htmllaundry 1.8

Simple HTML cleanup utilities

Introduction

This package contains several handy python methods to cleanup HTML markup or perform other common changes. The cleanup is strict enough to be able to clean HTML pasted from MS Word or Apple Pages. This package also contains integration code for z3c.form to provide fields that automatically sanitize HTML on save.

The implementation is based on the Cleaner class from lxml.

Cleanup routines

All cleanup routines can be invoked through the single sanitize function. This functions takes an input string as input and will return a cleaned up version of that string. Here is a simple example:

>>> from htmllaundry import sanitize
>>> sanitize('Hello, <em>world</em>')
'<p>Hello, <em>world</em></p>'

The sanitize method takes an extra optional parameter with a lxml Cleaner instance, which can be used to use different filtering rules. htmllaundry includes three cleaners:

• htmllaundry.cleaners.DocumentCleaner, which is the default cleaner. This cleaner will allow most safe tags, while stripping out inline styles and insecure markup.

• htmllaundry.cleaners.LineCleaner is a more strict cleaner which only allows a few inline elements. This is useful in places where you only want to accept single-line input, for example in document titles.

• htmllaundry.cleaners.CommentCleaner only allows a very limited set of HTML elements, and is designed to be useful for user provided comments. It will also force all external links to open in a new browser window.

If you want to go all the way you can also use StripMarkup to strip all markup from your input:

>>> from htmllaundry import StripMarkup
>>> StripMarkup('Hello, <em>world</em>')
'Hello, world'

z3c.form integration

If you want to use the z3c.form integration you should use the z3cform extra for this package:

install_requires=[
     ....
     htmllaundry [z3cform]
     ...
     ],

In addition you will need to load the ZCML. In your configure.zcml add a line like this:

<include package="htmllaundry" />

You can then use the HtmlText field type in your schemas. For example:

from zope.interface import Interface
from zope import schema
from htmllaundry.z3cform import HtmlText

class IDocument(Interface):
    title = schema.TextLine(
            title = _(u"Title"),
            required = True)

    description = HtmlText(
            title = _(u"Description"),
          required = True)

Please note that using HtmlText will not automatically give you a WYSYWIG widget.

Changelog

1.8 - November 30, 2010

• Remove link target enforcement from hardcoded code path from sanitize. This makes it possible to use the new link_target cleaner option.

1.7 - November 30, 2010

• Make forcing of target attributes on externals linke configurable via a new link_target option in the cleaners. Only enable this option for the CommentCleaner.

1.6 - November 18, 2010

• Correct whitespace test for wrapping bare text as well.

1.5 - November 18, 2010

• Correct whitespace checks to handle all unicode whitespace. This fixes problems with xA0 (or   in HTML-speak) being treated as text.

1.4 - August 3, 2010

• Small code cleanup.

• Strip leading breaks.

1.3 - July 30, 2010

• Strip all top level br elements. Breaks are fine in blocklevel elements, but should not be used to add vertical spacing between block elements.

1.2 - February 15, 2010

• Fix a typo in the documentation.

• Strip trailing breaks.

1.1 - February 5, 2010

• Add a simple StripMarkup method.

• Add ZCML necessary for z3c.form integration.

1.0 - February 5, 2010

• First release