Interface Cache

All Known Implementing Classes:
AbstractValueAdaptingCache, CaffeineCache, ConcurrentMapCache, JCacheCache, NoOpCache, TransactionAwareCacheDecorator

public interface Cache
Interface that defines common cache operations.

Serves primarily as an SPI for Spring's annotation-based caching model (Cacheable and co) and secondarily as an API for direct usage in applications.

Note: Due to the generic use of caching, it is recommended that implementations allow storage of null values (for example to cache methods that return null).

Since:
3.1
Author:
Costin Leau, Juergen Hoeller, Stephane Nicoll
See Also:
  • Nested Class Summary

    Nested Classes
    Modifier and Type
    Interface
    Description
    static class 
    Wrapper exception to be thrown from get(Object, Callable) in case of the value loader callback failing with an exception.
    static interface 
    A (wrapper) object representing a cache value.
  • Method Summary

    Modifier and Type
    Method
    Description
    void
    Clear the cache through removing all mappings.
    void
    Evict the mapping for this key from this cache if it is present.
    default boolean
    Evict the mapping for this key from this cache if it is present, expecting the key to be immediately invisible for subsequent lookups.
    get(Object key)
    Return the value to which this cache maps the specified key.
    <T> T
    get(Object key, Class<T> type)
    Return the value to which this cache maps the specified key, generically specifying a type that return value will be cast to.
    <T> T
    get(Object key, Callable<T> valueLoader)
    Return the value to which this cache maps the specified key, obtaining that value from valueLoader if necessary.
    Return the cache name.
    Return the underlying native cache provider.
    default boolean
    Invalidate the cache through removing all mappings, expecting all entries to be immediately invisible for subsequent lookups.
    void
    put(Object key, Object value)
    Associate the specified value with the specified key in this cache.
    putIfAbsent(Object key, Object value)
    Atomically associate the specified value with the specified key in this cache if it is not set already.
    Return the value to which this cache maps the specified key, wrapped in a CompletableFuture.
    default <T> CompletableFuture<T>
    retrieve(Object key, Supplier<CompletableFuture<T>> valueLoader)
    Return the value to which this cache maps the specified key, obtaining that value from valueLoader if necessary.
  • Method Details

    • getName

      String getName()
      Return the cache name.
    • getNativeCache

      Object getNativeCache()
      Return the underlying native cache provider.
    • get

      Return the value to which this cache maps the specified key.

      Returns null if the cache contains no mapping for this key; otherwise, the cached value (which may be null itself) will be returned in a Cache.ValueWrapper.

      Parameters:
      key - the key whose associated value is to be returned
      Returns:
      the value to which this cache maps the specified key, contained within a Cache.ValueWrapper which may also hold a cached null value. A straight null being returned means that the cache contains no mapping for this key.
      See Also:
    • get

      @Nullable <T> T get(Object key, @Nullable Class<T> type)
      Return the value to which this cache maps the specified key, generically specifying a type that return value will be cast to.

      Note: This variant of get does not allow for differentiating between a cached null value and no cache entry found at all. Use the standard get(Object) variant for that purpose instead.

      Parameters:
      key - the key whose associated value is to be returned
      type - the required type of the returned value (may be null to bypass a type check; in case of a null value found in the cache, the specified type is irrelevant)
      Returns:
      the value to which this cache maps the specified key (which may be null itself), or also null if the cache contains no mapping for this key
      Throws:
      IllegalStateException - if a cache entry has been found but failed to match the specified type
      Since:
      4.0
      See Also:
    • get

      @Nullable <T> T get(Object key, Callable<T> valueLoader)
      Return the value to which this cache maps the specified key, obtaining that value from valueLoader if necessary. This method provides a simple substitute for the conventional "if cached, return; otherwise create, cache and return" pattern.

      If possible, implementations should ensure that the loading operation is synchronized so that the specified valueLoader is only called once in case of concurrent access on the same key.

      If the valueLoader throws an exception, it is wrapped in a Cache.ValueRetrievalException

      Parameters:
      key - the key whose associated value is to be returned
      Returns:
      the value to which this cache maps the specified key
      Throws:
      Cache.ValueRetrievalException - if the valueLoader throws an exception
      Since:
      4.3
      See Also:
    • retrieve

      @Nullable default CompletableFuture<?> retrieve(Object key)
      Return the value to which this cache maps the specified key, wrapped in a CompletableFuture. This operation must not block but is allowed to return a completed CompletableFuture if the corresponding value is immediately available.

      Can return null if the cache can immediately determine that it contains no mapping for this key (e.g. through an in-memory key map). Otherwise, the cached value will be returned in the CompletableFuture, with null indicating a late-determined cache miss. A nested Cache.ValueWrapper potentially indicates a nullable cached value; the cached value may also be represented as a plain element if null values are not supported. Calling code needs to be prepared to handle all those variants of the result returned by this method.

