Flavio Brasil fwbrasil

Please help us validate Quill 2.0.0-SNAPSHOT so we can make the final release. Steps:

Read the migration notes
Add the snapshots repo to your sbt build: resolvers += Resolver.sonatypeRepo("snapshots")
Update the Quill version to 2.0.0-SNAPSHOT
Fix the compilation errors, feel free to ask questions on the gitter channel
Let us know the results

This version has 0️⃣ (yes, ZERO!) known bugs


	t1.flatMap(a => t2.filter(b => b.s == a.s).map(b => b.s))

	SELECT t2.s FROM t1, t2 WHERE t2.s = t1.s

	t1.flatMap(a => t2.map(b => b.s).take(10))

	SELECT x.s FROM t1, (SELECT * FROM t2 LIMIT 10) x

	t1.flatMap(a => t2.filter(b => b.s == a.s).map(b => b.s).take(10))

	package fix

	import scalafix._
	import scala.meta._

	case class ReplaceSymbols(sctx: SemanticCtx) extends SemanticRewrite(sctx) {

	def rewrite(ctx: RewriteCtx): Patch =
	ctx.replaceSymbols(
	"scala.concurrent.Await" -> "io.trane.future.scala.Await",

	# Requirement: the project must be on Scala 2.12.3

	# Step 1 - Add the scalafix sbt plugin
	echo 'addSbtPlugin("ch.epfl.scala" % "sbt-scalafix" % "0.5.0-M4")' >> projecs/plugins.sbt

	# Step 2 (MANUAL) - Add the traneio Scala Future dependency to the sbt build
	# resolvers += Resolver.sonatypeRepo("snapshots")
	# libraryDependencies += "io.trane" % "future-scala" % "0.2.3-SNAPSHOT"

	# Step 2 - Run the scalafix rewrite

	/**
	*
	* THIS IS JUST A DSL PROTOTYPE
	*
	* First tentative of a DSL for defining macro transformations in Java.
	* There's a lot of noise because of the absence of pattern matching and tuples.
	* This is a prototype of one of the if/else transformations in Monadless:
	*
	* case q"if($cond) $ifTrue else $ifFalse" =>
	* (ifTrue, ifFalse) match {

	def topHashtags(tweets: Dataset[Tweet], n: Int): Dataset[(String, Long)] =
	run { // produce a dataset from the Quill query
	liftQuery(tweets) // trasform the dataset into a Quill query
	.concatMap(_.text.split(" ")) // split into words and unnest results
	.filter(_.startsWith("#")) // filter hashtag words
	.map(_.toLowerCase) // normalize hashtags
	.groupBy(word => word) // group by each hashtag
	.map { // map word list to its count
	case (word, list) =>
	(word, list.size)

	def topHashtags(tweets: Dataset[Tweet], n: Int): Dataset[(String, BigInt)] =
	tweets
	.select($"text".as[String]) // select the text column (Dataframe)
	.flatMap(_.split("\\s+")) // split it into words (Dataset)
	.filter(_.startsWith("#")) // filter hashtag words (Dataset)
	.map(_.toLowerCase) // normalize hashtags (Dataset)
	.groupBy($"value") // group by each hashtag (Dataframe)
	.agg(count("*") as "count") // aggregate the count (Dataframe)
	.orderBy($"count" desc) // order (Datafeame)
	.limit(n) // limit to top results (Dataframe)

	def topHashtags(tweets: DataFrame, n: Int): DataFrame =
	tweets
	.select(explode(split($"text", "\\s+"))) // split it into words
	.select(lower($"col") as "word") // normalize hashtags
	.filter("word like '#%'") // filter hashtag words
	.groupBy($"word") // group by each hashtag
	.agg(count("*") as "count") // aggregate the count
	.orderBy($"count" desc) // order
	.limit(n) // limit to top results

	def topHashtags(tweets: RDD[Tweet], n: Int): Array[(String, BigInt)] =
	tweets
	.flatMap(_.text.split("\\s+")) // split it into words
	.filter(_.startsWith("#")) // filter hashtag words
	.map(_.toLowerCase) // normalize hashtags
	.map((_, BigInt(1))) // create tuples for counting
	.reduceByKey((a, b) => a + b) // accumulate counters
	.top(n)(Ordering.by(_._2)) // return ordered top hashtags