com.neurosys.dao.jpa

Interface QueryOptimizer

All Superinterfaces:

org.apache.commons.collections.Transformer

All Known Implementing Classes:

QueryResultArrayToEntityTransformer

public interface QueryOptimizer
extends org.apache.commons.collections.Transformer

Many times its convenient to express Queries in a general way that are left unoptimized. A QueryOptimizer implementation can wrap the queries and perform various optimization functions on them. To the calling API the transformations are transparent.

More literally, it allows to take a Query string an optimize the Query String. And on the other hand takes the result and allows to normalize the result based on the optimized Query String.
Normalize - When a Query string is manipulated, the result will not match the intended original result. Hence we need a transformation process to normalize it; so the calling API does not feel the difference. Example:

 
	<bean id="qlOptimizer_OrderSummary" class="com.neurosys.modules.etl.QueryResultArrayToEntityTransformer">
		<constructor-arg value="com.neurosys.ecom.orders.domain.Order" /> <!-- type -->
		<constructor-arg><list>
			<value>id</value>
			<value>purchaseDate</value>
			<value>status</value>
			<value>userContactDetails</value>
		</list></constructor-arg>		  	<!-- field names -->
		<!-- <constructor-arg><bean class="com.neurosys.etl.common.transformers.MapToCollectionTransformer" /></constructor-arg>
		<constructor-arg ref="orderRecordToColumnIndexMapper" /> --> <!-- Only fetch fields needed, MAP of column index from 0... mapped-to field-name or Transformer(entity). Interchangable -->				
		<constructor-arg value="false" />   <!-- Not Inverse : Field names are inclusive not exclusive -->
		<property name="overrideFromClause"><value><![CDATA[ Order orders LEFT JOIN orders.transaction ]]></value></property>
		<property name="nestedObjectInitMap"> 
			-- Map of field name and types key="amount" value="#{T(com.neurosys.ecom.orders.domain.Amount)}" --
                        
		</property>
</bean>	 
<bean id="allOrdersQuery" class="com.neurosys.dao.jpa.JPAQueryTransformer">
		<property name="query"><value><![CDATA[
SELECT orders FROM Order orders order by orders.id desc
		]]></value></property>
		<property name="queryIsNative" value="false" />
		<property name="queryOptimizer" ref="qlOptimizer_OrderSummary" />
		<property name="queryOptimizerEntityAlias" value="orders" />
</bean>
 

Author:

Arjun Dhar

Method Summary

Modifier and Type	Method and Description
void	morphQuery(Query query) Allow compatibility with Query object.
String	replaceEntityString(String query, String entityAlias) Manipulates the input Query String
Object	transform(Object result) Normalize the result.

Method Detail

replaceEntityString

String replaceEntityString(String query,
                           String entityAlias)

Manipulates the input Query String

Parameters:

query - as raw Query String

entityAlias - What represents the entity or Tables alias(s) in the query

Returns:

Optimized query String to execute

morphQuery

void morphQuery(Query query)

Allow compatibility with Query object. Many API calls may use Query instead of a direct Query String, in this case this allows us to morph the Query object to its optimized version.

Parameters:

query - as Query

transform

Object transform(Object result)

Normalize the result. See interface definition for what Normalization implies.

Specified by:

transform in interface org.apache.commons.collections.Transformer