gbrgr commented on code in PR #1824:
URL: https://github.com/apache/iceberg-rust/pull/1824#discussion_r2601424055
##########
crates/iceberg/src/arrow/record_batch_transformer.rs:
##########
@@ -539,86 +619,224 @@ impl RecordBatchTransformer {
prim_lit: &Option<PrimitiveLiteral>,
num_rows: usize,
) -> Result<ArrayRef> {
- Ok(match (target_type, prim_lit) {
- (DataType::Boolean, Some(PrimitiveLiteral::Boolean(value))) => {
- Arc::new(BooleanArray::from(vec![*value; num_rows]))
- }
- (DataType::Boolean, None) => {
- let vals: Vec<Option<bool>> = vec![None; num_rows];
- Arc::new(BooleanArray::from(vals))
- }
- (DataType::Int32, Some(PrimitiveLiteral::Int(value))) => {
- Arc::new(Int32Array::from(vec![*value; num_rows]))
- }
- (DataType::Int32, None) => {
- let vals: Vec<Option<i32>> = vec![None; num_rows];
- Arc::new(Int32Array::from(vals))
- }
- (DataType::Date32, Some(PrimitiveLiteral::Int(value))) => {
- Arc::new(Date32Array::from(vec![*value; num_rows]))
- }
- (DataType::Date32, None) => {
- let vals: Vec<Option<i32>> = vec![None; num_rows];
- Arc::new(Date32Array::from(vals))
- }
- (DataType::Int64, Some(PrimitiveLiteral::Long(value))) => {
- Arc::new(Int64Array::from(vec![*value; num_rows]))
- }
- (DataType::Int64, None) => {
- let vals: Vec<Option<i64>> = vec![None; num_rows];
- Arc::new(Int64Array::from(vals))
- }
- (DataType::Float32, Some(PrimitiveLiteral::Float(value))) => {
- Arc::new(Float32Array::from(vec![value.0; num_rows]))
- }
- (DataType::Float32, None) => {
- let vals: Vec<Option<f32>> = vec![None; num_rows];
- Arc::new(Float32Array::from(vals))
- }
- (DataType::Float64, Some(PrimitiveLiteral::Double(value))) => {
- Arc::new(Float64Array::from(vec![value.0; num_rows]))
- }
- (DataType::Float64, None) => {
- let vals: Vec<Option<f64>> = vec![None; num_rows];
- Arc::new(Float64Array::from(vals))
- }
- (DataType::Utf8, Some(PrimitiveLiteral::String(value))) => {
- Arc::new(StringArray::from(vec![value.clone(); num_rows]))
- }
- (DataType::Utf8, None) => {
- let vals: Vec<Option<String>> = vec![None; num_rows];
- Arc::new(StringArray::from(vals))
- }
- (DataType::Binary, Some(PrimitiveLiteral::Binary(value))) => {
- Arc::new(BinaryArray::from_vec(vec![value; num_rows]))
- }
- (DataType::Binary, None) => {
- let vals: Vec<Option<&[u8]>> = vec![None; num_rows];
- Arc::new(BinaryArray::from_opt_vec(vals))
- }
- (DataType::Struct(fields), None) => {
- // Create a StructArray filled with nulls. Per Iceberg spec,
optional struct fields
- // default to null when added to the schema. We defer non-null
default struct values
- // and leave them as not implemented yet.
- let null_arrays: Vec<ArrayRef> = fields
- .iter()
- .map(|field| Self::create_column(field.data_type(), &None,
num_rows))
- .collect::<Result<Vec<_>>>()?;
-
- Arc::new(StructArray::new(
- fields.clone(),
- null_arrays,
- Some(NullBuffer::new_null(num_rows)),
+ // Check if this is a RunEndEncoded type (for constant fields)
+ if let DataType::RunEndEncoded(_, values_field) = target_type {
+ // Helper to create a Run-End Encoded array
+ let create_ree_array = |values_array: ArrayRef| ->
Result<ArrayRef> {
+ let run_ends = if num_rows == 0 {
+ Int32Array::from(Vec::<i32>::new())
+ } else {
+ Int32Array::from(vec![num_rows as i32])
+ };
+ Ok(Arc::new(
+ RunArray::try_new(&run_ends, &values_array).map_err(|e| {
+ Error::new(
+ ErrorKind::Unexpected,
+ "Failed to create RunArray for constant value",
+ )
+ .with_source(e)
+ })?,
))
- }
- (DataType::Null, _) => Arc::new(NullArray::new(num_rows)),
- (dt, _) => {
- return Err(Error::new(
- ErrorKind::Unexpected,
- format!("unexpected target column type {}", dt),
- ));
- }
- })
+ };
+
+ // Create the values array based on the literal value
+ let values_array: ArrayRef = match (values_field.data_type(),
prim_lit) {
+ (DataType::Boolean, Some(PrimitiveLiteral::Boolean(v))) => {
Review Comment:
Created two helpers in value.rs
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
