Merge branch 'master' into aggregators

Author: Jolanrensen

core/api/core.api

@@ -5404,6 +5404,8 @@ public final class org/jetbrains/kotlinx/dataframe/impl/api/SchemaKt {
 public final class org/jetbrains/kotlinx/dataframe/impl/api/ToDataFrameKt {
 	public static final fun convertToDataFrame (Ljava/lang/Iterable;Lkotlin/reflect/KClass;Ljava/util/List;Ljava/util/Set;Ljava/util/Set;Ljava/util/Set;I)Lorg/jetbrains/kotlinx/dataframe/DataFrame;
 	public static final fun createDataFrameImpl (Ljava/lang/Iterable;Lkotlin/reflect/KClass;Lkotlin/jvm/functions/Function1;)Lorg/jetbrains/kotlinx/dataframe/DataFrame;
+	public static final fun getHasProperties (Lkotlin/reflect/KClass;)Z
+	public static final fun isValueType (Lkotlin/reflect/KClass;)Z
 }
 
 public final class org/jetbrains/kotlinx/dataframe/impl/api/ToSequenceKt {

core/build.gradle.kts

@@ -185,6 +185,7 @@ tasks.withType<ShadowJar> {
         exclude(dependency("org.jetbrains.kotlinx:kotlinx-datetime-jvm:.*"))
         exclude(dependency("commons-io:commons-io:.*"))
         exclude(dependency("commons-io:commons-csv:.*"))
+        exclude(dependency("org.apache.commons:commons-csv:.*"))
         exclude(dependency("org.slf4j:slf4j-api:.*"))
         exclude(dependency("io.github.microutils:kotlin-logging-jvm:.*"))
         exclude(dependency("org.jetbrains.kotlinx:kotlinx-serialization-core-jvm:.*"))

core/generated-sources/src/main/kotlin/org/jetbrains/kotlinx/dataframe/DataFrame.kt

@@ -4,6 +4,8 @@ import org.jetbrains.kotlinx.dataframe.aggregation.Aggregatable
 import org.jetbrains.kotlinx.dataframe.aggregation.AggregateGroupedBody
 import org.jetbrains.kotlinx.dataframe.annotations.AccessApiOverload
 import org.jetbrains.kotlinx.dataframe.annotations.HasSchema
+import org.jetbrains.kotlinx.dataframe.annotations.Interpretable
+import org.jetbrains.kotlinx.dataframe.annotations.Refine
 import org.jetbrains.kotlinx.dataframe.api.ColumnsSelectionDsl
 import org.jetbrains.kotlinx.dataframe.api.add
 import org.jetbrains.kotlinx.dataframe.api.cast
@@ -71,6 +73,8 @@ public interface DataFrame<out T> :
 
     // endregion
 
+    @Refine
+    @Interpretable("AggregateRow")
     public fun <R> aggregate(body: AggregateGroupedBody<T, R>): DataRow<T>
 
     // region get columns

core/generated-sources/src/main/kotlin/org/jetbrains/kotlinx/dataframe/api/DataColumnType.kt

@@ -7,8 +7,6 @@ import org.jetbrains.kotlinx.dataframe.columns.ColumnGroup
 import org.jetbrains.kotlinx.dataframe.columns.ColumnKind
 import org.jetbrains.kotlinx.dataframe.columns.FrameColumn
 import org.jetbrains.kotlinx.dataframe.columns.ValueColumn
-import org.jetbrains.kotlinx.dataframe.impl.isNothing
-import org.jetbrains.kotlinx.dataframe.impl.projectTo
 import org.jetbrains.kotlinx.dataframe.type
 import org.jetbrains.kotlinx.dataframe.typeClass
 import org.jetbrains.kotlinx.dataframe.util.IS_COMPARABLE
@@ -21,8 +19,11 @@ import kotlin.contracts.contract
 import kotlin.reflect.KClass
 import kotlin.reflect.KType
 import kotlin.reflect.KTypeProjection
+import kotlin.reflect.KVariance
+import kotlin.reflect.full.createType
 import kotlin.reflect.full.isSubclassOf
 import kotlin.reflect.full.isSubtypeOf
+import kotlin.reflect.full.withNullability
 import kotlin.reflect.typeOf
 
 public fun AnyCol.isColumnGroup(): Boolean {
@@ -53,14 +54,13 @@ public fun AnyCol.isBigNumber(): Boolean = isSubtypeOf<BigInteger?>() || isSubty
 
