Fooster-web

Features
- Indication whether server is using SSL in HTTPServer.using_ssl
- Support for HTTP/1.1 Pipelining (which also makes requests a bit cleaner)
- Allow address reuse (meaning that it ignores TIME_WAIT on the address when starting the server)

Fixes
- handle_error uses log now
- Finish request_queue when stopping

Features
- Use worker threads in a thread pooling scheme to reduce thread overhead
- HTTPResponse has client_address as well
- Partial nose unit tests completed

Features
- Custom directory indexing in file.py
- Allow specifying of custom FileHandler (mostly for custom directory indexing)

Fixes
- No longer errors if MIME is unknown
- Giving a 403 where a 404 should be given for requests on a file as a directory
- file.py can now be successfully inited multiple times at different locations
- Headers cleared before a 500 given
- Fix file.py regex matching unintended things that could possibly be a security flaw
- Ensure headers are strings so that writing will fail less often
- Get rid of globals allowing multiple instances of HTTPServer

Features
- Full support for the HTTP/1.1 protocol including keep-alive
- Built-in HTTP error handling and size bounds
- SSL encryption support
- Threads for each connection allowing many simultaneous connections with built-in blocking for non-atomic access to a resource
- Easy creation of a RESTful resource by the implementation of HTTP methods
- OPTIONS and HEAD are automatically handled for resources
- Handling of 100-continue requests that check with the resource first
- Automatically get the request body for certain resources
- Ability to raise an HTTPError anywhere in a request and a guarantee that it will be handled properly
- Ability to create custom error handlers
- file.py, a built-in addon, allowing for quick hosting of files with simple directory indexing and an optional modification interface for files