Optimize 2025 day 8 using PriorityQueue

Author: sim642

src/main/scala/eu/sim642/adventofcode2025/Day8.scala

@@ -4,6 +4,8 @@ import eu.sim642.adventofcodelib.IteratorImplicits.*
 import eu.sim642.adventofcodelib.graph.Kruskal
 import eu.sim642.adventofcodelib.pos.Pos3
 
+import scala.collection.mutable
+
 object Day8 {
 
   extension (pos: Pos3) {
@@ -13,20 +15,28 @@ object Day8 {
     }
   }
 
-  def closestPairsSeq(junctionBoxes: Seq[Pos3]): Seq[(Pos3, Pos3)] = {
-    //noinspection ConvertibleToMethodValue
-    (for {
+  extension [A](queue: mutable.PriorityQueue[A]) {
+    // normal queue.iterator does not yield dequeue order
+    def dequeueIterator: Iterator[A] = new Iterator[A] {
+      override def hasNext: Boolean = queue.nonEmpty
+
+      override def next(): A = queue.dequeue()
+    }
+  }
+
+  def iterateClosestPairs(junctionBoxes: Seq[Pos3]): Iterator[(Pos3, Pos3)] = {
+    // it is faster to use a PriorityQueue than sort the Seq of all pairs because Kruskal will only need some closest pairs, not all
+    val queue = mutable.PriorityQueue.empty[((Pos3, Pos3), Long)](using Ordering.by(-_._2))
+    for {
       // faster than combinations(2)
       (p1, i) <- junctionBoxes.iterator.zipWithIndex
       p2 <- junctionBoxes.view.slice(i + 1, junctionBoxes.size).iterator
-    } yield (p1, p2) -> (p1 euclideanDistanceSqr p2)) // no need to sqrt distance just for sorting
-      .toSeq
-      .sortBy(_._2)
-      .map(_._1)
+    } queue.enqueue((p1, p2) -> (p1 euclideanDistanceSqr p2)) // no need to sqrt distance just for sorting
+    queue.dequeueIterator.map(_._1)
   }
 
   def multiplySizesAfter(junctionBoxes: Seq[Pos3], after: Int = 1000, sizes: Int = 3): Int = {
-    val closestPairs = closestPairsSeq(junctionBoxes)
+    val closestPairs = iterateClosestPairs(junctionBoxes)
     val (ufAfter, _) = Kruskal.iterate(junctionBoxes, closestPairs)(after)
 
     ufAfter.groups()
@@ -38,7 +48,7 @@ object Day8 {
 
   // TODO: deduplicate
   def multiplyLastXs(junctionBoxes: Seq[Pos3]): Int = {
-    val closestPairs = closestPairsSeq(junctionBoxes)
+    val closestPairs = iterateClosestPairs(junctionBoxes)
     val lastPair = Kruskal.iterateEdges(junctionBoxes, closestPairs).last
     lastPair._1.x * lastPair._2.x
   }