1. wsgi: Interaction standard between webserver and app
1. Enter env,
2、callback(status header),
3. Return the body in the form of an array
def app(environ, start_response):
start_response('200 OK', [('Content-Type', 'text/plain')])
yield "Hello world!\n"
Second, the main process of processing
1. All applications are unified by django/core/handlers/wsgi.py WSGIHandler
2. WSGIHandler internally divides the entire processing into several stages
1" load_middleware (once during initialization)
request_middleware
self._view_middleware = []
# The last 3 are from back to front
self._template_response_middleware = []
self._response_middleware = []
self._exception_middleware = []
2" trigger the callback signals.request_started.send(sender=self.__class__, environ=environ)
In fact, only the db part will be disconnected and reconnected
3" Construct WSGIRequest
4" get_response ----- main process
1. Call back process_request in turn, and interrupt if there is a return value, otherwise execute the next one
2. Find the view according to the url
3. Call process_view in turn
4. The specific callback, ----- must return the response
5. If there is a render, apply process_template_response (before using the template, you can modify the template related logic)
6. Use process_response ---- after generating the response
7. When there is an exception in this process, trigger process_exception
简单说：process_request parseUrl process_view callback process_template_response render process_response
3. How to keep db connection alive? Directly do not keep alive
django.db.backends.base.base.py BaseDatabaseWrapper get_new_connection
django.db.models.sql.compiler.py execute_sql
The default is to open a new database connection for each request, and then close (stateless). If you want to maintain it, pay attention to setting the state and quantity limit of the connection
Third, the directory structure of a typical app
manager.py -- specify the setting file
project
settings
urls
sleep
tests
app
models
views
urls
apps
admin
static
template