On 04/24/2015 08:13 AM, George Neuner wrote:
I'm not using parameterize at all (at least not explicitly). I
receive a web request that has up to 9 arguments contained in its
bindings.
(define (search request)
(let* [
(params (request-bindings request))
(cookies (request-cookies request))
(pid (find-cookie "pid" cookies))
(county (exists-binding? 'county params))
(state (exists-binding? 'state params))
(search-txt (exists-binding? 'query params))
(search-and (exists-binding? 'all params))
(search-or (exists-binding? 'any params))
(search-not (exists-binding? 'none params))
(page-size (exists-binding? 'page params))
(sort (exists-binding? 'sort params))
:
Using those arguments, I retrieve information from the database
regarding the client that issued the request and construct a complex
database query. When that query succeeds, I need to enter additional
accounting data for each result based on the client - but only for
results that the client actually sees (paging), not necessarily for
every result.
So I need to maintain a lot of state through a continuation: the open
database connection [to cache the query results in the database], the
current page that the client is looking at (for navigation), and
ideally the client information for accounting [though I could look
that up again if necessary].
George
My intuition is the more state you have, the better the continuation
model serves you, as you don't have to pass and manage all that state
explicitly through cookies, url params, etc. It sounds like you are not
happy with the continuation model. You can of course use the "regular"
web patterns in a racket web service, but you'll still have many of the
same issues. Even in a non-continuation based service, you'll need some
way of maintaining a client->database connection mapping, and find some
way of removing stale db connections once the client is "gone", which
you can never know for sure.
Are you trying to decide between using continuations and regular web
state patterns?
Thanks,
Dave
--
You received this message because you are subscribed to the Google Groups "Racket
Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
