Nti.contentfragments

==================

- Nothing changed yet.

==================

- Fix adapting base string input to plain text to behave more like 1.7
by only running the HTML to plain text algorithm if the input looks
like it may contain HTML markup. Note that in some instances where
characters like '<' were previously escaped to '&lt;', this will no
longer happen if the rest of the string doesn't look like HTML. See `issue 44
<https://github.com/NextThought/nti.contentfragments/issues/44>`_.
- Add schema fields ``VerbatimPlainText`` and
``VerbatimPlainTextLine`` to assume any incoming unicode value
already represents a plain text content fragment, instead of
(possibly) passing it through the HTML to plain text algorithm.

==================

- Add support for Python 3.9 and 3.10.
- Move to Github Actions from Travis CI.
- The algorithm for converting HTML to plain text has been changed and
produces higher quality output. For example, links are preserved in
a human-readable fashion. See `issue 39 <https://github.com/NextThought/nti.contentfragments/issues/39>`_.
- Fix an error getting link text when there was no link formatter
utility installed. See `PR 42
<https://github.com/NextThought/nti.contentfragments/pull/41#event-5422788226>`_.

==================

- Allow conversion of reStructuredText fragments to plain text.

==================

- Ensure disallowed tags nested within anchors do not raise.
See `issue 34 <https://github.com/NextThought/nti.contentfragments/issues/34>`_.

==================

- Add support for reStructuredText content fragments and corresponding
fields.