comments

Author: jan-elastic

x-pack/plugin/esql/compute/src/main/java/org/elasticsearch/compute/operator/SampleOperator.java

@@ -41,6 +41,14 @@ public String describe() {
     }
 
     private final Deque<Page> outputPages;
+
+    /**
+     * At any time this iterator will point to be next document that still
+     * needs to be sampled. If this document is on the current page, it's
+     * added to the output and the iterator is advanced. It the document is
+     * not on the current page, the current page is finished and the index
+     * is used for the next page.
+     */
     private final RandomSamplingQuery.RandomSamplingIterator randomSamplingIterator;
     private boolean finished;
 
@@ -55,6 +63,7 @@ private SampleOperator(double probability, int seed) {
         outputPages = new LinkedList<>();
         SplittableRandom random = new SplittableRandom(seed);
         randomSamplingIterator = new RandomSamplingQuery.RandomSamplingIterator(Integer.MAX_VALUE, probability, random::nextInt);
+        // Initialize the iterator to the next document that needs to be sampled.
         randomSamplingIterator.nextDoc();
     }
 

x-pack/plugin/esql/qa/server/src/main/java/org/elasticsearch/xpack/esql/qa/rest/RestSampleTestCase.java

@@ -75,12 +75,18 @@ protected boolean matchesSafely(List<List<Integer>> lists) {
         }
     };
 
+    /**
+     * This tests sampling in the Lucene query.
+     */
     public void testSample_withFrom() throws IOException {
         createTestIndex();
         test("FROM sample-test-index | SAMPLE 0.5 | LIMIT 1000");
         deleteTestIndex();
     }
 
+    /**
+     * This tests sampling in the ES|QL operator.
+     */
     public void testSample_withRow() throws IOException {
         List<Integer> numbers = IntStream.range(0, 999).boxed().toList();
         test("ROW value = " + numbers + " | MV_EXPAND value | SAMPLE 0.5 | LIMIT 1000");

x-pack/plugin/esql/src/main/java/org/elasticsearch/xpack/esql/expression/function/aggregate/Count.java

@@ -108,6 +108,9 @@ private Count(Source source, Expression field, Expression filter, boolean isSamp
 
     private Count(StreamInput in) throws IOException {
         super(in);
+        // isSampleCorrected is only used during query optimization to mark
+        // whether this function has been processed. Hence there's no need to
+        // serialize it.
         this.isSampleCorrected = false;
     }
 

x-pack/plugin/esql/src/main/java/org/elasticsearch/xpack/esql/expression/function/aggregate/Sum.java

@@ -79,6 +79,9 @@ private Sum(Source source, Expression field, Expression filter, boolean isSample
 
     private Sum(StreamInput in) throws IOException {
         super(in);
+        // isSampleCorrected is only used during query optimization to mark
+        // whether this function has been processed. Hence there's no need to
+        // serialize it.
         this.isSampleCorrected = false;
     }