I am happy to announce the release of CherryPy-2.1.0
It is the result of 6 months of intense development since the
last stable release and the work of a growing number of
CherryPy has become increasingly popular these past few months
(the mailing lists now have more than 500 people) and it is
also used in other popular products such as Subway and Turbogears.
This release is a major step forward for CherryPy. It is packed
with new features and bug fixes.
Here are the main improvements in this release:
- New WSGI interface, which allow CherryPy sites to be deployed on
any WSGI server. People are already running it on mod_python, FastCGI,
SCGI, IIS or CherryPy's own built-in HTTP server.
- New implementation for sessions, which supports multiple backends
- Built-in list of convenient "filters" for things like gzip
XHTML validation, caching, unicode decoding/encoding, authentication,
XML-RPC wrapper, etc ... These filters can easily be enabled/disabled
- New "development" mode which provides things like autoreload (no
need to manually restart your server when you make a change), logging
of page stats, etc ...
- Better handling of file uploads
- Internal implementation now uses generators everywhere (no more
- New built-in HTTP server implementation
CherryPy-2 is a pythonic, object-oriented web development framework.
- this code starts a multi-threaded HTTP server that dispatches
requests to methods
- requests like "http://domain/dir/page?arg1=va l1&arg2=val2" are
mapped to "dir.page(arg1='val1', arg2='val2')"
- CherryPy also supports "RESTful" URLs like
http://domain/book/science/9 - requests are mapped to an object tree that is "mounted" on
(for instance: "cherrypy.root.user", "cherrypy.root.user.remi", ...)
- method must be explicitly exposed with a decorator "@cherrypy.expose"
(or "index.exposed = True" for Python-2.3)