io.appform.dropwizard.sharding.dao

Class LookupDao<T>

java.lang.Object

io.appform.dropwizard.sharding.dao.LookupDao<T>

All Implemented Interfaces:

ShardedDao<T>

Direct Known Subclasses:

CacheableLookupDao

public class LookupDao<T>
extends Object
implements ShardedDao<T>

A dao to manage lookup and top level elements in the system. Can save and retrieve an object (tree) from any shard. Note: - The element must have only one String key for lookup. - The key needs to be annotated with LookupKey The entity can be retrieved from any shard using the key.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	LookupDao.ReadOnlyContext<T>

Constructor Summary

Constructors

Constructor and Description
LookupDao(List<org.hibernate.SessionFactory> sessionFactories, Class<T> entityClass, ShardCalculator<String> shardCalculator, ShardingBundleOptions shardingOptions, ShardInfoProvider shardInfoProvider, TransactionObserver observer) Creates a new sharded DAO.

Method Summary

Modifier and Type	Method and Description
List<Long>	count(org.hibernate.criterion.DetachedCriteria criteria) Queries using the specified criteria across all shards and returns the counts of rows satisfying the criteria.
Optional<T>	createOrUpdate(String id, UnaryOperator<T> updater, Supplier<T> entityGenerator)
boolean	delete(String id)
boolean	exists(String key) Check if object with specified key exists in any shard.
List<T>	get(List<String> keys) Queries across various shards and returns the results.
Optional<T>	get(String key) Get an object on the basis of key (value of field annotated with LookupKey) from any shard.
<U> U	get(String key, Function<T,U> handler) Get an object on the basis of key (value of field annotated with LookupKey) from any shard and applies the provided function/lambda to it.
Optional<T>	get(String key, UnaryOperator<org.hibernate.Criteria> criteriaUpdater)
<U> U	get(String key, UnaryOperator<org.hibernate.Criteria> criteriaUpdater, Function<T,U> handler)
protected Field	getKeyField()
LockedContext<T>	lockAndGetExecutor(String id)
LookupDao.ReadOnlyContext<T>	readOnlyExecutor(String id)
LookupDao.ReadOnlyContext<T>	readOnlyExecutor(String id, Supplier<Boolean> entityPopulator)
LookupDao.ReadOnlyContext<T>	readOnlyExecutor(String id, UnaryOperator<org.hibernate.Criteria> criteriaUpdater)
LookupDao.ReadOnlyContext<T>	readOnlyExecutor(String id, UnaryOperator<org.hibernate.Criteria> criteriaUpdater, Supplier<Boolean> entityPopulator)
Map<Integer,List>	run(org.hibernate.criterion.DetachedCriteria criteria) Run arbitrary read-only queries on all shards and return results.
<U> U	run(org.hibernate.criterion.DetachedCriteria criteria, Function<Map<Integer,List>,U> translator) Run read-only queries on all shards and transform them into required types
<U,V> V	runInSession(BiFunction<Integer,org.hibernate.Session,U> sessionHandler, Function<Map<Integer,U>,V> translator)
<U> U	runInSession(String id, Function<org.hibernate.Session,U> handler)
Optional<T>	save(T entity) Saves an entity on proper shard based on hash of the value in the key field in the object.
<U> U	save(T entity, Function<T,U> handler) Save an object on the basis of key (value of field annotated with LookupKey) to target shard and applies the provided function/lambda to it.
LockedContext<T>	saveAndGetExecutor(T entity)
List<T>	scatterGather(org.hibernate.criterion.DetachedCriteria criteria) Queries using the specified criteria across all shards and returns the result.
ScrollResult<T>	scrollDown(org.hibernate.criterion.DetachedCriteria inCriteria, ScrollPointer inPointer, int pageSize, @NonNull String sortFieldName) Provides a scroll api for records across shards.
ScrollResult<T>	scrollUp(org.hibernate.criterion.DetachedCriteria inCriteria, ScrollPointer inPointer, int pageSize, @NonNull String sortFieldName) Provides a scroll api for records across shards.
boolean	update(String id, Function<Optional<T>,T> updater)
boolean	updateInLock(String id, Function<Optional<T>,T> updater)
int	updateUsingQuery(String id, UpdateOperationMeta updateOperationMeta)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface io.appform.dropwizard.sharding.dao.ShardedDao

getShardCalculator

Constructor Detail

LookupDao

public LookupDao(List<org.hibernate.SessionFactory> sessionFactories,
                 Class<T> entityClass,
                 ShardCalculator<String> shardCalculator,
                 ShardingBundleOptions shardingOptions,
                 ShardInfoProvider shardInfoProvider,
                 TransactionObserver observer)

Creates a new sharded DAO. The number of managed shards and bucketing is controlled by the ShardManager.

