(arrow-go) branch main updated: fix(array): clear dictionary data on reset (#876)

[zeroshade]

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

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


The following commit(s) were added to refs/heads/main by this push:
     new 26ba9465 fix(array): clear dictionary data on reset (#876)
26ba9465 is described below

commit 26ba9465b6753ea737a77c93897312a0cb7179df
Author: Minh Vu <vuhoangmin...@gmail.com>
AuthorDate: Wed Jul 1 18:58:33 2026 +0200

    fix(array): clear dictionary data on reset (#876)
    
    ### What changed
    
    Data.Reset now releases and clears any existing dictionary data when
    reusing an ArrayData value.
    
    ### Why
    
    Reset replaces the buffers, children, type, length, null count, and
    offset, but previously left the old dictionary attached. Reusing
    dictionary-backed data as non-dictionary data could keep returning stale
    dictionary data and retain the old dictionary longer than expected.
    
    ### Validation
    
    - go test ./arrow/array -run 'TestDataReset' -count=1
    - go test ./arrow/array -count=1
    -
    PARQUET_TEST_DATA=/Users/hoangvu/Code/OSS/arrow-go/parquet-testing/data
    ARROW_TEST_DATA=/Users/hoangvu/Code/OSS/arrow-go/arrow-testing/data go
    test ./... -count=1
---
 arrow/array/data.go      |  5 +++++
 arrow/array/data_test.go | 26 ++++++++++++++++++++++++++
 2 files changed, 31 insertions(+)

diff --git a/arrow/array/data.go b/arrow/array/data.go
index eaed4717..1772acb4 100644
--- a/arrow/array/data.go
+++ b/arrow/array/data.go
@@ -122,6 +122,11 @@ func (d *Data) Reset(dtype arrow.DataType, length int, 
buffers []*memory.Buffer,
        }
        d.childData = childData
 
+       if d.dictionary != nil {
+               d.dictionary.Release()
+               d.dictionary = nil
+       }
+
        d.dtype = dtype
        d.length = length
        d.nulls = nulls
diff --git a/arrow/array/data_test.go b/arrow/array/data_test.go
index 435d8b69..7b690100 100644
--- a/arrow/array/data_test.go
+++ b/arrow/array/data_test.go
@@ -74,6 +74,32 @@ func TestNewSliceDataInvalidBounds(t *testing.T) {
        }
 }
 
+func TestDataResetClearsDictionary(t *testing.T) {
+       mem := memory.NewCheckedAllocator(memory.NewGoAllocator())
+
+       values := memory.NewResizableBuffer(mem)
+       values.Resize(16)
+
+       dictData := NewData(arrow.PrimitiveTypes.Int32, 4, 
[]*memory.Buffer{nil, values}, nil, 0, 0)
+       values.Release()
+
+       dictType := &arrow.DictionaryType{
+               IndexType: arrow.PrimitiveTypes.Int8,
+               ValueType: arrow.PrimitiveTypes.Int32,
+       }
+       data := NewDataWithDictionary(dictType, 0, []*memory.Buffer{nil, nil}, 
0, 0, dictData)
+       defer data.Release()
+
+       assert.NotNil(t, data.Dictionary())
+
+       data.Reset(arrow.PrimitiveTypes.Int8, 0, []*memory.Buffer{nil, nil}, 
nil, 0, 0)
+
+       assert.Nil(t, data.Dictionary())
+
+       dictData.Release()
+       mem.AssertSize(t, 0)
+}
+
 func TestSizeInBytes(t *testing.T) {
        var buffers1 = make([]*memory.Buffer, 0, 3)