Templating

In normal Django fashion, templates are used to render various data to the web browser. We choose to implement a “central” template directory, in templates/ This is configured in openl2m/settings.py as:

TEMPLATES = [
  {
      ...
      'DIRS': [
          os.path.join(BASE_DIR, 'templates'),
      ],
      'APP_DIRS': True,
      ...
  }

Note that we still allow templates in app-specific sub-directories (APP_DIRS = True).

From switches/view.py we call then call these templates to render the pages.

The base.html template

This is the main template that defines the screen layout of all others. It is included in all other templates. This template defines the HTML headers, includes the various script files, and defines the template tags {% block content %}{% endblock %}

This tag that can be overwritten in other templates, and is the placeholder for the content in individual pages.

Static files

We store static images, javascript, etc in /project-static/. See Django docs on static files here.

The upgrade.sh script moves these to the /static/ location as needed.

We use a “context preprocessor” to get access to Python variables inside the template engine. In the settings file, we add the following function to map those variables:

TEMPLATES = [
  ...
  'switches.context_processors.add_variables',
  ...
  ]

This means we use the switches/context_processors.py file, and the add_variables() function map data from Python to templating.

Template tags

In some templates, we use ‘template tags’. Simply put, these are python functions that can be called from the template engine. Template tags are stored in switches/templatetags/helpers.py and can be included in templates with {% load helpers %} Note that each template needs to load the helper!

See more about template tags here.

We use several important functions to get strings rendered in the helpers files. Notably are the get_*_nms_link() functions, where we have to render based on objects passed from the template, and strings that are django templates. We create the html using render().

Additionally, get_lldp_capabilities_icon() produces html code, and we pass that back via mark_safe() to make sure it does not get marked up, i.e. gets interpreted as html.