I am ok with renaming lazy_cache (although some people will ask, isn't the
cache lazy already by definition?). We cannot call it current_cache because
it is not current.cache and that would cause confusion. It is a lazy
current.cache.
On Wednesday, 25 July 2012 16:41:57 UTC-5, Anthony wrote:
>
> I'm sure technically there's some double laziness going on, but from an
> API standpoint, maybe it would be sufficient to just include one "lazy" in
> the name. :-) Another option might be current_cache.
>
> Anthony
>
> On Wednesday, July 25, 2012 5:25:36 PM UTC-4, Massimo Di Pierro wrote:
>>
>> Because while the cache decorator wraps the function once. The
>> lazy_lazy_cache wraps cache once, therefore the function twice. This is
>> because cache requires current.request but this is only available when the
>> function is called, not when the module is first imported and the cache
>> decorator executed. I am not attached to the name anyway, we can change it.
>>
>> On Wednesday, 25 July 2012 09:44:20 UTC-5, Anthony wrote:
>>>
>>> Why is it called lazy_lazy_cache, as opposed to just lazy_cache?
>>>
>>> On Wednesday, July 25, 2012 10:36:50 AM UTC-4, villas wrote:
>>>>
>>>> Hi All,
>>>>
>>>> For anyone who was interested in this thread, Massimo did implement
>>>> cache decorators for use *inside modules* and this is now available in
>>>> trunk. It seemed to work OK for me and for anyone else wishing to test
>>>> this, here is the description....
>>>>
>>>> Example Code for use inside your module:
>>>>
>>>> from gluon.cache import lazy_lazy_cache
>>>> @lazy_lazy_cache('key', time_expire=10, cache_model='ram')
>>>> def myfunction():
>>>> import time
>>>> return time.time()
>>>>
>>>> 'key' is optional
>>>> time_expire defaults to 300 seconds.
>>>> cache_model defaults to 'ram'.
>>>>
>>>> Note the system is the same as regular cache except that cache_model is
>>>> 'ram, not cache.ram. You cannot use current.cache because cache.ram only
>>>> exists per-request.
>>>>
>>>>
>>>>
>>>>
>>>> On Monday, July 16, 2012 4:05:38 AM UTC+1, villas wrote:
>>>>>
>>>>> Ok as you asked, I raised it as issue
>>>>> 895<http://code.google.com/p/web2py/issues/detail?id=895>so you can
>>>>> ponder on it.
>>>>> As I think this might be a tricky issue, I shall pass the cache in as
>>>>> Bruno suggested and I am happy with that solution. Please mark as
>>>>> wontfix
>>>>> if you don't see an obvious way to implement this.
>>>>> Thanks
>>>>>
>>>>>
>>>>> On Monday, July 16, 2012 2:33:46 AM UTC+1, Massimo Di Pierro wrote:
>>>>>>
>>>>>> Not yet. Please open a ticket about this.
>>>>>>
>>>>>> On Mar 1, 2011, at 5:05 AM, Carl Roach wrote:
>>>>>>
>>>>>> thanks Massimo.
>>>>>> The ability to have cache decorators in modules would be great.
>>>>>>
>>>>>> On 1 March 2011 02:09, Massimo Di Pierro
>>>>>> <[email protected]>wrote:
>>>>>>
>>>>>>> Jonathan and I have a plan to make this easy but it will not be in
>>>>>>> until 1.93 or 1.94.
>>>>>>>
>>>>>>> On Feb 28, 3:12 pm, pbreit <[email protected]> wrote:
>>>>>>> > Yeah, I'm having a hard time figuring out when and how to put
>>>>>>> stuff in
>>>>>>> > modules. And what the implications are of putting functions in
>>>>>>> /models.
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
--