 public fun AnyCol.isList(): Boolean = typeClass == List::class
 
-/** Returns `true` if [this] column is inter-comparable, i.e.
+/** Returns `true` if [this] column is intra-comparable, i.e.
  * its values can be compared with each other and thus ordered.
  *
  * If true, operations like [`min()`][AnyCol.min], [`max()`][AnyCol.max], [`median()`][AnyCol.median], etc.
  * will work.
  *
- * Technically, this means the values' common type is a subtype of [Comparable] with
- * the type argument not being [Nothing]. */
+ * Technically, this means the values' common type `T(?)` is a subtype of [Comparable]`<in T>(?)` */
 @Deprecated(
     message = IS_COMPARABLE,
     replaceWith = ReplaceWith(IS_COMPARABLE_REPLACE, IS_INTER_COMPARABLE_IMPORT),
@@ -69,21 +69,24 @@ public fun AnyCol.isList(): Boolean = typeClass == List::class
 public fun AnyCol.isComparable(): Boolean = valuesAreComparable()
 
 /**
- * Returns `true` if [this] column is inter-comparable, i.e.
+ * Returns `true` if [this] column is intra-comparable, i.e.
  * its values can be compared with each other and thus ordered.
  *
  * If true, operations like [`min()`][AnyCol.min], [`max()`][AnyCol.max], [`median()`][AnyCol.median], etc.
  * will work.
  *
- * Technically, this means the values' common type is a subtype of [Comparable] with
- * the type argument not being [Nothing].
+ * Technically, this means the values' common type `T(?)` is a subtype of [Comparable]`<in T>(?)`
  */
 public fun AnyCol.valuesAreComparable(): Boolean =
-    isSubtypeOf<Comparable<*>?>() &&
-        type().projectTo(Comparable::class).arguments[0].let {
-            it != KTypeProjection.STAR &&
-                it.type?.isNothing != true
-        }
+    isValueColumn() &&
+        isSubtypeOf(
+            Comparable::class.createType(
+                arguments = listOf(
+                    KTypeProjection(KVariance.IN, type().withNullability(false)),
+                ),
+                nullable = hasNulls(),
+            ),
+        )
 
 @PublishedApi
 internal fun AnyCol.isPrimitive(): Boolean = typeClass.isPrimitive()

core/generated-sources/src/main/kotlin/org/jetbrains/kotlinx/dataframe/api/toDataFrame.kt

@@ -12,6 +12,8 @@ import org.jetbrains.kotlinx.dataframe.annotations.Refine
 import org.jetbrains.kotlinx.dataframe.columns.ColumnPath
 import org.jetbrains.kotlinx.dataframe.impl.ColumnNameGenerator
 import org.jetbrains.kotlinx.dataframe.impl.api.createDataFrameImpl
+import org.jetbrains.kotlinx.dataframe.impl.api.hasProperties
+import org.jetbrains.kotlinx.dataframe.impl.api.isValueType
 import org.jetbrains.kotlinx.dataframe.impl.asList
 import org.jetbrains.kotlinx.dataframe.impl.columnName
 import org.jetbrains.kotlinx.dataframe.impl.columns.createColumnGuessingType
@@ -26,7 +28,15 @@ import kotlin.reflect.KProperty
 @Interpretable("toDataFrameDefault")
 public inline fun <reified T> Iterable<T>.toDataFrame(): DataFrame<T> =
     toDataFrame {
-        properties()
+        // check if type is value: primitives, primitive arrays, datetime types etc.,
+        // or has no properties
+        if (T::class.isValueType || !T::class.hasProperties) {
+            // create a single `value` column
+            ValueProperty<T>::value from { it }
+        } else {
+            // otherwise creates columns based on properties
+            properties()
+        }
     }
 
 @Refine
@@ -217,99 +227,6 @@ public inline fun <reified T> Iterable<T>.toDataFrame(columnName: String): DataF
 
