fix: don't let S3A committer, delete and read defaults block native Iceberg writes

spark/src/main/scala/org/apache/comet/serde/operator/CometIcebergNativeWrite.scala

@@ -158,10 +158,19 @@ object CometIcebergNativeWrite extends CometOperatorSerde[IcebergWriteExec] {
   // from spark.hadoop.*. They do not alter an Iceberg data-file write request, so they must not
   // make every otherwise-clean S3 write ineligible. This also permits explicit overrides, which
   // are harmless on the write path for the same reason.
+  //
+  // Some Spark distributions also set S3A delete, read and committer defaults cluster-wide. The
+  // native writer does not go through S3A, `fadvise` is a read hint, and Iceberg commits through
+  // table metadata rather than a Hadoop output committer, so these are harmless here as well.
   private val IgnoredHadoopS3Keys: Set[String] = Set(
     "fs.s3a.downgrade.syncable.exceptions",
     "fs.s3a.vectored.read.max.merged.size",
-    "fs.s3a.vectored.read.min.seek.size")
+    "fs.s3a.vectored.read.min.seek.size",
+    "fs.s3a.bulk.delete.page.size",
+    "fs.s3a.experimental.input.fadvise",
+    "fs.s3a.committer.magic.enabled",
+    "fs.s3a.committer.name",
+    "fs.s3a.committer.threads")
 
   // Audited against the pinned iceberg-rust S3 parser
   // (`iceberg/src/io/storage/config/s3.rs` and `storage/opendal/src/s3.rs`). Do not broaden this

spark/src/test/scala/org/apache/comet/CometIcebergWriteDetectionSuite.scala

@@ -772,6 +772,23 @@ class CometIcebergWriteDetectionSuite extends CometTestBase with CometIcebergTes
     }
   }
 
+  test("cluster-wide S3A delete, read and committer defaults do not block a native write") {
+    val hadoopConf = new Configuration(false)
+    Seq(
+      "fs.s3a.bulk.delete.page.size" -> "1000",
+      "fs.s3a.experimental.input.fadvise" -> "random",
+      "fs.s3a.committer.magic.enabled" -> "true",
+      "fs.s3a.committer.name" -> "magic",
+      "fs.s3a.committer.threads" -> "8").foreach { case (k, v) => hadoopConf.set(k, v) }
+    assert(
+      CometIcebergNativeWrite.unsupportedHadoopS3Settings(hadoopConf, Some("target")).isEmpty)
+
+    hadoopConf.set("fs.s3a.encryption.algorithm", "SSE-KMS")
+    assert(
+      CometIcebergNativeWrite.unsupportedHadoopS3Settings(hadoopConf, Some("target")) == Seq(
+        "fs.s3a.encryption.algorithm"))
+  }
+
   test("Hadoop S3A built-in defaults are ignored but custom resources are effective") {
     def xml(key: String, value: String): java.io.ByteArrayInputStream =
       new java.io.ByteArrayInputStream(s"""<configuration>