Merge pull request #3 from audienceproject/scala3

Upgrade to Scala 3

README.md

@@ -7,77 +7,93 @@ Single node, in-memory DataFrame analytics library.
 * Fluent expression DSL
 * Immutable public API; `Array` under the hood
 
+## News
+
+* 2024-03-22: Upgraded to Scala 3 with version `0.2.0`. Some parts of the source code have been completely rewritten,
+  but the feature set remains the same. The intention is for Crossbow to be a Scala 3 library going forward.
+
 ## Installing
+
 The library is available through Maven Central.
 
 SBT style dependency: `"com.audienceproject" %% "crossbow" % "latest"`
 
 # API
-```scala
-import com.audienceproject.crossbow.DataFrame
-import com.audienceproject.crossbow.Implicits._
+
+```scala 3
+import com.audienceproject.crossbow.{*, given}
 
 val data = Seq(("a", 1), ("b", 2), ("c", 3))
 val df = DataFrame.fromSeq(data)
 
 df.printSchema()
+
 /**
- * _0: String
+ * _0: string
  * _1: int
  */
 
 df.as[(String, Int)].foreach(println)
+
 /**
- * ("a", 1)
- * ("b", 2)
- * ("c", 3)
+ * (a, 1)
+ * (b, 2)
+ * (c, 3)
  */
 ```
 
 ## Transforming
-```scala
+
+```scala 3
+val df = Seq((1, 2), (3, 4)).toDataFrame("x", "y")
 df.select($"x" + $"y" / 2d as "avg", ($"x", $"y") as "tuple")
 
 // Lambda functions
-val toUpper = lambda[String, String](_.toUpperCase)
-df.select(toUpper($"a") as "upperCaseA")
+val pythagoras = lambda[(Int, Int), Double]:
+  (a, b) => math.sqrt(a * a + b * b)
+df.select(pythagoras($"x", $"y"))
 ```
 
 ## Filtering
-```scala
+
+```scala 3
 df.filter($"x" >= 2 && $"y" % 2 =!= 0)
 ```
 
 ## Grouping
-```scala
-df.groupBy($"someKey").agg(sum($"x") / count($"x") as "avg", collect($"x") as "xs")
+
+```scala 3
+val df = Seq(("foo", 1), ("foo", 2), ("bar", 3)).toDataFrame("someKey", "x")
+df.groupBy($"someKey").agg(sum($"x") / count() as "avg", collect($"x") as "xs")
 
 // Custom aggregators
 val product = reducer[Int, Int](1)(_ * _)
 df.groupBy($"someKey").agg(product($"x") as "product")
 ```
 
 ## Sorting
-```scala
+
+```scala 3
 df.sortBy($"x")
 
 // Sorting on 'x' first, then 'y'
-df.sortBy(($"x", $"y"))
+df.sortBy(($"someKey", $"x"))
 
 // Sorting with explicit ordering (e.g. integer descending)
-import com.audienceproject.crossbow.expr.Order
-df.sortBy($"x", Order.by(Ordering.Int.reverse))
+df.sortBy($"x")(using Order.by(Ordering.Int.reverse))
 ```
 
 ## Joining
-```scala
+
+```scala 3
+val otherDf = Seq(("foo", 1, 10d), ("foo", 2, 20d), ("bar", 3, 30d)).toDataFrame("someKey", "x", "y")
+
 // Inner join
 df.join(otherDf, $"someKey")
 
 // Inner join on multiple columns
-df.join(otherDf, ($"key1", $"key2"))
+df.join(otherDf, ($"someKey", $"x"))
 
 // Other join types
-import com.audienceproject.crossbow.JoinType
 df.join(otherDf, $"someKey", JoinType.LeftOuter)
 ```

build.sbt

@@ -2,15 +2,13 @@ organization := "com.audienceproject"
 
 name := "crossbow"
 
-version := "0.1.6"
+version := "0.2.0"
 
-scalaVersion := "2.13.6"
-crossScalaVersions := Seq(scalaVersion.value, "2.12.12", "2.11.12")
+scalaVersion := "3.3.3"
 
-scalacOptions ++= Seq("-deprecation", "-feature", "-language:existentials")
+scalacOptions ++= Seq("-deprecation", "-feature", "-language:implicitConversions")
 
-libraryDependencies += "org.scala-lang" % "scala-reflect" % scalaVersion.value
-libraryDependencies += "org.scalatest" %% "scalatest-funsuite" % "3.2.0" % "test"
+libraryDependencies += "org.scalatest" %% "scalatest-funsuite" % "3.2.17" % "test"
 
 /**
  * Maven specific settings for publishing to Maven central.

project/build.properties

@@ -1 +1 @@
-sbt.version = 1.6.2
+sbt.version = 1.9.8