Hi all, Just thought I'd throw in a couple of thoughts into the discussion -
Firstly you don't need to do 'myFuncRef.toString()' to get the java class name, just get myFuncRef, for example: <cfset key = hash(myFuncRef)/> Don't know if anyone's done it already, but an interesting benefit of this is that you can use them as a sort of pointers to functions. Another point is that the string you get from the function itself (i.e. [EMAIL PROTECTED]) changes with every change of that file because CF recompiles it. So if you changed the function that you're caching the data from, you'd end up with a new copy of the data for every change. If I were you, I'd use the location to the function and the function name because as mentioned earlier, it's probably going to be unique. Tim Blair wrote: > Evening, > > I'm in the process of creating an application, part of which is a series > of CFCs which perform and cache data, allowing quick retrieval on the > next call. This is done by passing a reference to a function to a > method of a CFC which then runs the function after doing some other > processing and returns the data. > > Now I want to be able to cache the data generated. The generated data > will be plain text and so I'm just planning on storing it (keyed on > something that uniquely identifies the funciton used) within an > structure as an instance variable in the CFC. > > So basically, I need a way to generate a structure key based on a > reference to a function. The way I've come up with is simply a hash of > the toString of the function, prefixed with an arbitrary string so the > variable name is valid, as so: > > <cfset this.KEYPREFIX = "i"> > <cfset key = this.KEYPREFIX & hash(myFuncRef.toString())> > > Which results in a key of "i8CF622BA70E703A796DB0FD8AC8ABFA8" -- a valid > variable name and unique to the function called. > > The myFuncRef.toString() returns an output similar to > "[EMAIL PROTECTED]" which is basically the > associated compiled java class. AFAIK this uniquely identifies the > function being used and means I don't have to manually associate an ID > with each function. The hash() + the string prefix call makes sure that > any invalid chars get removed and the chance of two hash() calls > returning the same value is supposedly around 1 in 2^64 so it should be > safe there... > > Basically, does anyone see any issue with doing it this way? > > Tim. > > -- Oliver Tupman Key Systems Geotechnical ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~| Archives: http://www.houseoffusion.com/lists.cfm?link=t:4 Subscription: http://www.houseoffusion.com/lists.cfm?link=s:4 Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4 Get the mailserver that powers this list at http://www.coolfusion.com
