In case anyone was wondering I worked out what was going on and it makes 
perfect sense. I was being stupid :)

If you ask for a value via 'lookup' expired values will not be returned at 
this point as 'has?' is called internally. 

I made a quick and dirty library to reflect the behaviour I want for my 
particular use case: https://github.com/danstone/clj-refresh-cache

On Friday, February 28, 2014 9:47:14 PM UTC, [email protected] wrote:
>
> Taking the code below, if I repeatedly read from a cache I will 
> occasionally get nil back from the cache. This seems to happen on my 
> machine for ttl's under 5 ms.
> Although I'm sure I would never use such a short TTL in the wild, I would 
> like to know why I am seeing this... Has anybody else noticed this issue? 
>
> (ns cache-test.core
>   (:require [clojure.core.cache :as cache]))
>
> (def data (atom "foo"))
>
> (def ttl
>   (atom (cache/ttl-cache-factory {} :ttl 1)))
>
> (defn get-data
>   []
>   (let [c @ttl
>         nc (if (cache/has? c {})
>             (cache/hit c {})
>             (cache/miss c {} @data))]
>     (reset! ttl nc)
>     (cache/lookup nc {})))
>
> (def results (atom '()))
>
> ;; there will *sometimes* be
> ;; nil elements in the seq, seems dependent on cpu speed, which would make 
> sense due to side effect of taking system/currentTimeMillis internally.
> (set (map (fn [_] (get-data)) (range 10000)))
>

-- 
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to [email protected]
Note that posts from new members are moderated - please be patient with your 
first post.
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"Clojure" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to