Add distinct in ColumnSelectionDsl.

Author: nikitinas

docs/StardustDocs/topics/ColumnSelectors.md

@@ -209,6 +209,9 @@ df.select { allDfs().filter { it.name.startsWith("year") } }
 
 // exclude columns from column set
 df.select { allDfs().except { age } }
+
+// keep only unique columns
+df.select { (colsOf<Int>() and age).distinct() }
 ```
 
 <!---END-->

src/main/kotlin/org/jetbrains/kotlinx/dataframe/api/ColumnsSelectionDsl.kt

@@ -20,6 +20,7 @@ import org.jetbrains.kotlinx.dataframe.impl.DataFrameReceiver
 import org.jetbrains.kotlinx.dataframe.impl.columnName
 import org.jetbrains.kotlinx.dataframe.impl.columns.ColumnAccessorImpl
 import org.jetbrains.kotlinx.dataframe.impl.columns.ColumnsList
+import org.jetbrains.kotlinx.dataframe.impl.columns.DistinctColumnSet
 import org.jetbrains.kotlinx.dataframe.impl.columns.allColumnsExcept
 import org.jetbrains.kotlinx.dataframe.impl.columns.createColumnSet
 import org.jetbrains.kotlinx.dataframe.impl.columns.getAt
@@ -327,6 +328,8 @@ public interface ColumnsSelectionDsl<out T> : ColumnSelectionDsl<T>, SingleColum
 
     // endregion
 
+    public fun <C> ColumnSet<C>.distinct(): ColumnSet<C> = DistinctColumnSet(this)
+
     public operator fun <C> String.invoke(newColumnExpression: RowExpression<T, C>): DataColumn<C> =
         newColumnWithActualType(this, newColumnExpression)
 

src/main/kotlin/org/jetbrains/kotlinx/dataframe/impl/columns/DistinctColumnSet.kt

@@ -0,0 +1,10 @@
+package org.jetbrains.kotlinx.dataframe.impl.columns
+
+import org.jetbrains.kotlinx.dataframe.columns.ColumnResolutionContext
+import org.jetbrains.kotlinx.dataframe.columns.ColumnSet
+import org.jetbrains.kotlinx.dataframe.columns.ColumnWithPath
+
+internal class DistinctColumnSet<T>(val src: ColumnSet<T>) : ColumnSet<T> {
+    override fun resolve(context: ColumnResolutionContext): List<ColumnWithPath<T>> =
+        src.resolve(context).distinctBy { it.path }
+}

src/test/kotlin/org/jetbrains/kotlinx/dataframe/samples/api/Access.kt

@@ -802,6 +802,9 @@ class Access : TestBase() {
 
         // exclude columns from column set
         df.select { allDfs().except { age } }
+
+        // keep only unique columns
+        df.select { (colsOf<Int>() and age).distinct() }
         // SampleEnd
     }