html5lib: TypeError: __init__() got an unexpected keyword argument 'encoding'

Question

I'm trying to install html5lib. at first I tried to install the latest version (8 or 9 nines), but it came into conflict with my BeautifulSoup, so I decided to try older verison (0.9999999, seven nines ). I installed it, but when I try to use it:

>>> with urlopen("http://example.com/") as f:
    document = html5lib.parse(f, encoding=f.info().get_content_charset())

I get an error:

Traceback (most recent call last):
  File "<pyshell#11>", line 2, in <module>
    document = html5lib.parse(f, encoding=f.info().get_content_charset())
  File "C:\Python\Python35-32\lib\site-packages\html5lib\html5parser.py", line 35, in parse
    return p.parse(doc, **kwargs)
  File "C:\Python\Python35-32\lib\site-packages\html5lib\html5parser.py", line 235, in parse
    self._parse(stream, False, None, *args, **kwargs)
  File "C:\Python\Python35-32\lib\site-packages\html5lib\html5parser.py", line 85, in _parse
    self.tokenizer = _tokenizer.HTMLTokenizer(stream, parser=self, **kwargs)
  File "C:\Python\Python35-32\lib\site-packages\html5lib\_tokenizer.py", line 36, in __init__
    self.stream = HTMLInputStream(stream, **kwargs)
  File "C:\Python\Python35-32\lib\site-packages\html5lib\_inputstream.py", line 151, in HTMLInputStream
    return HTMLBinaryInputStream(source, **kwargs)
TypeError: __init__() got an unexpected keyword argument 'encoding'

What is wrong and what should I do?

Answer 1

I see something was broken in the latest versions of html5lib in regard to bs4, html5lib.treebuilders._base is no longer there, usng bs4 4.4.1 the latest compatible version seems to be the one with 7 nines, once you install it as below it works fine:

 pip3 install -U html5lib=="0.9999999"

Tested using bs4 4.4.1:

In [1]: import bs4

In [2]: bs4.__version__
Out[2]: '4.4.1'

In [3]: import html5lib

In [4]: html5lib.__version__
Out[4]: '0.9999999'

In [5]: from urllib.request import  urlopen

In [6]: with urlopen("http://example.com/") as f:
   ...:         document = html5lib.parse(f, encoding=f.info().get_content_charset())
   ...:     

In [7]: 

You can see the change in this commit Rename treebuilders._base to .base to reflect public status the name was changed:

The error you see is because you are still using the newest version, in html5lib/_inputstream.py, HTMLBinaryInputStream has no encoding arg:

class HTMLBinaryInputStream(HTMLUnicodeInputStream):
    """Provides a unicode stream of characters to the HTMLTokenizer.

    This class takes care of character encoding and removing or replacing
    incorrect byte-sequences and also provides column and line tracking.

    """

    def __init__(self, source, override_encoding=None, transport_encoding=None,
                 same_origin_parent_encoding=None, likely_encoding=None,
                 default_encoding="windows-1252", useChardet=True):

Setting override_encoding=f.info().get_content_charset() should do the trick.

Also upgrading to the latest version of bs4 works fine with the latest version of html5lib:

In [16]: bs4.__version__
Out[16]: '4.5.1'

In [17]: html5lib.__version__
Out[17]: '0.999999999'

In [18]: with urlopen("http://example.com/") as f:
             document = html5lib.parse(f, override_encoding=f.info().get_content_charset())
   ....:     

In [19]: