[core] Fix incorrect sequenceNumber in manifest after row-tracking compaction

Author: juntaozhang

paimon-core/src/main/java/org/apache/paimon/append/AppendCompactTask.java

@@ -35,6 +35,7 @@
 
 import java.util.ArrayList;
 import java.util.Collections;
+import java.util.Comparator;
 import java.util.List;
 import java.util.Objects;
 
@@ -49,6 +50,7 @@ public class AppendCompactTask {
 
     public AppendCompactTask(BinaryRow partition, List<DataFileMeta> files) {
         Preconditions.checkArgument(files != null);
+        files.sort(Comparator.comparingLong(DataFileMeta::minSequenceNumber));
         this.partition = partition;
         compactBefore = new ArrayList<>(files);
         compactAfter = new ArrayList<>();

paimon-spark/paimon-spark-ut/src/test/scala/org/apache/paimon/spark/sql/RowTrackingTestBase.scala

@@ -252,6 +252,14 @@ abstract class RowTrackingTestBase extends PaimonSparkTestBase {
         sql("SELECT *, _ROW_ID, _SEQUENCE_NUMBER FROM t ORDER BY id"),
         Seq(Row(1, 1, 0, 1), Row(2, 2, 1, 2), Row(3, 3, 2, 3))
       )
+
+      sql("INSERT INTO t VALUES (4, '4')")
+      sql("INSERT INTO t VALUES (5, '5')")
+      sql("CALL sys.compact(table => 't')")
+      checkAnswer(
+        sql("SELECT min_sequence_number, max_sequence_number FROM `t$files`"),
+        Seq(Row(1, 5))
+      )
     }
   }