Added HTML.from_html and HTML.from_url methods to create an HTML instance from an existing HTML source
Added HTML.codify method to produce Python code from an HTML instance. In combination with one of the HTML.from_ methods, can make Python code to reproduce existing HTML similar to what you need.
Added ability to handle 'blank' attrs like 'nowrap' using e.g. nowrap=None attribute.
Added class_ as alternative to klass in avoiding Python's class keyword
Release notes from html3 and html libaries
1.18 Fixed support of python 3.8
1.17 First release of html3
1.16 detect and raise a more useful error when some WSGI frameworks
attempt to call HTML.read(). Also added ability to add new content using
the += operator.
1.15 fix Python 3 compatibility (unit tests)
1.14 added plain XML support
1.13 allow adding (X)HTML instances (tags) as new document content
1.12 fix handling of XHTML empty tags when generating unicode
output (thanks Carsten Eggers)
1.11 remove setuptools dependency
1.10 support plain ol' distutils again
1.9 added unicode support for Python 2.x
1.8 added Python 3 compatibility
1.7 added Python 2.5 compatibility and escape argument to tag
construction