Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Added implementation of row level headers , params in http method-3.1 #40

Merged
merged 5 commits into from
Aug 31, 2023
Merged

Added implementation of row level headers , params in http method-3.1 #40

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
9db288a
Added Http Row with headers and params from row and its test cases
sharath-sg2706 Aug 23, 2023
1425395
Added Http Row with headers and params from row and its test cases
sharath-sg2706 Aug 23, 2023
feadfe3
Added implementation of row level headers , params in http method
sharath-sg2706 Aug 25, 2023
c783bac
Added implementation of row level headers , params in http method
sharath-sg2706 Aug 25, 2023
1bdc1c7
Added implementation of row level headers , params in http method-3.1
sharath-sg2706 Aug 31, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
447 changes: 339 additions & 108 deletions README.md
View file
Open in desktop

Large diffs are not rendered by default.

184 changes: 107 additions & 77 deletions src/main/scala/com/github/music/of/the/ainur/almaren/http/HTTPConnector.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,72 +13,100 @@ import scala.concurrent.duration.Duration
import scala.util.{Failure, Success, Try}

final case class Response(
`__ID__`:String,
`__BODY__`:Option[String] = None,
`__HEADER__`:Map[String,Seq[String]] = Map(),
`__STATUS_CODE__`:Option[Int] = None,
`__STATUS_MSG__`:Option[String] = None,
`__ERROR__`:Option[String] = None,
`__ELAPSED_TIME__`:Long,
`__URL__`:String,
`__DATA__`: String)
`__ID__`: String,
`__BODY__`: Option[String] = None,
`__HEADER__`: Map[String, Seq[String]] = Map(),
`__STATUS_CODE__`: Option[Int] = None,
`__STATUS_MSG__`: Option[String] = None,
`__ERROR__`: Option[String] = None,
`__ELAPSED_TIME__`: Long,
`__URL__`: String,
`__DATA__`: String)

final case class ResponseBatch(
`__ID__`:Seq[String],
`__BODY__`:Option[String] = None,
`__HEADER__`:Map[String,Seq[String]] = Map(),
`__STATUS_CODE__`:Option[Int] = None,
`__STATUS_MSG__`:Option[String] = None,
`__ERROR__`:Option[String] = None,
`__ELAPSED_TIME__`:Long,
`__URL__`:String,
`__DATA__`:String)
`__ID__`: Seq[String],
`__BODY__`: Option[String] = None,
`__HEADER__`: Map[String, Seq[String]] = Map(),
`__STATUS_CODE__`: Option[Int] = None,
`__STATUS_MSG__`: Option[String] = None,
`__ERROR__`: Option[String] = None,
`__ELAPSED_TIME__`: Long,
`__URL__`: String,
`__DATA__`: String)


object Alias {
val DataCol = "__DATA__"
val IdCol = "__ID__"
val UrlCol = "__URL__"
val ParamsCol = "__PARAMS__"
val HiddenParamsCol = "__HIDDEN_PARAMS__"
val HeadersCol = "__HEADERS__"
}


private[almaren] case class HTTP(
headers: Map[String, String],
params: Map[String, String],
hiddenParams: Map[String, String],
method: String,
requestHandler: (Row, Session, String, Map[String, String], Map[String, String], String, Int, Int) => requests.Response,
session: () => requests.Session,
connectTimeout: Int,
readTimeout: Int,
threadPoolSize: Int,
batchSize: Int) extends Main {
headers: Map[String, String],
params: Map[String, String],
hiddenParams: Map[String, String],
method: String,
requestHandler: (Row, Session, String, Map[String, String], Map[String, String], String, Int, Int) => requests.Response,
session: () => requests.Session,
connectTimeout: Int,
readTimeout: Int,
threadPoolSize: Int,
batchSize: Int) extends Main {

private def columnExists(df: DataFrame, colName: String): Option[String] =
Some(colName).filter(df.columns.contains)

private def getRowParam(row: Row, colNameExists: Option[String]): Map[String, String] =
colNameExists.map(name => row.getAs[Map[String, String]](name)).getOrElse(Map())

override def core(df: DataFrame): DataFrame = {
logger.info(s"headers:{$headers},params:{$params}, method:{$method}, connectTimeout:{$connectTimeout}, readTimeout{$readTimeout}, threadPoolSize:{$threadPoolSize}, batchSize:{$batchSize}")

import df.sparkSession.implicits._

val headersColExists = columnExists(df,Alias.HeadersCol)
val paramsColExists = columnExists(df,Alias.ParamsCol)
val hiddenParamsColExists = columnExists(df,Alias.HiddenParamsCol)

