django-cleanhtmlfield

A simple Django app that will give you a Cleaned HTML Field.


License
MIT
Install
pip install django-cleanhtmlfield==1.0.2

Documentation

Django CleanHtmlField

image

Django CleanHtmlField is a simple Django application (supporting Django 1.8, 1.11, 2.0 and 2.1) that defines an HTMLField that automatically removes potentially malicious content. This app should work with Python 2.7+ and Python3.4+.

For instance, if you allow the user to freely input HTML Content, and the user decides to inject a JavaScript snippet:

<h1>Hello Friend</h1>
<script type="text/javascript">
    steal_all_passwords();
</script>
<p>This is for you!</p>

HTMLField will filter this to

<h1>Hello Friend</h1>
<p>This is for you!</p>

Quick start

  1. Download and install using pip install from PyPi:
pip install django-cleanhtmlfield
  1. Create a field HTMLField(strip_unsafe=True) in your model:
from django.db import models
from django_cleanhtmlfield.fields import HTMLField


class MyModel(models.Model)
    some_content = HTMLField(strip_unsafe=True)
  1. Don't forget to create and run migrations for changes on Django models, e.g.:
python manage.py makemigrations
python manage.py migrate

Requirements

For HTML content to be parsed and processed we depend on BeautifulSoup4 (this is installed as a dependency). As this is a Django app, it obviously requires Django (though we expect this to be already installed).

Optional: If you want a WYSIWYG Interface in your Admin Panel (or any other Django Form), you need the django-ckeditor package.

alt text

Field options

  • strip_unsafe (Default: False) needs to be set to True to enable stripping of unsafe HTML content
  • widget_form_class (Default: None) - allows overwriting the css form class for the widget (e.g., if you want to use django-ckeditor)

Configuration options

The following Django Settings are available (see below for a full example)

  • ACCEPTABLE_ELEMENTS - tuple that contains all allowed HTML tags (e.g., 'a', 'span', 'p', 'div', ...)
  • ACCEPTABLE_ATTRIBUTES - tuple that contains all allowed HTMl attributes (e.g., 'alt', 'style', 'target', 'title',...)
  • ACCEPTABLE_STYLES - tuple that contains all allowed CSS styles (e.g., 'background-color', 'border-color', 'font-size', ...)
  • REMOVE_WITH_CONTENT - tuple that contains potentially malicious HTML tags that will automatically be removed (e.g., 'script', 'object', ...)

Example:

ACCEPTABLE_ELEMENTS = (
    'a', 'abbr', 'acronym', 'address', 'area', 'aria-label', 'b', 'big',
    'blockquote', 'br', 'button', 'caption', 'center', 'cite', 'code', 'col',
    'colgroup', 'dd', 'del', 'dfn', 'dir', 'div', 'dl', 'dt', 'em',
    'font', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'hr', 'i', 'img',
    'ins', 'kbd', 'label', 'legend', 'li', 'map', 'menu', 'ol',
    'p', 'pre', 'q', 's', 'samp', 'small', 'span', 'strike',
    'strong', 'sub', 'sup', 'table', 'tbody', 'td', 'tfoot', 'th',
    'thead', 'tr', 'tt', 'u', 'ul', 'var', 'iframe', 'section', 'article',
)

ACCEPTABLE_ATTRIBUTES = (
    'abbr', 'accept', 'accesskey',
    'action', 'align', 'alt', 'axis', 'border', 'cellpadding', 'cellspacing',
    'char', 'charoff', 'charset', 'checked', 'cite', 'class', 'clear', 'cols',
    'colspan', 'color', 'compact', 'coords', 'data-mlang', 'data-equation', 'datetime', 'dir',
    'enctype', 'for', 'headers', 'height', 'href', 'hreflang', 'hspace',
    'id', 'ismap', 'label', 'lang', 'longdesc', 'maxlength', 'method',
    'multiple', 'name', 'nohref', 'noshade', 'nowrap', 'prompt',
    'rel', 'rev', 'rows', 'rowspan', 'role', 'rules', 'scope', 'shape', 'size', 'style',
    'span', 'src', 'start', 'summary', 'tabindex', 'target', 'title', 'type',
    'usemap', 'valign', 'value', 'vspace', 'width',
)

ACCEPTABLE_STYLES = (
    'background-color', 'background', 'background-image', 'background-position', 'background-size', 'background-repeat',
    'background-attachment', 'background-origin', 'background-clip',
    'font-family', 'font-size', 'font-weight', 'font-style', 'color',
    'width', 'height', 'min-width', 'max-width', 'min-height', 'max-height', 'line-height',
    'text-decoration', 'text-transform', 'text-align', 'border', 'border-style', 'border-width',
    'border-top', 'border-bottom', 'border-left', 'border-right', 'border-top-style',
    'border-bottom-style', 'border-left-style', 'border-right-style', 'border-top-width',
    'border-bottom-width', 'border-left-width', 'border-right-width',
    'border-color',
    'border-top-color', 'border-bottom-color', 'border-left-color', 'border-spacing', 'border-collapse',
    'border-right-color',
    'border-radius',
    'vertical-align', 'clear', 'float',
    'margin', 'margin-left', 'margin-right', 'margin-top', 'margin-bottom',
    'outline',
    'padding', 'padding-left', 'padding-right', 'padding-top', 'padding-bottom',

)

REMOVE_WITH_CONTENT = ('script', 'object', 'embed', 'style', 'form', )

Compatibility with Django REST Framework

If you are using Django Rest Framework you need to add the following code to register a serializer/field handler:

from django.utils.translation import ugettext_lazy as _

from rest_framework.serializers import ModelSerializer

from django_cleanhtmlfield.fields import HTMLField
from django_cleanhtmlfield.helpers import clean_html

class RestHtmlField(fields.CharField):
    default_error_messages = {
        'invalid': _('"{input}" is not a valid html.')
    }
    default_empty_html = False
    initial = False

    def __init__(self, **kwargs):
        super(HtmlField, self).__init__(**kwargs)

    def to_internal_value(self, data):
        return clean_html(data, strip_unsafe=True)


ModelSerializer.serializer_field_mapping[HTMLField] = RestHtmlField

Changelog

1.0.1

  • Renamed restricted field option to custom_css_class
  • Updated docs/readme

1.0

  • Initial Release

License

MIT License

Development and Tests

The test app is located in the tests subfolder. Please run tests against all Django and Python versions (e.g., using tox - see tox.ini).