use beautifulsoup to parse the description from the first paragraph

Signed-off-by: Brian S. Stephan <bss@incorporeal.org>

CHANGELOG.md

@@ -2,22 +2,6 @@
 
 Included is a summary of changes to the project, by version. Details can be found in the commit history.
 
-## v2.1.0
-
-### Features
-
-* The page title (also used in the `og:title` header) and the optional description used in the `og:description` header
-  can be derived from the contents of the page content, if the markdown meta tags are not supplied. The first `h1` is
-  used for the title, and the first `p` is used for the description. This is largely to save some time writing pages
-  that one wants to look nice, especially in a social media card, and removes some repetition.
-
-### Miscellaneous
-
-* Requirements bumped, which led to...
-* Python 3.9 has been removed from the supported versions.
-* Added some miscellaneous unit tests and coverage changes to keep us at 95% (which only dropped for a library reason I
-  don't understand).
-
 ## v2.0.5
 
 ### Features

README.md

@@ -7,8 +7,8 @@ A lightweight static site generator for Markdown-based sites.
 Something like the following should suffice:
 
 ```
-% virtualenv --python=python3.10 env-py3.10
-% source env-py3.10/bin/activate
+% virtualenv --python=python3.9 env-py3.9
+% source env-py3.9/bin/activate
 % pip install -U pip
 % pip install incorporeal-cms
 % incorporealcms-build ./path/to/instance ./path/to/output/www/root