cannot import name 'sanitizer'

[erikzenker]

I want install the django cms which uses the sanitizer of the html5lib,
but I get the error shown above in the title.

when I just import the sanitizer from html5lib:

from html5lib import sanitizer

I get the following error:

Traceback (most recent call last):
  File "test.py", line 1, in <module>
    from html5lib import sanitizer
ImportError: cannot import name 'sanitizer'

I installed html5lib from pip:

pip install html5lib

Is the sanitizer module not available anymore ?

[stefanfoulis]

@erikzenker It looks like it has been removed (or moved). Quickfix is to use a slightly older version of the package where sanitizer is still there:

pip install html5lib<0.99999999

ot: what is it with those weird version numbers anyway? You know you could also make it 0.10 .

[gsnedders]

That's #72 and #110. The sanitizer now exists only as a filter for the treewalkers, rather than something touching the tokenizer in odd ways that was likely broken.

Our state of terrible documentation probably doesn't help here, and the changelog entry is rather useless in hindsight. The common-case is something like:

from html5lib import parse, serialize

def clean(s):
    d = parse(s)
    return serialize(s, sanitize=True)

(0.10 is less than 0.9999999, so no, that option's long gone.)

[macolo]

You could make it 1.0 though. Counting the 9s makes me dizzy.

[DanielTate]

This is still not working.... Are there any real solutions to this?

[vhquang]

As a2917e9, sanitizer is moved into filters. So to import (as of version 1.1), you would do:

from html5lib.filters import sanitizer