Parameters:

sessionFactories - a session provider for each shard

shardCalculator - calculator for shards

Method Detail

get

public Optional<T> get(String key)
                throws Exception

Get an object on the basis of key (value of field annotated with LookupKey) from any shard. Note: Lazy loading will not work once the object is returned. If you need lazy loading functionality use the alternate get(String, Function) method.

Parameters:

key - The value of the key field to look for.

Returns:

The entity

Throws:

Exception - if backing dao throws

get

public Optional<T> get(String key,
                       UnaryOperator<org.hibernate.Criteria> criteriaUpdater)
                throws Exception

Throws:

Exception

get

public <U> U get(String key,
                 Function<T,U> handler)
          throws Exception

Get an object on the basis of key (value of field annotated with LookupKey) from any shard and applies the provided function/lambda to it. The return from the handler becomes the return to the get function. Note: The transaction is open when handler is applied. So lazy loading will work inside the handler. Once get returns, lazy loading will nt owrok.

Parameters:

key - The value of the key field to look for.

handler - Handler function/lambda that receives the retrieved object.

Returns:

Whatever is returned by the handler function

Throws:

Exception - if backing dao throws

get

public <U> U get(String key,
                 UnaryOperator<org.hibernate.Criteria> criteriaUpdater,
                 Function<T,U> handler)
          throws Exception

Throws:

Exception

exists

public boolean exists(String key)
               throws Exception

Check if object with specified key exists in any shard.

Parameters:

key - id of the element to look for

Returns:

true/false depending on if it's found or not.

Throws:

Exception - if backing dao throws

save

public Optional<T> save(T entity)
                 throws Exception

Saves an entity on proper shard based on hash of the value in the key field in the object. The updated entity is returned. If Cascade is specified, this can be used to save an object tree based on the shard of the top entity that has the key field. Note: Lazy loading will not work on the augmented entity. Use the alternate save(Object, Function) for that.

Parameters:

entity - Entity to save

Returns:

Entity

Throws:

Exception - if backing dao throws

save

public <U> U save(T entity,
                  Function<T,U> handler)
           throws Exception

Save an object on the basis of key (value of field annotated with LookupKey) to target shard and applies the provided function/lambda to it. The return from the handler becomes the return to the get function. Note: Handler is executed in the same transactional context as the save operation. So any updates made to the object in this context will also get persisted.

Parameters:

entity - The value of the key field to look for.

handler - Handler function/lambda that receives the retrieved object.

Returns:

The entity

Throws:

Exception - if backing dao throws

createOrUpdate

public Optional<T> createOrUpdate(String id,
                                  UnaryOperator<T> updater,
                                  Supplier<T> entityGenerator)

updateInLock

public boolean updateInLock(String id,
                            Function<Optional<T>,T> updater)

update

public boolean update(String id,
                      Function<Optional<T>,T> updater)

updateUsingQuery

public int updateUsingQuery(String id,
                            UpdateOperationMeta updateOperationMeta)

lockAndGetExecutor

public LockedContext<T> lockAndGetExecutor(String id)

readOnlyExecutor

public LookupDao.ReadOnlyContext<T> readOnlyExecutor(String id)

readOnlyExecutor

public LookupDao.ReadOnlyContext<T> readOnlyExecutor(String id,
                                                     UnaryOperator<org.hibernate.Criteria> criteriaUpdater)

readOnlyExecutor

public LookupDao.ReadOnlyContext<T> readOnlyExecutor(String id,
                                                     Supplier<Boolean> entityPopulator)

readOnlyExecutor

public LookupDao.ReadOnlyContext<T> readOnlyExecutor(String id,
                                                     UnaryOperator<org.hibernate.Criteria> criteriaUpdater,
                                                     Supplier<Boolean> entityPopulator)

saveAndGetExecutor

public LockedContext<T> saveAndGetExecutor(T entity)

scatterGather

public List<T> scatterGather(org.hibernate.criterion.DetachedCriteria criteria)

Queries using the specified criteria across all shards and returns the result. Note: This method runs the query serially, and it's usage is not recommended.

