driver.save_screenshot("debug_redirect.png")
+    assert context.browser.url.startswith(context.get_url('login'))
 ```
 
 Both steps are parameterized, so we can use the steps to login users with any username and password, as long as we have previously created them and the password matches.

Pipfile.lock

@@ -1,8 +1,47 @@
+import re
+import os
 from splinter.browser import Browser
 
+def chrome_browser(headless=True):
+    from selenium.webdriver.chrome.options import Options
+    chrome_options = Options()
+    if headless:
+        chrome_options.add_argument("--headless=new")  # recommended headless mode
+    chrome_options.add_argument("--disable-notifications")
+    chrome_options.add_argument("--disable-infobars")
+    chrome_options.add_argument("--no-sandbox")
+    chrome_options.add_argument("--disable-dev-shm-usage")
+    chrome_options.add_argument("--window-size=1920,1080")
+    prefs = {
+        "credentials_enable_service": False,  # disable password manager
+        "profile.password_manager_enabled": False,
+        "profile.password_manager_leak_detection": False
+    }
+    chrome_options.add_experimental_option("prefs", prefs)
+    return Browser("chrome", options=chrome_options)
+
+def firefox_browser(headless=True):
+    return Browser("firefox", headless=headless)
+
 def before_all(context):
-    context.browser = Browser('chrome', headless=True)
+    context.browser = chrome_browser(headless=True)
+    # Alternatively, use `firefox_browser` and headless=False to see the browser while testing
 
 def after_all(context):
     context.browser.quit()
     context.browser = None
+
+def slugify(text):
+    text = text.lower()
+    text = re.sub(r"[^\w\s-]", "", text)
+    text = re.sub(r"[\s]+", "-", text)
+    return text
+
+def after_step(context, step):
+    if step.status == "failed":
+        feature = slugify(context.feature.name)
+        scenario = slugify(context.scenario.name)
+        step_name = slugify(step.name)
+        os.makedirs("screenshots", exist_ok=True)
+        filename = f"screenshots/{feature}__{scenario}__{step_name}.png"
+        context.browser.driver.save_screenshot(filename)