paste.urlparser – Handle URL paths and server static files

WSGI applications that parse the URL and dispatch to on-disk resources

Module Contents

class paste.urlparser.StaticURLParser(directory, root_directory=None, cache_max_age=None)

Like URLParser but only serves static files.

cache_max_age:

integer specifies Cache-Control max_age in seconds

paste.urlparser.make_static(global_conf, document_root, cache_max_age=None)

Return a WSGI application that serves a directory (configured with document_root)

cache_max_age - integer specifies CACHE_CONTROL max_age in seconds

class paste.urlparser.PkgResourcesParser(egg_or_spec, resource_name, manager=None, root_resource=None)

paste.urlparser.make_pkg_resources(global_conf, egg, resource_name='')

A static file parser that loads data from an egg using pkg_resources. Takes a configuration value egg, which is an egg spec, and a base resource_name (default empty string) which is the path in the egg that this starts at.

class paste.urlparser.URLParser(global_conf, directory, base_python_name, index_names=<class 'paste.urlparser.NoDefault'>, hide_extensions=<class 'paste.urlparser.NoDefault'>, ignore_extensions=<class 'paste.urlparser.NoDefault'>, constructors=None, **constructor_conf)

WSGI middleware

Application dispatching, based on URL. An instance of URLParser is an application that loads and delegates to other applications. It looks for files in its directory that match the first part of PATH_INFO; these may have an extension, but are not required to have one, in which case the available files are searched to find the appropriate file. If it is ambiguous, a 404 is returned and an error logged.

By default there is a constructor for .py files that loads the module, and looks for an attribute application, which is a ready application object, or an attribute that matches the module name, which is a factory for building applications, and is called with no arguments.

URLParser will also look in __init__.py for special overrides. These overrides are:

urlparser_hook(environ)

This can modify the environment. Its return value is ignored, and it cannot be used to change the response in any way. You can use this, for example, to manipulate SCRIPT_NAME/PATH_INFO (try to keep them consistent with the original URL – but consuming PATH_INFO and moving that to SCRIPT_NAME is ok).

urlparser_wrap(environ, start_response, app):

After URLParser finds the application, it calls this function (if present). If this function doesn’t call app(environ, start_response) then the application won’t be called at all! This can be used to allocate resources (with try:finally:) or otherwise filter the output of the application.

not_found_hook(environ, start_response):

If no file can be found (in this directory) to match the request, then this WSGI application will be called. You can use this to change the URL and pass the request back to URLParser again, or on to some other application. This doesn’t catch all 404 Not Found responses, just missing files.

application(environ, start_response):

This basically overrides URLParser completely, and the given application is used for all requests. urlparser_wrap and urlparser_hook are still called, but the filesystem isn’t searched in any way.

paste.urlparser.make_url_parser(global_conf, directory, base_python_name, index_names=None, hide_extensions=None, ignore_extensions=None, **constructor_conf)

Create a URLParser application that looks in directory, which should be the directory for the Python package named in base_python_name. index_names are used when viewing the directory (like 'index' for 'index.html'). hide_extensions are extensions that are not viewable (like '.pyc') and ignore_extensions are viewable but only if an explicit extension is given.