val result = df.mapPartitions(partition => {

implicit val ec:ExecutionContext = ExecutionContext.fromExecutor(Executors.newFixedThreadPool(threadPoolSize))
val data:Iterator[Future[Seq[Response]]] = partition.grouped(batchSize).map(rows => Future {
implicit val ec: ExecutionContext = ExecutionContext.fromExecutor(Executors.newFixedThreadPool(threadPoolSize))
val data: Iterator[Future[Seq[Response]]] = partition.grouped(batchSize).map(rows => Future {
val s = session()
rows.map(row => request(row,s))
rows.map(row => request(row, s, headersColExists, paramsColExists, hiddenParamsColExists))
})
val requests:Future[Iterator[Seq[Response]]] = Future.sequence(data)
val requests: Future[Iterator[Seq[Response]]] = Future.sequence(data)
Await.result(requests, Duration.Inf).flatten
})
result.toDF
}

private def request(row:Row, session:Session): Response = {
private def request(row: Row, session: Session, headersColExists: Option[String], paramsColExists: Option[String], hiddenParamsColExists: Option[String]): Response = {
val url = row.getAs[Any](Alias.UrlCol).toString
val startTime = System.currentTimeMillis()
val response = Try(requestHandler(row,session,url,headers,params ++ hiddenParams,method,connectTimeout,readTimeout))

val headersRow = getRowParam(row, headersColExists)
val paramsRow = getRowParam(row, paramsColExists)
val hiddenParamsRow = getRowParam(row, hiddenParamsColExists)

val allHeaders = headers ++ headersRow
val allParams = params ++ paramsRow
val allHiddenParams = hiddenParams ++ hiddenParamsRow

logger.info(s"headers:{$allHeaders},params:{$allParams}")
val response = Try(requestHandler(row, session, url, allHeaders, allParams ++ allHiddenParams, method, connectTimeout, readTimeout))
val elapsedTime = System.currentTimeMillis() - startTime
val id = row.getAs[Any](Alias.IdCol).toString
val data = row.getAs[Any](Alias.DataCol).toString

val data = method.toUpperCase match {
case "PUT" | "POST" | "DELETE" => row.getAs[Any](Alias.DataCol).toString
case _ => ""
}

def getResponse(r: requests.Response) = Response(
id,
Some(r.text()),
Expand All @@ -105,18 +133,18 @@ private[almaren] case class HTTP(
}

private[almaren] case class HTTPBatch(
url: String,
headers: Map[String, String],
params: Map[String, String],
hiddenParams: Map[String, String],
method: String,
requestHandler: (String, Session, String, Map[String, String], Map[String, String], String, Int, Int) => requests.Response,
session: () => requests.Session,
connectTimeout: Int,
readTimeout: Int,
batchSize: Int,
batchDelimiter: (Seq[Row]) => String
) extends Main {
url: String,
headers: Map[String, String],
params: Map[String, String],
hiddenParams: Map[String, String],
method: String,
requestHandler: (String, Session, String, Map[String, String], Map[String, String], String, Int, Int) => requests.Response,
session: () => requests.Session,
connectTimeout: Int,
readTimeout: Int,
batchSize: Int,
batchDelimiter: (Seq[Row]) => String
) extends Main {

override def core(df: DataFrame): DataFrame = {
logger.info(s"url:{$url}, headers:{$headers},params:{$params}, method:{$method}, connectTimeout:{$connectTimeout}, readTimeout{$readTimeout}, batchSize:{$batchSize}")
Expand All @@ -143,7 +171,9 @@ private[almaren] case class HTTPBatch(
`__DATA__` = data
)

Try{request(data,s)} match {
Try {
request(data, s)
} match {
case Success(r) => getResponse(r)
case Failure(re: RequestFailedException) => getResponse(re.response)
case Failure(f) => {
Expand All @@ -155,33 +185,33 @@ private[almaren] case class HTTPBatch(
`__URL__` = url,
`__DATA__` = data
)

}
}
})
})
result.toDF
}

private def request(data:String, session:Session): requests.Response =
requestHandler(data,session,url,headers,params ++ hiddenParams,method,connectTimeout,readTimeout)
private def request(data: String, session: Session): requests.Response =
requestHandler(data, session, url, headers, params ++ hiddenParams, method, connectTimeout, readTimeout)


}

private[almaren] trait HTTPConnector extends Core {

def http(
headers: Map[String, String] = Map(),
params: Map[String, String] = Map(),
hiddenParams: Map[String, String] = Map(),
method: String,
requestHandler: (Row, Session, String, Map[String, String], Map[String, String], String, Int, Int) => requests.Response = HTTPConn.defaultHandler,
session: () => requests.Session = HTTPConn.defaultSession,
connectTimeout: Int = 60000,
readTimeout: Int = 1000,
threadPoolSize: Int = 1,
batchSize: Int = 5000): Option[Tree] =
headers: Map[String, String] = Map(),
params: Map[String, String] = Map(),
hiddenParams: Map[String, String] = Map(),
method: String,
requestHandler: (Row, Session, String, Map[String, String], Map[String, String], String, Int, Int) => requests.Response = HTTPConn.defaultHandler,
session: () => requests.Session = HTTPConn.defaultSession,
connectTimeout: Int = 60000,
readTimeout: Int = 1000,
threadPoolSize: Int = 1,
batchSize: Int = 5000): Option[Tree] =
HTTP(
headers,
params,
Expand All @@ -196,18 +226,18 @@ private[almaren] trait HTTPConnector extends Core {
)

