Extend
Arc can be exended in four ways by registering:
- Dynamic Configuration Plugins.
- Lifecycle Plugins.
- Pipeline Stage Plugins.
- User Defined Functions which extend the Spark SQL dialect.
Dynamic Configuration Plugins
Since: 1.3.0
Dynamic vs Deterministic Configuration
Use of this functionality is discouraged as it goes against the principles of Arc specifically around statelessness/deterministic behaviour but is inlcuded here for users who have not yet committed to a job orchestrator such as Apache Airflow and have dynamic configuration requirements.
The Dynamic Configuration Plugin
plugin allow users to inject custom configuration parameters which will be processed before resolving the job configuration file. The plugin must return a Java Map[String, Object]
which will be included in the job configuration resolution step.
Examples
For example a custom runtime configuration plugin could be used calculate a formatted list of dates to be used with an Extract stage to read only a subset of documents:
package au.com.agl.arc.plugins.config
import scala.collection.JavaConverters._
import au.com.agl.arc.plugins._
import au.com.agl.arc.util.log.logger.Logger
import java.sql.Date
import java.time.LocalDate
import java.time.format.{DateTimeFormatter, DateTimeFormatterBuilder}
import java.time.format.ResolverStyle
class DeltaPeriodDynamicConfigurationPlugin extends DynamicConfigurationPlugin {
override def values(params: Map[String, String])(implicit logger: au.com.agl.arc.util.log.logger.Logger): java.util.Map[String, Object] = {
val startTime = System.currentTimeMillis()
val stageDetail = new java.util.HashMap[String, Object]()
stageDetail.put("type", "DeltaPeriodDynamicConfigurationPlugin")
stageDetail.put("pluginVersion", BuildInfo.version)
stageDetail.put("params", params.asJava)
logger.info()
.field("event", "enter")
.map("stage", stageDetail)
.log()
// input validation
val returnName = params.get("returnName") match {
case Some(returnName) => returnName.trim
case None => throw new Exception("required parameter 'returnName' not found.")
}
val lagDays = params.get("lagDays") match {
case Some(lagDays) => {
try {
lagDays.toInt * -1
} catch {
case e: Exception => throw new Exception(s"cannot convert lagDays ('${lagDays}') to integer.")
}
}
case None => throw new Exception("required parameter 'lagDays' not found.")
}
val leadDays = params.get("leadDays") match {
case Some(leadDays) => {
try {
leadDays.toInt
} catch {
case e: Exception => throw new Exception(s"cannot convert leadDays ('${leadDays}') to integer.")
}
}
case None => throw new Exception("required parameter 'leadDays' not found.")
}
val formatter = params.get("pattern") match {
case Some(pattern) => {
try {
DateTimeFormatter.ofPattern(pattern).withResolverStyle(ResolverStyle.SMART)
} catch {
case e: Exception => throw new Exception(s"cannot parse pattern ('${pattern}').")
}
}
case None => throw new Exception("required parameter 'pattern' not found.")
}
val currentDate = params.get("currentDate") match {
case Some(currentDate) => {
try {
LocalDate.parse(currentDate, formatter)
} catch {
case e: Exception => throw new Exception(s"""cannot parse currentDate ('${currentDate}') with formatter '${params.get("pattern").getOrElse("")}'.""")
}
}
case None => java.time.LocalDate.now
}
// calculate the range
// produces a value that looks like "2018-12-31,2019-01-01,2019-01-02,2019-01-03,2019-01-04,2019-01-05,2019-01-06"
val res = (lagDays to leadDays).map { v =>
formatter.format(currentDate.plusDays(v))
}.mkString(",")
// set the return value
val values = new java.util.HashMap[String, Object]()
values.put(returnName, res)
logger.info()
.field("event", "exit")
.field("duration", System.currentTimeMillis() - startTime)
.map("stage", stageDetail)
.log()
values
}
}
The plugin then needs to be registered in the plugins.config
section of the job configuration and the full plugin name must be listed in your project’s /resources/META-INF/services/au.com.agl.arc.plugins.DynamicConfigurationPlugin
file. See this example.
Note that the resolution order of these plugins is in descending order in that if the the ETL_CONF_LAST_PROCESSING_DAY
was declared in multiple plugins the value set by the plugin with the lower index in the plugins.config
array will take precedence.
The ETL_CONF_LAST_PROCESSING_DAY
variable is then available to be resolved in a standard configuration:
{
"plugins": {
"config": [
{
"type": "au.com.agl.arc.plugins.config.DeltaPeriodDynamicConfigurationPlugin",
"environments": [
"production",
"test"
],
"params": {
"returnName": "ETL_CONF_DELTA_PERIOD",
"lagDays": "10",
"leadDays": "1",
"pattern": "yyyy-MM-dd"
}
}
]
},
"stages": [
{
"type": "DelimitedExtract",
"name": "load customer extract",
"environments": [
"production",
"test"
],
"inputURI": "hdfs://datalake/input/customer/customers_{"${ETL_CONF_DELTA_PERIOD}"}.csv",
"outputView": "customer"
}
]
}
Lifecycle Plugins
Since: 1.3.0
Custom Lifecycle Plugins
allow users to extend the base Arc framework with logic which is executed before
or after
each Arc stage (lifecycle hooks). These stages are useful for implementing things like dataset logging after each stage execution for debugging.
Examples
package au.com.agl.arc.plugins.lifecycle
import java.util
import org.apache.spark.sql.{DataFrame, SparkSession}
import au.com.agl.arc.api.API._
import au.com.agl.arc.plugins.LifecyclePlugin
import au.com.agl.arc.util.Utils
import au.com.agl.arc.util.log.logger.Logger
class DataFramePrinterLifecyclePlugin extends LifecyclePlugin {
var params = Map[String, String]()
override def setParams(p: Map[String, String]) {
params = p
}
override def before(stage: PipelineStage)(implicit spark: SparkSession, logger: au.com.agl.arc.util.log.logger.Logger) {
logger.trace()
.field("event", "before")
.field("stage", stage.name)
.field("stageType", stage.getType)
.log()
}
override def after(stage: PipelineStage, result: Option[DataFrame], isLast: Boolean)(implicit spark: SparkSession, logger: au.com.agl.arc.util.log.logger.Logger) {
logger.trace()
.field("event", "after")
.field("stage", stage.name)
.field("stageType", stage.getType)
.field("isLast", java.lang.Boolean.valueOf(isLast))
.log()
result match {
case Some(df) => {
val numRows = params.get("numRows") match {
case Some(n) => n.toInt
case None => 20
}
val truncate = params.get("truncate") match {
case Some(t) => t.toBoolean
case None => true
}
df.show(numRows, truncate)
}
case None =>
}
}
}
The plugin then needs to be registered by adding the full plugin name must be listed in your project’s /resources/META-INF/services/au.com.agl.arc.plugins.LifecyclePlugin
file.
To execute:
{
"plugins": {
"lifecycle": [
{
"type": "au.com.agl.arc.plugins.lifecycle.DataFramePrinterLifecyclePlugin",
"environments": [
"production",
"test"
],
"params": {
"numRows": "100",
"truncate": "false",
}
}
]
},
"stages": [
...
]
}
Pipeline Stage Plugins
Since: 1.3.0
Custom Pipeline Stage Plugins
allow users to extend the base Arc framework with custom stages which allow the full use of the Spark Scala API. This means that private business logic or code which relies on libraries not included in the base Arc framework can be used - however it is strongly advised to use the inbuilt SQL stages where possible. These stages can use the params
map to be able to pass configuration parameters.
If stages are general purpose enough for use outside your organisation consider creating a pull request against the main Arc repository so that others can benefit.
Examples
package au.com.myfakebusiness.plugins
import au.com.agl.arc.plugins
import au.com.agl.arc.util.log.logger.Logger
import org.apache.spark.sql.{DataFrame, SparkSession}
class MyFakeBusinessAddCopyrightStage extends PipelineStagePlugin {
override def execute(name: String, params: Map[String, String])(implicit spark: SparkSession, logger: Logger): Option[DataFrame] = {
val startTime = System.currentTimeMillis()
val inputView = params.get("inputView").getOrElse("")
val outputView = params.get("outputView").getOrElse("")
val copyrightStatement = params.get("copyrightStatement").getOrElse("")
val stageDetail = new java.util.HashMap[String, Object]()
stageDetail.put("name", name)
stageDetail.put("inputView", inputView)
stageDetail.put("outputView", outputView)
stageDetail.put("copyrightStatement", copyrightStatement)
logger.info()
.field("event", "enter")
.map("stage", stageDetail)
.log()
// get existing dataframe
val df = spark.table(inputView)
// add copyright statement
val enrichedDF = df.withColumn("copyright", lit(copyrightStatement))
// register output view
enrichedDF.createOrReplaceTempView(outputView)
logger.info()
.field("event", "exit")
.field("duration", System.currentTimeMillis() - startTime)
.map("stage", stageDetail)
.log()
Option(enrichedDF)
}
}
The plugin then needs to be registered by adding the full plugin name must be listed in your project’s /resources/META-INF/services/au.com.agl.arc.plugins.PipelineStagePlugin
file.
To execute:
{
"stages": [
{
"type": "au.com.mybusiness.plugins.MyFakeBusinessAddCopyrightStage",
"name": "add copyright to each row",
"environments": [
"production",
"test"
],
"params": {
"inputView": "calculated_dataset",
"copyrightStatement": "copyright 2018 MyBusiness.com.au",
"outputView": "final_dataset"
}
}
]
}
User Defined Functions
Since: 1.3.0
User Defined Functions vs Spark SQL Functions
The inbuilt Spark SQL Functions are heavily optimised by the internal Spark code to a level which custom User Defined Functions cannot be (byte code) - so where possible it is better to use the inbuilt functions.
User Defined Functions
allow users to extend the Spark SQL dialect.
Arc already includes some addtional functions which are not included in the base Spark SQL dialect so any useful generic functions can be included in the Arc repository so that others can benefit.
Examples
Write the code to define the custom User Defined Function
:
package au.com.agl.arc.plugins
import java.util
import org.apache.spark.sql.SQLContext
import au.com.agl.arc.util.log.logger.Logger
class UDFPluginTest extends UDFPlugin {
// one udf plugin can register multiple user defined functions
override def register(sqlContext: SQLContext)(implicit logger: au.com.agl.arc.util.log.logger.Logger): Seq[String] = {
// register the functions so they can be accessed via Spark SQL
// SELECT add_ten(1) AS one_plus_ten
sqlContext.udf.register("add_ten", UDFPluginTest.addTen _ )
// return the list of udf names that were registered for logging
Seq("add_ten")
}
}
object UDFPluginTest {
// add 10 to an incoming integer - DO NOT DO THIS IN PRODUCTION INSTEAD USE SPARK SQL DIRECTLY
def addTen(input: Int): Int = {
input + 10
}
}
The plugin then needs to be registered by adding the full plugin name must be listed in your project’s /resources/META-INF/services/au.com.agl.arc.plugins.UDFPlugin
file.