Spring Hadoop Schema
<?xml version="1.0" encoding="UTF-8"?> <xsd:schema xmlns="http://www.springframework.org/schema/hadoop" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:beans="http://www.springframework.org/schema/beans" xmlns:tool="http://www.springframework.org/schema/tool" targetNamespace="http://www.springframework.org/schema/hadoop" elementFormDefault="qualified" attributeFormDefault="unqualified" version="1.0.0.M1"> <xsd:import namespace="http://www.springframework.org/schema/beans" /> <xsd:import namespace="http://www.springframework.org/schema/tool" /> <xsd:annotation> <xsd:documentation><![CDATA[ Defines the configuration elements for Spring Data Hadoop. ]]></xsd:documentation> </xsd:annotation> <xsd:element name="tasklet"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a Spring Batch tasklet for Hadoop Jobs. ]]> </xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.springframework.data.hadoop.mapreduce.HadoopTasklet"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:complexType> <!-- the job reference --> <xsd:attribute name="job-ref"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.mapreduce.Job"><![CDATA[ Hadoop Job]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.mapreduce.Job" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="wait-for-job" type="xsd:string" use="optional" default="true"> <xsd:annotation> <xsd:documentation><![CDATA[ Whether to synchronously wait for the job to finish (the default) or not. ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="id" type="xsd:ID" use="required" /> <xsd:attribute name="scope" type="xsd:string" use="optional" /> </xsd:complexType> </xsd:element> <!-- common attributes shared by properties based configurations NOT meant for extensibility - do NOT rely on this type as it might be removed in the future --> <xsd:complexType name="propertiesConfigurableType" mixed="true"> <xsd:attribute name="properties-ref" type="xsd:string"> <xsd:annotation> <xsd:documentation><![CDATA[ Reference to a Properties object. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="java.util.Properties" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="properties-location" type="xsd:string"> <xsd:annotation> <xsd:documentation><![CDATA[ Properties location(s). Multiple locations can be specified using comma (,) as a separator. ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> </xsd:complexType> <xsd:element name="configuration"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a Hadoop Configuration. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.apache.hadoop.conf.Configuration"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:complexType mixed="true"> <xsd:complexContent> <xsd:extension base="propertiesConfigurableType"> <xsd:attribute name="id" type="xsd:ID" use="optional"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id (default is "hadoop-configuration"). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="configuration-ref"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.conf.Configuration"><![CDATA[ Reference to another Hadoop configuration (useful for chaining)]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.conf.Configuration" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="resources"> <xsd:annotation> <xsd:documentation source="java:org.springframework.core.io.Resource"><![CDATA[ Hadoop Configuration resources. Multiple resources can be specified, using comma (,) as a separator.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="org.springframework.core.io.Resource[]" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="register-url-handler" use="optional" default="false"> <xsd:annotation> <xsd:documentation><![CDATA[ Registers an HDFS url handler in the running VM. Note that this operation can be executed at most once in a given JVM hence the default is false. ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> </xsd:element> <xsd:element name="resource-loader"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a HDFS-aware resource loader. ]]> </xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.springframework.data.hadoop.fs.HdfsResourceLoader"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:complexType> <xsd:attribute name="id" type="xsd:ID" use="optional"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id (default is "hadoop-resource-loader"). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="configuration-ref" use="optional" default="hadoop-configuration"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.conf.Configuration"><![CDATA[ Reference to the Hadoop Configuration. Defaults to 'hadoop-configuration'.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.conf.Configuration" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="uri" use="optional"/> </xsd:complexType> </xsd:element> <!-- common attributes shared by properties based configurations NOT meant for extensibility - do NOT rely on this type as it might be removed in the future --> <xsd:complexType name="jobType"> <xsd:complexContent> <xsd:extension base="propertiesConfigurableType"> <xsd:attribute name="id" type="xsd:ID" use="required" /> <xsd:attribute name="scope" type="xsd:string" use="optional" /> <xsd:attribute name="mapper" use="required"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> <tool:assignable-to type="org.apache.hadoop.mapreduce.Mapper" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="reducer" use="required"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> <tool:assignable-to type="org.apache.hadoop.mapreduce.Reducer" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="combiner"> <xsd:annotation> <xsd:documentation><![CDATA[ The combiner class name. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> <tool:assignable-to type="org.apache.hadoop.mapreduce.Reducer" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="input-format"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> <tool:assignable-to type="org.apache.hadoop.mapreduce.InputFormat" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="output-format"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> <tool:assignable-to type="org.apache.hadoop.mapreduce.OutputFormat" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="partitioner"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> <tool:assignable-to type="org.apache.hadoop.mapreduce.Partitioner" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="input-path" use="required"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.String[]" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="output-path" use="required"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.String" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="configuration-ref" default="hadoop-configuration"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.conf.Configuration"><![CDATA[ Reference to the Hadoop Configuration. Defaults to 'hadoop-configuration'.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.conf.Configuration" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> <xsd:element name="job"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a Hadoop Job. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.apache.hadoop.mapreduce.Job"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:complexType> <xsd:complexContent mixed="true"> <xsd:extension base="jobType"> <xsd:attribute name="sort-comparator"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> <tool:assignable-to type="org.apache.hadoop.io.RawComparator" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="grouping-comparator"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> <tool:assignable-to type="org.apache.hadoop.io.RawComparator" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="key"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="value"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="map-key"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="map-value"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="codec"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="jar"> <xsd:annotation> <xsd:documentation><![CDATA[ Indicates the user jar for the map-reduce job. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:expected-type type="org.springframework.core.io.Resource" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="jar-by-class"> <xsd:annotation> <xsd:documentation><![CDATA[ Indicates the job's jar file by finding an example class location. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="validate-paths" default="true"> <xsd:annotation> <xsd:documentation><![CDATA[ Indicates whether the job input/output paths are validated before submitting. This saves time as the validation is done locally without having to interact with the job tracker. The validation checks whether the input path exists and the output does not. ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> </xsd:element> <xsd:element name="streaming"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a Hadoop Streaming Job. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.apache.hadoop.mapreduce.Job"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:complexType> <xsd:complexContent mixed="true"> <xsd:extension base="jobType"> <xsd:sequence> <xsd:element name="cmd-env" minOccurs="0" maxOccurs="1"> <xsd:annotation> <xsd:documentation><![CDATA[Environment variables (-cmdenv)]]></xsd:documentation> </xsd:annotation> </xsd:element> </xsd:sequence> <xsd:attribute name="number-reducers"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Integer" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="file"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.String[]" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="archive"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.String[]" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="lib"> <xsd:annotation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.String[]" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> </xsd:element> <!-- common attributes shared by properties based configurations NOT meant for extensibility - do NOT rely on this type as it might be removed in the future --> <xsd:complexType name="toolRunnerType"> <xsd:complexContent mixed="true"> <xsd:extension base="propertiesConfigurableType"> <xsd:sequence> <xsd:any namespace="##any" processContents="lax" minOccurs="0" maxOccurs="1"/> <xsd:element name="arg" minOccurs="0" maxOccurs="unbounded"> <xsd:annotation> <xsd:documentation><![CDATA[ Tool argument.]]></xsd:documentation> </xsd:annotation> <xsd:complexType> <xsd:attribute name="value" type="xsd:string" use="required"/> </xsd:complexType> </xsd:element> </xsd:sequence> <xsd:attribute name="id" type="xsd:ID" use="required"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id.]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="tool-class" type="xsd:string"> <xsd:annotation> <xsd:documentation><![CDATA[ Indicates the job's jar file by finding an example class location. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="direct"> <tool:expected-type type="java.lang.Class" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="tool-ref"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.util.Tool"><![CDATA[ Reference to a Hadoop Tool instance.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.util.Tool" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="configuration-ref" default="hadoop-configuration"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.conf.Configuration"><![CDATA[ Reference to the Hadoop Configuration. Defaults to 'hadoop-configuration'.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.conf.Configuration" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> <xsd:element name="tool-runner"> <xsd:annotation> <xsd:documentation><![CDATA[ Executes a Hadoop Tool.]]></xsd:documentation> </xsd:annotation> <xsd:complexType> <xsd:complexContent mixed="true"> <xsd:extension base="toolRunnerType"> <xsd:attribute name="run-at-startup" type="xsd:boolean" default="false"> <xsd:annotation> <xsd:documentation><![CDATA[ Whether the Tool runs at startup or not (default). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> </xsd:element> <xsd:element name="tool-tasklet"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a Hadoop Tool Tasklet.]]></xsd:documentation> </xsd:annotation> <xsd:complexType> <xsd:complexContent> <xsd:extension base="toolRunnerType"> <xsd:attribute name="scope" type="xsd:string" use="optional" /> </xsd:extension> </xsd:complexContent> </xsd:complexType> </xsd:element> <xsd:complexType name="entryType"> <xsd:attribute name="value" type="xsd:string" use="required"/> </xsd:complexType> <xsd:element name="cache"> <xsd:annotation> <xsd:documentation><![CDATA[ Configures Hadoop Distributed Cache. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.apache.hadoop.io.DistributedCacheFactoryBean"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:complexType> <xsd:sequence minOccurs="1" maxOccurs="unbounded"> <xsd:choice> <xsd:element name="classpath" type="entryType"/> <xsd:element name="cache" type="entryType"/> <xsd:element name="local" type="entryType"/> </xsd:choice> </xsd:sequence> <xsd:attribute name="id" type="xsd:ID" use="optional"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id (default is "hadoop-cache"). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="create-symlink" type="xsd:boolean"/> <xsd:attribute name="configuration-ref" default="hadoop-configuration"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.conf.Configuration"><![CDATA[ Reference to the Hadoop Configuration. Defaults to 'hadoop-configuration'.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.conf.Configuration" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="file-system-ref"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.fs.FileSystem"><![CDATA[ Reference to the Hadoop FileSystem.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.fs.FileSystem" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> </xsd:complexType> </xsd:element> <xsd:complexType name="scriptType" mixed="true"> <xsd:attribute name="location" type="xsd:string"> <xsd:annotation> <xsd:documentation><![CDATA[ Location of the script. As an alternative one can inline the script by using a nested, text declaration.]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:expected-type type="org.springframework.core.io.Resource"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> </xsd:complexType> <xsd:complexType name="scriptWithArgumentsType" mixed="true"> <xsd:complexContent> <xsd:extension base="scriptType"> <xsd:sequence> <xsd:element name="arguments" minOccurs="0" maxOccurs="1"> <xsd:annotation> <xsd:documentation><![CDATA[ Argument(s) to pass to this script. Defined in Properties format (key=value). ]]></xsd:documentation> </xsd:annotation> </xsd:element> </xsd:sequence> </xsd:extension> </xsd:complexContent> </xsd:complexType> <xsd:element name="pig"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a PigServer 'template' (note that since PigServer is not thread-safe, each bean invocation will create a new PigServer instance). ]]> </xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.apache.pig.PigServer"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:complexType> <xsd:complexContent mixed="true"> <xsd:extension base="propertiesConfigurableType"> <xsd:sequence> <xsd:element name="script" type="scriptWithArgumentsType" minOccurs="0" maxOccurs="unbounded"> <xsd:annotation> <xsd:documentation><![CDATA[ Pig script.]]></xsd:documentation> </xsd:annotation> </xsd:element> </xsd:sequence> <xsd:attribute name="id" type="xsd:ID" use="optional"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id (default is "hadoop-pig-server"). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="paths-to-skip"> <xsd:annotation> <xsd:documentation><![CDATA[ The path to be skipped while automatically shipping binaries for streaming. Multiple resources can be specified, using comma (,) as a separator. ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="parallelism" type="xsd:integer"/> <xsd:attribute name="validate-each-statement" type="xsd:boolean"/> <xsd:attribute name="job-priority" type="xsd:string"/> <xsd:attribute name="job-name" type="xsd:string"/> <xsd:attribute name="job-tracker" type="xsd:string"/> <xsd:attribute name="configuration-ref"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.conf.Configuration"><![CDATA[ Reference to the Hadoop Configuration. Can be tweaked through the 'configuration' element or the other attributes.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.conf.Configuration" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="exec-type" default="MAPREDUCE"> <xsd:simpleType> <xsd:restriction base="xsd:string"> <xsd:enumeration value="MAPREDUCE"/> <xsd:enumeration value="LOCAL"/> </xsd:restriction> </xsd:simpleType> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> </xsd:element> <xsd:element name="pig-tasklet"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a PigTasklet. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.springframework.data.hadoop.batch.PigTasket"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:complexType> <xsd:sequence> <xsd:element name="script" type="scriptWithArgumentsType" minOccurs="1" maxOccurs="unbounded"> <xsd:annotation> <xsd:documentation><![CDATA[ Pig script.]]></xsd:documentation> </xsd:annotation> </xsd:element> </xsd:sequence> <xsd:attribute name="id" type="xsd:ID" use="required"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id.]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="pig-server-ref" type="xsd:string" use="optional" default="pig"> <xsd:annotation> <xsd:documentation source="java:org.apache.pig.PigServer"><![CDATA[ Reference to a PigServer instance. Defaults to 'hadoop-pig-server'. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.pig.PigServer" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="scope" type="xsd:string" use="optional" /> </xsd:complexType> </xsd:element> <!-- HBase --> <xsd:element name="hbase-configuration"> <xsd:complexType> <xsd:complexContent mixed="true"> <xsd:extension base="propertiesConfigurableType"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines an HBase configuration. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.apache.hadoop.conf.Configuration"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:attribute name="id" type="xsd:ID" use="optional"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id (default is "hbase-configuration"). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="stop-proxy" type="xsd:boolean" default="true"/> <xsd:attribute name="delete-connection" type="xsd:boolean" default="true"/> <xsd:attribute name="configuration-ref"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.conf.Configuration"><![CDATA[ Reference to the Hadoop configuration.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.conf.Configuration" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> </xsd:element> <!-- Hive --> <xsd:element name="hive-client"> <xsd:complexType> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a Hive client for connecting to a Hive server through the Thrift protocol. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.apache.hadoop.hive.service.HiveClient"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:sequence> <xsd:element name="script" type="scriptType" minOccurs="0" maxOccurs="unbounded"> <xsd:annotation> <xsd:documentation><![CDATA[ Hive script to be executed during start-up.]]></xsd:documentation> </xsd:annotation> </xsd:element> </xsd:sequence> <xsd:attribute name="id" type="xsd:ID" use="optional"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id (default is "hive-client"). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="host" type="xsd:string" default="localhost"/> <xsd:attribute name="port" type="xsd:string" default="10000"/> <xsd:attribute name="auto-startup" type="xsd:boolean" default="true"/> </xsd:complexType> </xsd:element> <xsd:element name="hive-server"> <xsd:complexType> <xsd:complexContent mixed="true"> <xsd:extension base="propertiesConfigurableType"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines an embedded Hive Server instance opened for access through the Thrift protocol. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.apache.thrift.server.TServer"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:attribute name="id" type="xsd:ID" use="optional"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id (default is "hive-server"). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="port" type="xsd:string" default="10000"/> <xsd:attribute name="min-threads" type="xsd:string" default="5"/> <xsd:attribute name="max-threads" type="xsd:string" default="100"/> <xsd:attribute name="auto-startup" type="xsd:boolean" default="true"/> <xsd:attribute name="configuration-ref"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.conf.Configuration"><![CDATA[ Reference to the Hadoop configuration.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.conf.Configuration" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> </xsd:element> <xsd:element name="hive-tasklet"> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a HiveTasklet. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="org.springframework.data.hadoop.batch.HiveTasket"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:complexType> <xsd:sequence> <xsd:element name="script" type="scriptType" minOccurs="1" maxOccurs="unbounded"> <xsd:annotation> <xsd:documentation><![CDATA[ Hive script.]]></xsd:documentation> </xsd:annotation> </xsd:element> </xsd:sequence> <xsd:attribute name="id" type="xsd:ID" use="required"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id.]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="hive-client-ref" type="xsd:string" use="optional" default="hive-client"> <xsd:annotation> <xsd:documentation source="java:org.apache.hadoop.hive.service.HiveClient"><![CDATA[ Reference to a HiveClient instance. Defaults to 'hive-client'. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="org.apache.hadoop.hive.service.HiveClient" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="scope" type="xsd:string" use="optional" /> </xsd:complexType> </xsd:element> <!-- Script type - NOT mean to be reused outside this schema --> <xsd:complexType name="scriptingType" abstract="true" mixed="true"> <xsd:sequence> <xsd:element name="property" type="beans:propertyType" minOccurs="0" maxOccurs="unbounded"> <xsd:annotation> <xsd:documentation><![CDATA[ Property to pass to the script. Can be used to enhance or override the default properties. ]]></xsd:documentation> </xsd:annotation> </xsd:element> </xsd:sequence> <xsd:attribute name="location" type="xsd:string"> <xsd:annotation> <xsd:documentation><![CDATA[ The location of the script. Can be any resource on the local filesystem, web or even hdfs. ]]> </xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="language" type="xsd:string"> <xsd:annotation> <xsd:documentation><![CDATA[ The language used for executing the script. If no value is given, the script source extension is used to determine the scripting engine. ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="evaluate" default="ALWAYS"> <xsd:annotation> <xsd:documentation><![CDATA[ When to evaluate the script. 'ALWAYS' (default) evaluates the script on all invocations, 'IF_MODIFIED' if the script source has been modified since the last invocation and 'ONCE' only once for the duration of the application. ]]></xsd:documentation> </xsd:annotation> <xsd:simpleType> <xsd:restriction base="xsd:string"> <xsd:enumeration value="ONCE"/> <xsd:enumeration value="IF_MODIFIED"/> <xsd:enumeration value="ALWAYS"/> </xsd:restriction> </xsd:simpleType> </xsd:attribute> </xsd:complexType> <xsd:element name="script"> <xsd:complexType mixed="true"> <xsd:complexContent> <xsd:annotation> <xsd:documentation><![CDATA[ Dedicated scripting facility for interacting with Hadoop. Allows Groovy, JavaScript (Rhino), Ruby (JRuby), Python (Jython) or any JSR-223 scripting language to be used for executing commands against Hadoop, in particular its file system. ]]></xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="java.lang.Object"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:extension base="scriptingType"> <xsd:attribute name="id" type="xsd:ID" use="optional"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id (if no value is given, a name will be generated). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="run-at-startup" type="xsd:boolean" default="false"> <xsd:annotation> <xsd:documentation><![CDATA[ Whether the script is evaluated automatically once the application context initializes or only when in use (the default). ]]></xsd:documentation> </xsd:annotation> </xsd:attribute> </xsd:extension> </xsd:complexContent> </xsd:complexType> </xsd:element> <xsd:element name="script-tasklet"> <xsd:complexType> <xsd:annotation> <xsd:documentation><![CDATA[ Defines a scripting Tasklet for interacting with Hadoop. Allows Groovy, JavaScript (Rhino), Ruby (JRuby), Python (Jython) or any JSR-223 scripting language to be used for executing commands against Hadoop, in particular its file system. ]]> </xsd:documentation> <xsd:appinfo> <tool:annotation> <tool:exports type="java.lang.Object"/> </tool:annotation> </xsd:appinfo> </xsd:annotation> <xsd:sequence> <xsd:element name="script" minOccurs="0" maxOccurs="1"> <xsd:annotation> <xsd:documentation><![CDATA[ Nested script declaration.]]></xsd:documentation> </xsd:annotation> <xsd:complexType mixed="true"> <xsd:complexContent> <xsd:extension base="scriptingType"/> </xsd:complexContent> </xsd:complexType> </xsd:element> </xsd:sequence> <xsd:attribute name="id" type="xsd:ID" use="required"> <xsd:annotation> <xsd:documentation><![CDATA[ Bean id.]]></xsd:documentation> </xsd:annotation> </xsd:attribute> <xsd:attribute name="script-ref" type="xsd:string"> <xsd:annotation> <xsd:documentation><![CDATA[ Reference to a script declaration.]]></xsd:documentation> <xsd:appinfo> <tool:annotation kind="ref"> <tool:expected-type type="java.lang.Object" /> </tool:annotation> </xsd:appinfo> </xsd:annotation> </xsd:attribute> <xsd:attribute name="scope" type="xsd:string" use="optional" /> </xsd:complexType> </xsd:element> </xsd:schema>