docs: add middlewares documentation

docs/contributing/index.rst

@@ -171,6 +171,11 @@ The Scrapy framework is very flexible. To maintain a good separation of concerns
    -  Connect signals, typically `item signals <https://docs.scrapy.org/en/latest/topics/signals.html#item-signals>`__ and `spider signals <https://docs.scrapy.org/en/latest/topics/signals.html#spider-signals>`__
    -  Raise a `NotConfigured <https://docs.scrapy.org/en/latest/topics/exceptions.html#notconfigured>`__ exception in its `from_crawler <https://docs.scrapy.org/en/latest/topics/extensions.html#writing-your-own-extension>`__ method, if a required `setting <https://docs.scrapy.org/en/latest/topics/settings.html>`__ isn't set
 
+-  A downloader middleware's responsibility is to process requests, before they are sent to the internet, and responses, before they are processed by the Spiders. It should only:
+   -  Yield a request
+   -  Return a Deferred
+   -  Yield items
+
 When setting a custom `Request.meta key <https://docs.scrapy.org/en/latest/topics/request-response.html#scrapy.http.Request.meta>`__, check that the attribute name isn't `already in use <https://docs.scrapy.org/en/latest/topics/request-response.html#topics-request-meta>`__ by Scrapy.
 
 Update requirements
@@ -193,3 +198,4 @@ API reference
    extensions.rst
    util.rst
    exceptions.rst
+   middlewares.rst

docs/contributing/middlewares.rst

@@ -0,0 +1,6 @@
+Middlewares
+===========
+
+.. automodule:: kingfisher_scrapy.middlewares
+   :members:
+   :undoc-members: