Merge pull request #650 from Kotlin/pojo-toDataFrame

POJO toDataFrame support (and array improvements)

Author: Jolanrensen

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

@@ -13,6 +13,7 @@ import org.jetbrains.kotlinx.dataframe.impl.asList
 import org.jetbrains.kotlinx.dataframe.impl.columnName
 import org.jetbrains.kotlinx.dataframe.impl.columns.guessColumnType
 import org.jetbrains.kotlinx.dataframe.index
+import kotlin.reflect.KCallable
 import kotlin.reflect.KClass
 import kotlin.reflect.KProperty
 
@@ -115,24 +116,26 @@ public fun Iterable<Pair<String, Iterable<Any?>>>.toDataFrameFromPairs(): AnyFra
 public interface TraversePropertiesDsl {
 
     /**
-     * Skip given [classes] during recursive (dfs) traversal
+     * Skip given [classes] during recursive (dfs) traversal.
      */
     public fun exclude(vararg classes: KClass<*>)
 
     /**
-     * Skip given [properties] during recursive (dfs) traversal
+     * Skip given [properties] during recursive (dfs) traversal.
+     * These can also be getter-like functions (like `getX()` or `isX()`).
      */
-    public fun exclude(vararg properties: KProperty<*>)
+    public fun exclude(vararg properties: KCallable<*>)
 
     /**
-     * Store given [classes] in ValueColumns without transformation into ColumnGroups or FrameColumns
+     * Store given [classes] in ValueColumns without transformation into ColumnGroups or FrameColumns.
      */
     public fun preserve(vararg classes: KClass<*>)
 
     /**
-     * Store given [properties] in ValueColumns without transformation into ColumnGroups or FrameColumns
+     * Store given [properties] in ValueColumns without transformation into ColumnGroups or FrameColumns.
+     * These can also be getter-like functions (like `getX()` or `isX()`).
      */
-    public fun preserve(vararg properties: KProperty<*>)
+    public fun preserve(vararg properties: KCallable<*>)
 }
 
 public inline fun <reified T> TraversePropertiesDsl.preserve(): Unit = preserve(T::class)
@@ -148,7 +151,7 @@ public abstract class CreateDataFrameDsl<T> : TraversePropertiesDsl {
     public infix fun AnyBaseCol.into(path: ColumnPath): Unit = add(this, path)
 
     public abstract fun properties(
-        vararg roots: KProperty<*>,
+        vararg roots: KCallable<*>,
         maxDepth: Int = 0,
         body: (TraversePropertiesDsl.() -> Unit)? = null,
     )

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

@@ -4,7 +4,7 @@ import org.jetbrains.kotlinx.dataframe.DataFrame
 import org.jetbrains.kotlinx.dataframe.DataRow
 import org.jetbrains.kotlinx.dataframe.annotations.ColumnName
 import org.jetbrains.kotlinx.dataframe.annotations.DataSchema
-import org.jetbrains.kotlinx.dataframe.impl.schema.getPropertiesOrder
+import org.jetbrains.kotlinx.dataframe.impl.schema.getPropertyOrderFromPrimaryConstructor
 import org.jetbrains.kotlinx.dataframe.schema.ColumnSchema
 import kotlin.reflect.KClass
 import kotlin.reflect.KType
@@ -53,7 +53,7 @@ internal object MarkersExtractor {
         }
 
     private fun getFields(markerClass: KClass<*>, nullableProperties: Boolean): List<GeneratedField> {
-        val order = getPropertiesOrder(markerClass)
+        val order = getPropertyOrderFromPrimaryConstructor(markerClass) ?: emptyMap()
         return markerClass.memberProperties.sortedBy { order[it.name] ?: Int.MAX_VALUE }.mapIndexed { _, it ->
             val fieldName = ValidFieldName.of(it.name)
             val columnName = it.findAnnotation<ColumnName>()?.name ?: fieldName.unquoted

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

@@ -16,7 +16,9 @@ import java.net.URL
 import java.time.LocalDateTime
 import java.time.LocalTime
 import kotlin.reflect.KType
+import kotlin.reflect.full.isSubtypeOf
 import kotlin.reflect.jvm.jvmErasure
+import kotlin.reflect.typeOf
 
 internal fun String.truncate(limit: Int): RenderedContent = if (limit in 1 until length) {
     if (limit < 4) RenderedContent.truncatedText("...", this)
@@ -57,6 +59,11 @@ internal fun renderType(type: KType?): String {
         else -> {
             val fullName = type.jvmErasure.qualifiedName ?: return type.toString()
             val name = when {
+                // catching cases like `typeOf<Array<Int>>().jvmErasure.qualifiedName == "IntArray"`
+                // https://github.com/Kotlin/dataframe/issues/678
+                type.isSubtypeOf(typeOf<Array<*>>()) ->
+                    "Array"
+
                 type.classifier == URL::class ->
                     fullName.removePrefix("java.net.")
 

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

@@ -1,3 +1,5 @@
+@file:OptIn(ExperimentalUnsignedTypes::class)
+
 package org.jetbrains.kotlinx.dataframe.impl
 
 import org.jetbrains.kotlinx.dataframe.AnyFrame
@@ -467,3 +469,100 @@ internal fun nothingType(nullable: Boolean): KType =
     } else {
         typeOf<List<Nothing>>()
     }.arguments.first().type!!
+
+@OptIn(ExperimentalUnsignedTypes::class)
+private val primitiveArrayClasses = setOf(
+    BooleanArray::class,
+    ByteArray::class,
+    ShortArray::class,
+    IntArray::class,
+    LongArray::class,
+    FloatArray::class,
+    DoubleArray::class,
+    CharArray::class,
+
+    UByteArray::class,
+    UShortArray::class,
+    UIntArray::class,
+    ULongArray::class,
+)
+
+/**
+ * Returns `true` if this class is a primitive array class like `XArray`.
+ *
+ * Use [KClass.isArray] to also check for `Array<>`.
+ */
+internal val KClass<*>.isPrimitiveArray: Boolean
+    get() = this in primitiveArrayClasses
+
+/**
+ * Returns `true` if this class is an array, either a primitive array like `XArray` or `Array<>`.
+ *
+ * Use [KClass.isPrimitiveArray] to only check for primitive arrays.
+ */
+internal val KClass<*>.isArray: Boolean
+    get() = this in primitiveArrayClasses ||
+        this.qualifiedName == Array::class.qualifiedName // instance check fails
+
+/**
+ * Returns `true` if this type is of a primitive array like `XArray`.
+ *
+ * Use [KType.isArray] to also check for `Array<>`.
+ */
+internal val KType.isPrimitiveArray: Boolean
+    get() =
+        if (arguments.isNotEmpty()) {
+            // Catching https://github.com/Kotlin/dataframe/issues/678
+            // as typeOf<Array<Int>>().classifier == IntArray::class
+            false
+        } else {
+            (classifier as? KClass<*>)?.isPrimitiveArray == true
+        }
+
+/**
+ * Returns `true` if this type is of an array, either a primitive array like `XArray` or `Array<>`.
+ *
+ * Use [KType.isPrimitiveArray] to only check for primitive arrays.
+ */
+internal val KType.isArray: Boolean
+    get() = (classifier as? KClass<*>)?.isArray == true
+
+/**
+ * Returns `true` if this object is a primitive array like `XArray`.
+ *
+ * Use [Any.isArray] to also check for the `Array<>` object.
+ */
+internal val Any.isPrimitiveArray: Boolean
+    get() = this::class.isPrimitiveArray
+
+/**
+ * Returns `true` if this object is an array, either a primitive array like `XArray` or `Array<>`.
+ *
+ * Use [Any.isPrimitiveArray] to only check for primitive arrays.
+ */
+internal val Any.isArray: Boolean
+    get() = this::class.isArray
+
+/**
+ * If [this] is an array of any kind, the function returns it as a list of values,
+ * else it returns `null`.
+ */
+internal fun Any.asArrayAsListOrNull(): List<*>? =
+    when (this) {
+        is BooleanArray -> asList()
+        is ByteArray -> asList()
+        is ShortArray -> asList()
+        is IntArray -> asList()
+        is LongArray -> asList()
+        is FloatArray -> asList()
+        is DoubleArray -> asList()
+        is CharArray -> asList()
+
+        is UByteArray -> asList()
+        is UShortArray -> asList()
+        is UIntArray -> asList()
+        is ULongArray -> asList()
+
+        is Array<*> -> asList()
+        else -> null
+    }

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

@@ -14,7 +14,9 @@ import org.jetbrains.kotlinx.dataframe.impl.columns.toColumnSet
 import org.jetbrains.kotlinx.dataframe.nrow
 import java.math.BigDecimal
 import java.math.BigInteger
+import kotlin.reflect.KCallable
 import kotlin.reflect.KClass
+import kotlin.reflect.KFunction
 import kotlin.reflect.KProperty
 import kotlin.reflect.KType
 import kotlin.reflect.KTypeProjection
@@ -23,6 +25,7 @@ import kotlin.reflect.full.createType
 import kotlin.reflect.full.findAnnotation
 import kotlin.reflect.full.isSubtypeOf
 import kotlin.reflect.full.starProjectedType
+import kotlin.reflect.full.valueParameters
 import kotlin.reflect.full.withNullability
 import kotlin.reflect.jvm.jvmErasure
 import kotlin.reflect.typeOf
@@ -337,5 +340,61 @@ internal fun List<String>.joinToCamelCaseString(): String {
         .replaceFirstChar { it.lowercaseChar() }
 }
 
+/** Returns `true` if this callable is a getter-like function.
+ *
+ * A callable is considered getter-like if it is either a property getter,
+ * or it's a function with no (type) parameters that starts with "get"/"is". */
+internal fun KFunction<*>.isGetterLike(): Boolean =
+    (name.startsWith("get") || name.startsWith("is")) &&
+        valueParameters.isEmpty() &&
+        typeParameters.isEmpty()
+
+/** Returns `true` if this callable is a getter-like function.
+ *
+ * A callable is considered getter-like if it is either a property getter,
+ * or it's a function with no (type) parameters that starts with "get"/"is". */
+internal fun KProperty<*>.isGetterLike(): Boolean = true
+
+/**
+ * Returns `true` if this callable is a getter-like function.
+ *
+ * A callable is considered getter-like if it is either a property getter,
+ * or it's a function with no (type) parameters that starts with "get"/"is".
+ */
+internal fun KCallable<*>.isGetterLike(): Boolean =
+    when (this) {
+        is KProperty<*> -> isGetterLike()
+        is KFunction<*> -> isGetterLike()
+        else -> false
+    }
+
+/** Returns the column name for this callable.
+ * If the callable contains the [ColumnName][org.jetbrains.kotlinx.dataframe.annotations.ColumnName] annotation, its [ColumnName.name][org.jetbrains.kotlinx.dataframe.annotations.ColumnName.name] is returned.
+ * Otherwise, the name of the callable is returned with proper getter-trimming if it's a [KFunction]. */
+@PublishedApi
+internal val KFunction<*>.columnName: String
+    get() = findAnnotation<ColumnName>()?.name
+        ?: name
+            .removePrefix("get")
+            .removePrefix("is")
+            .replaceFirstChar { it.lowercase() }
+
+/** Returns the column name for this callable.
+ * If the callable contains the [ColumnName][org.jetbrains.kotlinx.dataframe.annotations.ColumnName] annotation, its [ColumnName.name][org.jetbrains.kotlinx.dataframe.annotations.ColumnName.name] is returned.
+ * Otherwise, the name of the callable is returned with proper getter-trimming if it's a [KFunction]. */
 @PublishedApi
-internal val <T> KProperty<T>.columnName: String get() = findAnnotation<ColumnName>()?.name ?: name
+internal val KProperty<*>.columnName: String
+    get() = findAnnotation<ColumnName>()?.name ?: name
+
+/**
+ * Returns the column name for this callable.
+ * If the callable contains the [ColumnName] annotation, its [ColumnName.name] is returned.
+ * Otherwise, the name of the callable is returned with proper getter-trimming if it's a [KFunction].
+ */
+@PublishedApi
+internal val KCallable<*>.columnName: String
+    get() = when (this) {
+        is KFunction<*> -> columnName
+        is KProperty<*> -> columnName
+        else -> findAnnotation<ColumnName>()?.name ?: name
+    }