      Parameters:
      key - the key whose associated value is to be returned
      Returns:
      the value to which this cache maps the specified key, contained within a CompletableFuture which may also be empty when a cache miss has been late-determined. A straight null being returned means that the cache immediately determined that it contains no mapping for this key. A Cache.ValueWrapper contained within the CompletableFuture indicates a cached value that is potentially null; this is sensible in a late-determined scenario where a regular CompletableFuture-contained null indicates a cache miss. However, a cache may also return a plain value if it does not support the actual caching of null values, avoiding the extra level of value wrapping. Spring's cache processing can deal with all such implementation strategies.
      Since:
      6.1
      See Also:
    • retrieve

      default <T> CompletableFuture<T> retrieve(Object key, Supplier<CompletableFuture<T>> valueLoader)
      Return the value to which this cache maps the specified key, obtaining that value from valueLoader if necessary. This method provides a simple substitute for the conventional "if cached, return; otherwise create, cache and return" pattern, based on CompletableFuture. This operation must not block.

      If possible, implementations should ensure that the loading operation is synchronized so that the specified valueLoader is only called once in case of concurrent access on the same key.

      Null values always indicate a user-level null value with this method. The provided CompletableFuture handle produces a value or raises an exception. If the valueLoader raises an exception, it will be propagated to the returned CompletableFuture handle.

      Parameters:
      key - the key whose associated value is to be returned
      Returns:
      the value to which this cache maps the specified key, contained within a CompletableFuture which will never be null. The provided future is expected to produce a value or raise an exception.
      Since:
      6.1
      See Also:
    • put

      void put(Object key, @Nullable Object value)
      Associate the specified value with the specified key in this cache.

      If the cache previously contained a mapping for this key, the old value is replaced by the specified value.

      Actual registration may be performed in an asynchronous or deferred fashion, with subsequent lookups possibly not seeing the entry yet. This may for example be the case with transactional cache decorators. Use putIfAbsent(java.lang.Object, java.lang.Object) for guaranteed immediate registration.

      If the cache is supposed to be compatible with CompletableFuture and reactive interactions, the put operation needs to be effectively non-blocking, with any backend write-through happening asynchronously. This goes along with a cache implemented and configured to support retrieve(Object) and retrieve(Object, Supplier).

      Parameters:
      key - the key with which the specified value is to be associated
      value - the value to be associated with the specified key
      See Also:
    • putIfAbsent

      @Nullable default Cache.ValueWrapper putIfAbsent(Object key, @Nullable Object value)
      Atomically associate the specified value with the specified key in this cache if it is not set already.

      This is equivalent to:

      
       ValueWrapper existingValue = cache.get(key);
       if (existingValue == null) {
           cache.put(key, value);
       }
       return existingValue;
       
      except that the action is performed atomically. While all out-of-the-box CacheManager implementations are able to perform the put atomically, the operation may also be implemented in two steps, e.g. with a check for presence and a subsequent put, in a non-atomic way. Check the documentation of the native cache implementation that you are using for more details.

      The default implementation delegates to get(Object) and put(Object, Object) along the lines of the code snippet above.

      Parameters:
      key - the key with which the specified value is to be associated
      value - the value to be associated with the specified key
      Returns:
      the value to which this cache maps the specified key (which may be null itself), or also null if the cache did not contain any mapping for that key prior to this call. Returning null is therefore an indicator that the given value has been associated with the key.
      Since:
      4.1
      See Also:
    • evict

      void evict(Object key)
      Evict the mapping for this key from this cache if it is present.

      Actual eviction may be performed in an asynchronous or deferred fashion, with subsequent lookups possibly still seeing the entry. This may for example be the case with transactional cache decorators. Use evictIfPresent(java.lang.Object) for guaranteed immediate removal.

      If the cache is supposed to be compatible with CompletableFuture and reactive interactions, the evict operation needs to be effectively non-blocking, with any backend write-through happening asynchronously. This goes along with a cache implemented and configured to support retrieve(Object) and retrieve(Object, Supplier).

      Parameters:
      key - the key whose mapping is to be removed from the cache
      See Also:
    • evictIfPresent

      default boolean evictIfPresent(Object key)
      Evict the mapping for this key from this cache if it is present, expecting the key to be immediately invisible for subsequent lookups.

      The default implementation delegates to evict(Object), returning false for not-determined prior presence of the key. Cache providers and in particular cache decorators are encouraged to perform immediate eviction if possible (e.g. in case of generally deferred cache operations within a transaction) and to reliably determine prior presence of the given key.

      Parameters:
      key - the key whose mapping is to be removed from the cache
      Returns:
      true if the cache was known to have a mapping for this key before, false if it did not (or if prior presence could not be determined)
      Since:
      5.2
      See Also:
    • clear

      void clear()
      Clear the cache through removing all mappings.

      Actual clearing may be performed in an asynchronous or deferred fashion, with subsequent lookups possibly still seeing the entries. This may for example be the case with transactional cache decorators. Use invalidate() for guaranteed immediate removal of entries.

      If the cache is supposed to be compatible with CompletableFuture and reactive interactions, the clear operation needs to be effectively non-blocking, with any backend write-through happening asynchronously. This goes along with a cache implemented and configured to support retrieve(Object) and retrieve(Object, Supplier).

      See Also:
    • invalidate

      default boolean invalidate()
      Invalidate the cache through removing all mappings, expecting all entries to be immediately invisible for subsequent lookups.
      Returns:
      true if the cache was known to have mappings before, false if it did not (or if prior presence of entries could not be determined)
      Since:
      5.2
      See Also: