 |
C++Talk.NET
C++ language newsgroups
|
| View previous topic :: View next topic |
| Author |
Message |
francis_r
Guest
|
 Posted: Tue Jul 25, 2006 5:25 am Post subject: What do you think of this caching optimization strategy? |
 |
|
Hello newsgroup readers,
I would like to know if the caching strategy used in code below is a
good idea. The code explains my question... (It is a simplified version
of what I'm currently working on.)
class X
{
// Datastructure Action
struct Action {
enum Type {
Type_A,
Type_B,
Type_C // etc...
};
int mID;
Type mType;
std::string mDescription;
};
// Collection of actions
std::vector< Action > mActions; // contains many (say 500+) actions
// Extract all ActionIDs for a given ActionType
const std::vector< int >& GetActionIDs( Action::Type inActionType )
{
// Use caching
static std::map< Action::Type, std::vector< int > > fCache;
// First search cache
std::map< Action::Type, std::vector< int > >::iterator theIt =
fCache.find( inActionType );
// If ActionType not found in cache then iterate all Actions
if( theIt == fCache.end() )
{
std::vector< int > theActionIDs;
for( int theIndex = 0; theIndex != mActions.size(); ++theIndex )
{
if( mActions[ theIndex ].mType == inActionType )
{
theActionIDs.push_back( mActions[ theIndex ].mID );
}
}
// Store result in cache
fCache[ inActionType ] = theActionIDs;
// Return result from cache ( so we can return a reference)
return fCache[ inActionType ];
}
// Else immediately return the result from the cache
return theIt->second;
}
};
I have recently been writing code like this. Do you think it's a good
practice?
Greetings,
Francis
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ] |
|
| Back to top |
|
 |
francis_r
Guest
|
| Posted: Wed Jul 26, 2006 12:57 am Post subject: Re: What do you think of this caching optimization strategy? |
|
|
| Quote: | // Store result in cache
fCache[ inActionType ] = theActionIDs;
This statement doesn't check if 'theActionIDs' is empty and will
create an empty entry in the cache. Is that acceptable?
|
Yes, I see no reason why not. If there are no actions for a certain
ActionID then the code will return the empty vector from the dict. This
seems logical to me.
| Quote: | // Return result from cache ( so we can return a reference)
return fCache[ inActionType ];
}
// Else immediately return the result from the cache
return theIt->second;
}
};
I have recently been writing code like this. Do you think it's a good
practice?
It seems incomplete. How do you clear the cache and force it to be
re-collected if the contents of 'mActions' change?
|
That's a valid concern. But in my case the mActions never changes at
runtime, so that won't pose a problem.
I also forgot to mention that the class X is a singleton. If it wasn't
then the same cache would be shared among all objects. That would
probably not be desirable ;)
| Quote: | If for some reason 'mActions' is a "grow only" container, your scheme
can work, but it still seems (a) underdesigned and (b) wasteful because
it stores copies of vectors in the cache instead of, say, pointers to
the vectors.
|
The copy waste can indeed be reduced by using pointers.
Thanks for your feedback.
Francis
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ] |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
