libfaster API Documentation  Development Version
Super fast distributted computing
Memory Model

Description

Automatic memory deallocation.

In order to allow for operator chains like this:

...
int result = someFdd -> map(&myMap) -> flatMap(&myFlatMap) -> reduce(&myReduce);
...

a automatic memory deallocation model was adopted. If a user apply some operators to a dataset, its distributed memory will be deallocated. In order to use a dataset more than once, the user needs to protect his dataset with the cache() function and discard its content once it is done with the discard() function.

Returns
pointer to self

Functions

void faster::fddCore< T >::discard ()
 deallocates previusly cached fdd
 
fdd< T > * faster::fdd< T >::cache ()
 Prevents automatic memory deallocation from hapenning. More...
 
groupedFdd< K > * faster::groupedFdd< K >::cache ()
 Prevents automatic memory deallocation from hapenning. More...
 
void faster::groupedFdd< K >::discard ()
 deallocates previously cached fdd
 
void faster::iFddCore< K, T >::discard ()
 deallocates previously cached FDD
 
indexedFdd< K, T > * faster::indexedFdd< K, T >::cache ()
 Prevents automatic memory deallocation from hapenning. More...
 

Function Documentation

§ cache() [1/3]

template<typename K>
groupedFdd<K>* faster::groupedFdd< K >::cache ( )
inline

Prevents automatic memory deallocation from hapenning.

Returns
pointer to the cached dataset (self)

Definition at line 118 of file groupedFdd.h.

§ cache() [2/3]

template<class T>
fdd<T>* faster::fdd< T >::cache ( )
inline

Prevents automatic memory deallocation from hapenning.

Returns
pointer to the cached dataset (self)

Definition at line 150 of file fdd.h.

§ cache() [3/3]

template<typename K, typename T>
indexedFdd<K,T>* faster::indexedFdd< K, T >::cache ( )
inline

Prevents automatic memory deallocation from hapenning.

Returns
pointer to the cached dataset (self)

Definition at line 222 of file indexedFdd.h.