(arrow-rs) branch main updated: [parquet] Add some tests to cover selective list child padding (#10258)

[jeffreyvo]

This is an automated email from the ASF dual-hosted git repository.

Jefffrey pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/arrow-rs.git


The following commit(s) were added to refs/heads/main by this push:
     new 1583c8e656 [parquet] Add some tests to cover selective list child 
padding (#10258)
1583c8e656 is described below

commit 1583c8e6563dca80f1380c93089a7c1052776943
Author: Andrew Lamb <and...@nerdnetworks.org>
AuthorDate: Sat Jul 4 21:38:08 2026 -0400

    [parquet] Add some tests to cover selective list child padding (#10258)
    
    ~Draft as it includes #9848 from @HippoBaro  as well~
    - Follow on from  #9848 from @HippoBaro
    
    There were a few uncovered cases I found via code coverage, so this PR
    adds additional coverage
    
    Changes:
    - Adds an ArrowWriter round-trip test for List<Utf8View> with an item
    null, a null parent list, and a long string to exercise selective
    padding end to end.
    - Adds a short definition-level bitmap test that covers the unfiltered
    full-word append path.
---
 parquet/src/arrow/arrow_writer/mod.rs                | 17 +++++++++++++++++
 parquet/src/arrow/record_reader/definition_levels.rs | 20 ++++++++++++++++++++
 2 files changed, 37 insertions(+)

diff --git a/parquet/src/arrow/arrow_writer/mod.rs 
b/parquet/src/arrow/arrow_writer/mod.rs
index e985c183ed..95d433a461 100644
--- a/parquet/src/arrow/arrow_writer/mod.rs
+++ b/parquet/src/arrow/arrow_writer/mod.rs
@@ -3862,6 +3862,23 @@ mod tests {
         one_column_roundtrip(Arc::new(list), true);
     }
 
+    #[test]
+    fn list_utf8_view_selective_padding_roundtrip() {
+        let item = Arc::new(Field::new_list_field(DataType::Utf8View, true));
+        let mut builder = 
ListBuilder::new(StringViewBuilder::new()).with_field(item);
+        builder.values().append_value("a");
+        builder.values().append_null();
+        builder.append(true);
+        // The null parent list covers selective padding dropping values below
+        // the list definition level while preserving the preceding item null.
+        builder.append(false);
+        // The long string covers the non-inlined Utf8View buffer path.
+        builder.values().append_value("large payload over 12 bytes");
+        builder.append(true);
+
+        one_column_roundtrip(Arc::new(builder.finish()), true);
+    }
+
     #[test]
     fn struct_single_column() {
         let a_values = Int32Array::from(vec![1, 2, 3, 4, 5, 6, 7, 8, 9, 10]);
diff --git a/parquet/src/arrow/record_reader/definition_levels.rs 
b/parquet/src/arrow/record_reader/definition_levels.rs
index 0720c6cdbe..a104c93a2c 100644
--- a/parquet/src/arrow/record_reader/definition_levels.rs
+++ b/parquet/src/arrow/record_reader/definition_levels.rs
@@ -504,6 +504,26 @@ mod tests {
     use crate::encodings::rle::RleEncoder;
     use rand::{Rng, rng};
 
+    #[test]
+    fn test_build_validity_bitmap_unfiltered_word_chunk() {
+        // 65 levels forces the unfiltered path to process one full u64 word
+        // with append_word, plus a remainder bit.
+        let def_levels = (0..65)
+            .map(|i| if i % 3 == 0 { 2 } else { 1 })
+            .collect::<Vec<_>>();
+        let mut bitmap = BooleanBufferBuilder::new(0);
+
+        assert_eq!(
+            build_filtered_validity_bitmap(&def_levels, None, None, 2, &mut 
bitmap),
+            def_levels.len()
+        );
+
+        let bitmap = bitmap.finish();
+        for (idx, def) in def_levels.iter().enumerate() {
+            assert_eq!(bitmap.value(idx), *def >= 2);
+        }
+    }
+
     #[test]
     fn test_packed_decoder() {
         let mut rng = rng();