Disable swagger document caching when flask debug mode is on

[strongbugman]

No description provided.

[coveralls]

Coverage remained the same at 91.184% when pulling e5b19f0 on strongbugman:feature/debug_without_caching into 20a960b on rochacbruno:master.

[javabrett]

Does this make an assumption that app is always set and is a Flask WSGI server?

I expect that in 90%+ of dev/debug situations, Flask will be being used and the dev/debug server enabled, which enables Flask's own reload/cache-busting, making this valid. But what if someone wanted to do this under another WSGI server e.g. Gunicorn with reload enabled?

Would it be better to provide Flasgger with a direct debug flag, which if not-specified and running under Flask would inherit from the Flask debug flag? That way it could be set if the Flask assumption was not true.

I haven't run Flasgger under Gunicorn, and my comments might be astray as a result.

[strongbugman]

Yes, Gunicorn with reloading won't enable flask's debug mode, but Gunicorn's reloading is not enough if someone wants to run flask app with whole debug mode. Gunicorn also provides an extra reloading If someone needs reloading only.

And it'll increase complexity if we provide an extra debug flag, imagine that the flask's debug is enabled while the flasgger's debug is closed.
So, in my opinion, it's enough if flasgger can be responsible for the flask, and other WSGI servers can active flask's debug mode too (FLASK_ENV or something else).

[javabrett]

Thanks @strongbugman .