For a long time I wanted to write an article on how to write auto populate field functionality for a Django project. This is a very useful feature that allows you to change the content of other model fields in Django by setting a value to a field that uses auto populate.
First, why is it needed? - Such functionality allows you to reduce the size of the code in those places where you need to rewrite other fields of the object when they change. That is, for example, you do not have to redefine the save method each time in order to rewrite some field in case other fields of the object change. Also, using auto populate is basically a more advanced and neat way to manage data models in Django.
And also a similar approach can solve some problems and improve the site.
For example, I use this approach on this site for writing the content of comments and articles, and in general in all fields where the markdown syntax is used.
- All data models have two fields content and content_markdown . The content_markdown field controls the content of the content field. That is, in a normal situation, the user does not have access to edit the contents of the content field, he always edits only the content_markdown field.
- The user, when creating a message, edits the content_markdown field, which automatically generates html code from the markdown markup and writes html in the content field.
Everyone chooses for himself why he should use auto populate. For example, you can automatically generate a thumbnail image from a source image. Such functionality is ideal for such actions.
And specifically in my case, it turned out that I did not find a suitable functionality for generating html from markdown. All I found at that time were libraries that generated html at runtime on every page load using template tags. And this approach greatly reduces the performance of the site. Therefore, I decided for myself that I was ready to sacrifice disk space for the sake of website performance. After all, a tenfold increase in page load speed is in some cases clearly preferable to double the amount of disk space occupied.
I will present a simplified MarkdownField code that you could use on your site.
MarkdownField will inherit from the standard TextField and will take as an argument the name of the field to use for html content. This is the html_field argument.
This field has a set_html method which is responsible for generating html from the markdown markup and also setting the value in the html field of the content.
The important point is to connect the set_html method to the pre_save signal in the contribute_to_class method. An earlier version I developed had the disadvantage that every time an object was requested, the html-generating method was called. It was a bug that I subsequently discovered and corrected in a similar way. Now the content is generated only if the object is saved.
# -*- coding: utf-8 -*- from django.db import models from django.db.models.signals import pre_save from .utils import MarkdownWorker class MarkdownField(models.TextField): """ This field save markdown text with auto-populate text to html field. This field must be used with second text field for html content. """ def set_html(self, instance=None, update_fields=None, **kwargs): value = getattr(instance, self.attname) if value and len(value) > 0: instance.__dict__[self.html_field] = MarkdownWorker(value).get_text() def contribute_to_class(self, cls, name, **kwargs): super().contribute_to_class(cls, name, **kwargs) pre_save.connect(self.set_html, sender=cls) def __init__(self, html_field=None, *args, **kwargs): self.html_field = html_field super().__init__(*args, **kwargs)
This class is responsible for generating html content from markdown markup. In this class, you can add any additional types of processing from html to get the finished result you need. For example, you can remove links or images that you don't think the user can add.
When creating an object of the MarkdownWorker class, the text to be processed is set, and then a method is called that converts the text to html.
Also in my case, extensions are used for:
- Table generation
- Code support
- Transfer Additions
- Upper and lower syntax
# -*- coding: utf-8 -*- import markdown class MarkdownWorker: """ Markdown converter. It will convert markdown text to html text """ def __init__(self, text): self.pre_markdown_text = text self.markdown_text = None self.make_html_from_markdown() def make_html_from_markdown(self): if self.pre_markdown_text: self.markdown_text = markdown.markdown( self.pre_markdown_text, extensions=['markdown.extensions.attr_list', 'markdown.extensions.tables', 'markdown.extensions.fenced_code', 'markdown.extensions.nl2br', 'superscript', 'subscript'], output_format='html5' ) def get_text(self): return self.markdown_text
To support all this functionality, you will need to install the libraries:
Maybe also beautifulsoup4 , or it will be installed in dependencies, but honestly, I don’t remember.
And now let's apply this field in the data model.
We have a Post model that contains the fields:
- user - foreign key per user
- content - field that will contain html content
- content_markdown - the field where the markdown text will be saved with subsequent generation of html text. In doing so, you can see that the html_field='content' argument is passed to it, which tells which field the MarkdownField is to work with.
# -*- coding: utf-8 -*- from django.conf import settings from django.db import models from django.utils.translation import gettext_lazy as _ from .fields import MarkdownField class Post(models.Model): user = models.ForeignKey(to=settings.AUTH_USER_MODEL, on_delete=models.CASCADE) content = models.TextField(verbose_name=_('Description')) content_markdown = MarkdownField(verbose_name=_('Description - Markdown'), html_field='content')
This way you can add markdown fields to your site that will automatically generate html content.
You will now have support for fancy markdown syntax on your site, which will automatically generate html content.