 // region toDataFrame overloads for built-in types
 
-/*
-Without overloads Iterable<String>.toDataFrame produces unexpected result
-
-
-```
-val string = listOf("aaa", "aa", null)
-string.toDataFrame()
-```
-=>
-  length
-0    3
-1    2
-2 null
- */
-
-@JvmName("toDataFrameByte")
-public inline fun <reified B : Byte?> Iterable<B>.toDataFrame(): DataFrame<ValueProperty<B>> =
-    toDataFrame {
-        ValueProperty<B>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameShort")
-public inline fun <reified S : Short?> Iterable<S>.toDataFrame(): DataFrame<ValueProperty<S>> =
-    toDataFrame {
-        ValueProperty<S>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameInt")
-public inline fun <reified I : Int?> Iterable<I>.toDataFrame(): DataFrame<ValueProperty<I>> =
-    toDataFrame {
-        ValueProperty<I>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameLong")
-public inline fun <reified L : Long?> Iterable<L>.toDataFrame(): DataFrame<ValueProperty<L>> =
-    toDataFrame {
-        ValueProperty<L>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameString")
-public inline fun <reified S : String?> Iterable<S>.toDataFrame(): DataFrame<ValueProperty<S>> =
-    toDataFrame {
-        ValueProperty<S>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameChar")
-public inline fun <reified C : Char?> Iterable<C>.toDataFrame(): DataFrame<ValueProperty<C>> =
-    toDataFrame {
-        ValueProperty<C>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameBoolean")
-public inline fun <reified B : Boolean?> Iterable<B>.toDataFrame(): DataFrame<ValueProperty<B>> =
-    toDataFrame {
-        ValueProperty<B>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameFloat")
-public inline fun <reified F : Float?> Iterable<F>.toDataFrame(): DataFrame<ValueProperty<F>> =
-    toDataFrame {
-        ValueProperty<F>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameDouble")
-public inline fun <reified D : Double?> Iterable<D>.toDataFrame(): DataFrame<ValueProperty<D>> =
-    toDataFrame {
-        ValueProperty<D>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameUByte")
-public inline fun <reified U : UByte?> Iterable<U>.toDataFrame(): DataFrame<ValueProperty<U>> =
-    toDataFrame {
-        ValueProperty<U>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameUShort")
-public inline fun <reified U : UShort?> Iterable<U>.toDataFrame(): DataFrame<ValueProperty<U>> =
-    toDataFrame {
-        ValueProperty<U>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameUInt")
-public inline fun <reified U : UInt?> Iterable<U>.toDataFrame(): DataFrame<ValueProperty<U>> =
-    toDataFrame {
-        ValueProperty<U>::value from { it }
-    }.cast()
-
-@JvmName("toDataFrameULong")
-public inline fun <reified U : ULong?> Iterable<U>.toDataFrame(): DataFrame<ValueProperty<U>> =
-    toDataFrame {
-        ValueProperty<U>::value from { it }
-    }.cast()
-
 @DataSchema
 public interface ValueProperty<T> {
     public val value: T

core/generated-sources/src/main/kotlin/org/jetbrains/kotlinx/dataframe/impl/api/toDataFrame.kt

@@ -23,7 +23,8 @@ import org.jetbrains.kotlinx.dataframe.impl.projectUpTo
 import org.jetbrains.kotlinx.dataframe.impl.schema.sortWithConstructor
 import java.lang.reflect.InvocationTargetException
 import java.lang.reflect.Method
-import java.time.temporal.Temporal
+import java.time.temporal.TemporalAccessor
+import java.time.temporal.TemporalAmount
 import kotlin.reflect.KCallable
 import kotlin.reflect.KClass
 import kotlin.reflect.KProperty
@@ -37,23 +38,51 @@ import kotlin.reflect.jvm.isAccessible
 import kotlin.reflect.jvm.javaField
 import kotlin.reflect.typeOf
 
+// non-standard value types (not supertypes, but exact types)
 private val valueTypes = setOf(
+    Any::class,
+    Unit::class,
+    Char::class,
+    UByte::class,
+    UShort::class,
+    UInt::class,
+    ULong::class,
     String::class,
     Boolean::class,
     kotlin.time.Duration::class,
     kotlinx.datetime.LocalDate::class,
     kotlinx.datetime.LocalDateTime::class,
     kotlinx.datetime.Instant::class,
+    kotlinx.datetime.TimeZone::class,
+    kotlinx.datetime.DateTimePeriod::class,
+    kotlinx.datetime.DateTimeUnit::class,
 )
 
+/**
+ * Checks if `KClass` is a value type (number, datetime, string, etc.)
+ * Should be aligned with `ConeKotlinType.isValueType()` in
+ * plugins/kotlin-dataframe/src/org/jetbrains/kotlinx/dataframe/plugin/impl/api/toDataFrame.kt
+ */
+@PublishedApi
 internal val KClass<*>.isValueType: Boolean
     get() =
         this in valueTypes ||
             this.isSubclassOf(Number::class) ||
             this.isSubclassOf(Enum::class) ||
-            this.isSubclassOf(Temporal::class) ||
+            // all java datetime types
+            this.isSubclassOf(TemporalAccessor::class) ||
+            this.isSubclassOf(TemporalAmount::class) ||
             this.isArray
 
+/**
+ * Checks if `KClass` has public properties / getter functions (for pojo-like classes).
+ */
+@PublishedApi
+internal val KClass<*>.hasProperties: Boolean
+    get() = this.memberProperties.any { it.visibility == KVisibility.PUBLIC } ||
+        // check pojo-like classes
+        this.memberFunctions.any { it.visibility == KVisibility.PUBLIC && it.isGetterLike() }
+
 internal class CreateDataFrameDslImpl<T>(
     override val source: Iterable<T>,
     private val clazz: KClass<*>,