Parameters:

criteria - The select criteria

Returns:

List of elements or empty if none match

scrollDown

public ScrollResult<T> scrollDown(org.hibernate.criterion.DetachedCriteria inCriteria,
                                  ScrollPointer inPointer,
                                  int pageSize,
                                  @NonNull
                                  @NonNull String sortFieldName)

Provides a scroll api for records across shards. This api will scroll down in ascending order of the 'sortFieldName' field. Newly added records can be polled by passing the pointer repeatedly. If nothing new is available, it will return an empty in ScrollResult#getResult(). If the passed pointer is null, it will return the first pageSize records with a pointer to be passed to get the next pageSize set of records.

NOTES: - Do not modify the criteria between subsequent calls - It is important to provide a sort field that is perpetually increasing - Pointer returned can be used to _only_ scroll down

Parameters:

inCriteria - The core criteria for the query

inPointer - Existing ScrollPointer, should be null at start of a scroll session

pageSize - Count of records per shard

sortFieldName - Field to sort by. For correct sorting, the field needs to be an ever-increasing one

Returns:

A ScrollResult object that contains a ScrollPointer and a list of results with max N * pageSize elements

scrollUp

public ScrollResult<T> scrollUp(org.hibernate.criterion.DetachedCriteria inCriteria,
                                ScrollPointer inPointer,
                                int pageSize,
                                @NonNull
                                @NonNull String sortFieldName)

Provides a scroll api for records across shards. This api will scroll up in descending order of the 'sortFieldName' field. As this api goes back in order, newly added records will not be available in the scroll. If the passed pointer is null, it will return the last pageSize records with a pointer to be passed to get the previous pageSize set of records.

NOTES: - Do not modify the criteria between subsequent calls - It is important to provide a sort field that is perpetually increasing - Pointer returned can be used to _only_ scroll up

Parameters:

inCriteria - The core criteria for the query

inPointer - Existing ScrollPointer, should be null at start of a scroll session

pageSize - Count of records per shard

sortFieldName - Field to sort by. For correct sorting, the field needs to be an ever-increasing one

Returns:

A ScrollResult object that contains a ScrollPointer and a list of results with max N * pageSize elements

count

public List<Long> count(org.hibernate.criterion.DetachedCriteria criteria)

Queries using the specified criteria across all shards and returns the counts of rows satisfying the criteria. Note: This method runs the query serially

Parameters:

criteria - The select criteria

Returns:

List of counts in each shard

run

public Map<Integer,List> run(org.hibernate.criterion.DetachedCriteria criteria)

Run arbitrary read-only queries on all shards and return results.

Parameters:

criteria - The detached criteria. Typically, a grouping or counting query

Returns:

A map of shard vs result-list

run

public <U> U run(org.hibernate.criterion.DetachedCriteria criteria,
                 Function<Map<Integer,List>,U> translator)

Run read-only queries on all shards and transform them into required types

Type Parameters:

U - Return type

Parameters:

criteria - The detached criteria. Typically, a grouping or counting query

translator - A method to transform results to required type

Returns:

Translated result

get

public List<T> get(List<String> keys)

Queries across various shards and returns the results. Note: This method runs the query serially and is efficient over scatterGather and serial get of all key

Parameters:

keys - The list of lookup keys

Returns:

List of elements or empty if none match

runInSession

public <U> U runInSession(String id,
                          Function<org.hibernate.Session,U> handler)

runInSession

public <U,V> V runInSession(BiFunction<Integer,org.hibernate.Session,U> sessionHandler,
                            Function<Map<Integer,U>,V> translator)

delete

public boolean delete(String id)

getKeyField

protected Field getKeyField()