def httpBatch(
url: String,
headers: Map[String, String] = Map(),
params: Map[String, String] = Map(),
hiddenParams: Map[String, String] = Map(),
method: String,
requestHandler: (String, Session, String, Map[String, String], Map[String, String], String, Int, Int) => requests.Response = HTTPConn.defaultHandlerBatch,
session: () => requests.Session = HTTPConn.defaultSession,
connectTimeout: Int = 60000,
readTimeout: Int = 1000,
batchSize: Int = 5000,
batchDelimiter: (Seq[Row]) => String = HTTPConn.defaultBatchDelimiter
): Option[Tree] =
url: String,
headers: Map[String, String] = Map(),
params: Map[String, String] = Map(),
hiddenParams: Map[String, String] = Map(),
method: String,
requestHandler: (String, Session, String, Map[String, String], Map[String, String], String, Int, Int) => requests.Response = HTTPConn.defaultHandlerBatch,
session: () => requests.Session = HTTPConn.defaultSession,
connectTimeout: Int = 60000,
readTimeout: Int = 1000,
batchSize: Int = 5000,
batchDelimiter: (Seq[Row]) => String = HTTPConn.defaultBatchDelimiter
): Option[Tree] =
HTTPBatch(
url,
headers,
Expand All @@ -224,7 +254,7 @@ private[almaren] trait HTTPConnector extends Core {
}

object HTTPConn {
val defaultHandler = (row:Row, session:Session, url:String, headers:Map[String, String], params:Map[String, String], method:String, connectTimeout:Int, readTimeout:Int) => {
val defaultHandler = (row: Row, session: Session, url: String, headers: Map[String, String], params: Map[String, String], method: String, connectTimeout: Int, readTimeout: Int) => {
method.toUpperCase match {
case "GET" => session.get(url, headers = headers, params = params, readTimeout = readTimeout, connectTimeout = connectTimeout)
case "DELETE" => session.delete(url, headers = headers, params = params, data = row.getAs[String](Alias.DataCol), readTimeout = readTimeout, connectTimeout = connectTimeout)
Expand All @@ -236,7 +266,7 @@ object HTTPConn {
}
}

val defaultHandlerBatch = (data:String, session:Session, url:String, headers:Map[String, String], params:Map[String, String], method:String, connectTimeout:Int, readTimeout:Int) => {
val defaultHandlerBatch = (data: String, session: Session, url: String, headers: Map[String, String], params: Map[String, String], method: String, connectTimeout: Int, readTimeout: Int) => {
method.toUpperCase match {
case "GET" => session.get(url, headers = headers, params = params, readTimeout = readTimeout, connectTimeout = connectTimeout)
case "DELETE" => session.delete(url, headers = headers, params = params, data = data, readTimeout = readTimeout, connectTimeout = connectTimeout)
Expand All @@ -248,9 +278,9 @@ object HTTPConn {
}
}

val defaultBatchDelimiter = (rows:Seq[Row]) => rows.map(row => row.getAs[String](Alias.DataCol)).mkString("\n")
val defaultBatchDelimiter = (rows: Seq[Row]) => rows.map(row => row.getAs[String](Alias.DataCol)).mkString("\n")

val defaultSession = () => requests.Session()

implicit class HTTPImplicit(val container: Option[Tree]) extends HTTPConnector
}
}
Binary file added src/test/resources/data/postRowSession.parquet/._SUCCESS.crc
View file
Open in desktop
Binary file not shown.
Binary file added ...wSession.parquet/.part-00000-08db27e8-3820-44fa-b4ab-95e079e04116-c000.snappy.parquet.crc
View file
Open in desktop
Binary file not shown.
Empty file added src/test/resources/data/postRowSession.parquet/_SUCCESS
View file
Open in desktop
Empty file.
Binary file added ...ostRowSession.parquet/part-00000-08db27e8-3820-44fa-b4ab-95e079e04116-c000.snappy.parquet
View file
Open in desktop
Binary file not shown.
Binary file added src/test/resources/data/postRowWithoutSession.parquet/._SUCCESS.crc
View file
Open in desktop
Binary file not shown.
Binary file added ...tSession.parquet/.part-00000-5d7e75fd-c2c8-4d17-a8cb-3782b2626f84-c000.snappy.parquet.crc
View file
Open in desktop
Binary file not shown.
Empty file added src/test/resources/data/postRowWithoutSession.parquet/_SUCCESS
View file
Open in desktop
Empty file.
Binary file added ...ithoutSession.parquet/part-00000-5d7e75fd-c2c8-4d17-a8cb-3782b2626f84-c000.snappy.parquet
View file
Open in desktop
Binary file not shown.
Loading
Loading