joegasewicz / bobtail Goto Github PK

Add the following metadata to the Request object:

• protocol
• domain
• port

WSGI environ dict URI metadata

'wsgi.url_scheme': 'http',  
  'SERVER_NAME': '127.0.0.1', 
  'SERVER_PORT': '8000',

Further reading: https://peps.python.org/pep-3333/#url-reconstruction

Link issue #24

e.g

res.set_str()

The doc strings need to be completed as many are missing / incomplete

• I will keep this issue open until we get 100% doc coverage

This should only raise a RouteClassError if the request method's class is uninitiated and not swallow other exceptions.

Duplicate the sample test and add one for each of the Status.codes

Check if vars has been assigned

  File "/Users/joegasewicz/PycharmProjects/bobtail/bobtail/wsgi.py", line 91, in __call__
    self._handle_route()
  File "/Users/joegasewicz/PycharmProjects/bobtail/bobtail/wsgi.py", line 63, in _handle_route
    self.request.set_args(self.parse_metadata["vars"])

This issue should add the following

• A default templating engine that should be able to render html files and parse variables and some basic logic. An improved version should follow with full AST parsing.
• Middleware for Jinja2 (https://github.com/joegasewicz/bobtail-jinja2)
• Include security for string manipulation e.g CSRF
  Link issue #60

If there are no arguments in the URI path & req.get_arg is called then RouteClassError is raised.

Fix:
provide a new custom exception for this type of error.

This will be implemented in an external library, possibly in Rust or C & will follow the
https://www.rfc-editor.org/rfc/rfc3986 ref.

• Handle multiple query params in URI ✅ fixed in #9
• Handle # - will be fixed in - #9
• Create binding for this C lib - https://github.com/joegasewicz/route-parser

Support wsgi.multithread=True

Create a form api that improves access to form data in the request

[21-04-23 20:13:10] GET / 200
[2023-04-21 20:13:10 +0100] [14073] [ERROR] Error handling request /favicon.ico
Traceback (most recent call last):
File "/home/joe/.local/share/virtualenvs/main_app-NxXa5jZN/lib/python3.11/site-packages/gunicorn/workers/sync.py", line 136, in handle
self.handle_request(listener, req, client, addr)
File "/home/joe/.local/share/virtualenvs/main_app-NxXa5jZN/lib/python3.11/site-packages/gunicorn/workers/sync.py", line 179, in handle_request
respiter = self.wsgi(environ, resp.start_response)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/joe/.local/share/virtualenvs/main_app-NxXa5jZN/lib/python3.11/site-packages/bobtail/wsgi.py", line 126, in call
self._handle_route()
File "/home/joe/.local/share/virtualenvs/main_app-NxXa5jZN/lib/python3.11/site-packages/bobtail/wsgi.py", line 94, in _handle_route
self.parse_metadata = p.route()
^^^^^^^^^
File "/home/joe/.local/share/virtualenvs/main_app-NxXa5jZN/lib/python3.11/site-packages/bobtail/parser.py", line 125, in route
route = self._match()
^^^^^^^^^^^^^
File "/home/joe/.local/share/virtualenvs/main_app-NxXa5jZN/lib/python3.11/site-packages/bobtail/parser.py", line 60, in _match
if route_segment[0] != "{" and route_segment != path_segment:
~~~~~~~~~~~~~^^^
IndexError: string index out of range

Examples using SQLAlchemy Core:

data = req.get_multipart_data()

ins = images.insert().values(
            user_id=int(data["user_id"]["value"]),
            file_name=data["logo"]["value"]["filename"],
            noticeboard_id=int(data["noticeboard_id"]["value"]),
        )

Create some test utilities

Make the path value available through a method.

For example:

req.get_path()

And update the README.md with an example

if route_segment[0] != "{" and route_segment != path_segment: ~~~~~~~~~~~~~^^^

Although the code is compatible with 3.10 the tests aren't currently running on 3.10

https://www.rfc-editor.org/rfc/rfc9110.html#name-options

We have a method to get the filename but not the actual file value.
See #49