The ibatis: component allows you to query, poll, insert, update and delete data in a relational database using Apache iBATIS.
Maven users will need to add the following dependency to their pom.xml
for this component:
<dependency> <groupId>org.apache.camel</groupId> <artifactId>camel-ibatis</artifactId> <version>x.x.x</version> <!-- use the same version as your Camel core version --> </dependency>
ibatis:statementName[?options]
Where statementName is the name in the iBATIS XML configuration file which maps to the query, insert, update or delete operation you wish to evaluate.
You can append query options to the URI in the following format,
?option=value&option=value&...
This component will by default load the iBatis SqlMapConfig file from the root of the classpath and expected named as SqlMapConfig.xml
.
It uses Spring resource loading so you can define it using classpath
, file
or http
as prefix to load resources with those schemes.
In Camel 2.2 you can configure this on the iBatisComponent with the setSqlMapConfig(String)
method.
Option | Type | Default | Description |
---|---|---|---|
consumer.onConsume
|
String
|
null
|
Statements to run after consuming. Can be used, for example, to update rows after they have been consumed and processed in Fuse Mediation Router. See sample later. Multiple statements can be separated with comma. |
consumer.useIterator
|
boolean
|
true
|
If true each row returned when polling will be processed
individually. If false the entire List of data is
set as the IN body. |
consumer.routeEmptyResultSet
|
boolean
|
false
|
Fuse Mediation Router 2.0: Sets whether empty result set should be routed or not. By default, empty result sets are not routed. |
statementType
|
StatementType
|
null
|
Fuse Mediation Router 1.6.1/2.0: Mandatory to specify for
IbatisProducer to control which iBatis SqlMapClient method to invoke.
The enum values are: QueryForObject , QueryForList ,
Insert , Update , Delete . |
maxMessagesPerPoll
|
int
|
0
|
Fuse Mediation Router 2.0: An integer to define a maximum messages to gather per poll. By default, no maximum is set. Can be used to set a limit of e.g. 1000 to avoid when starting up the server that there are thousands of files. Set a value of 0 or negative to disabled it. |
Fuse Mediation Router will populate the result message, either IN or OUT with a header with the operationName used:
Header | Type | Description |
---|---|---|
org.apache.camel.ibatis.queryName
|
String
|
Fuse Mediation Router 1.x: The statementName used (for example: insertAccount). |
CamelIBatisStatementName
|
String
|
Fuse Mediation Router 2.0: The statementName used (for example: insertAccount). |
CamelIBatisResult
|
Object
|
Fuse Mediation Router 1.6.2/2.0: The response returned from iBatis
in any of the operations. For instance an INSERT could return the
auto-generated key, or number of rows etc. |
Fuse Mediation Router 1.6.1: The response from iBatis will be set as OUT body.
Fuse Mediation Router 1.6.2/2.0: The response from
iBatis will only be set as body if it's a SELECT
statement. That means, for
example, for INSERT
statements Fuse Mediation Router will not replace the body. This
allows you to continue routing and keep the original body. The response from iBatis is always
stored in the header with the key CamelIBatisResult
.
For example if you wish to consume beans from a JMS queue and insert them into a database you could do the following:
from("activemq:queue:newAccount"). to("ibatis:insertAccount?statementType=Insert");
Notice we have to specify the statementType
, as we need to instruct
Fuse Mediation Router which SqlMapClient
operation to invoke.
Where insertAccount is the iBatis ID in the SQL map file:
<!-- Insert example, using the Account parameter class --> <insert id="insertAccount" parameterClass="Account"> insert into ACCOUNT ( ACC_ID, ACC_FIRST_NAME, ACC_LAST_NAME, ACC_EMAIL ) values ( #id#, #firstName#, #lastName#, #emailAddress# ) </insert>
Available as of Fuse Mediation Router 1.6.1/2.0 When routing to an
iBatis endpoint you want more fine grained control so you can control whether the SQL
statement to be executed is a SELEECT
, UPDATE
,
DELETE
or INSERT
etc. This is now possible in Fuse Mediation Router
1.6.1/2.0. So for instance if we want to route to an iBatis endpoint in which the IN body
contains parameters to a SELECT
statement we can do:
from("direct:start") .to("ibatis:selectAccountById?statementType=QueryForObject") .to("mock:result");
In the code above we can invoke the iBatis statement selectAccountById
and the IN body should contain the account id we want to retrieve, such as an
Integer
type.
We can do the same for some of the other operations, such as
QueryForList
:
from("direct:start") .to("ibatis:selectAllAccounts?statementType=QueryForList") .to("mock:result");
And the same for UPDATE
, where we can send an
Account
object as IN body to iBatis:
from("direct:start") .to("ibatis:updateAccount?statementType=Update") .to("mock:result");
Since this component does not support scheduled polling, you need to use another mechanism for triggering the scheduled polls, such as the Timer or Quartz components.
In the sample below we poll the database, every 30 seconds using the Timer component and send the data to the JMS queue:
from("timer://pollTheDatabase?delay=30000").to("ibatis:selectAllAccounts?statementType=QueryForList").to("activemq:queue:allAccounts");
And the iBatis SQL map file used:
<!-- Select with no parameters using the result map for Account class. --> <select id="selectAllAccounts" resultMap="AccountResult"> select * from ACCOUNT </select>
This component supports executing statements after data
have been consumed and processed by Fuse Mediation Router. This allows you to do post updates in the
database. Notice all statements must be UPDATE
statements. Fuse Mediation Router
supports executing multiple statements whose name should be separated by comma.
The route below illustrates we execute the consumeAccount statement data is processed. This allows us to change the status of the row in the database to processed, so we avoid consuming it twice or more.
from("ibatis:selectUnprocessedAccounts?consumer.onConsume=consumeAccount").to("mock:results");
And the statements in the sqlmap file:
<select id="selectUnprocessedAccounts" resultMap="AccountResult"> select * from ACCOUNT where PROCESSED = false </select>
<update id="consumeAccount" parameterClass="Account"> update ACCOUNT set PROCESSED = true where ACC_ID = #id# </update>