mbutrovich commented on code in PR #1821:
URL: https://github.com/apache/iceberg-rust/pull/1821#discussion_r2517988370
##########
crates/iceberg/src/arrow/reader.rs:
##########
@@ -919,6 +954,77 @@ fn build_fallback_field_id_map(parquet_schema:
&SchemaDescriptor) -> HashMap<i32
column_map
}
+/// Apply name mapping to Arrow schema for Parquet files lacking field IDs.
+///
+/// Assigns Iceberg field IDs based on column names using the name mapping,
+/// enabling correct projection on migrated files (e.g., from Hive/Spark via
add_files).
+///
+/// Per Iceberg spec Column Projection rule #2:
+/// "Use schema.name-mapping.default metadata to map field id to columns
without field id"
+/// https://iceberg.apache.org/spec/#column-projection
+///
+/// Corresponds to Java's ParquetSchemaUtil.applyNameMapping() and
ApplyNameMapping visitor.
+/// The key difference is Java operates on Parquet MessageType, while we
operate on Arrow Schema.
+///
+/// # Arguments
+/// * `arrow_schema` - Arrow schema from Parquet file (without field IDs)
+/// * `name_mapping` - Name mapping from table metadata
(TableProperties.DEFAULT_NAME_MAPPING)
+///
+/// # Returns
+/// Arrow schema with field IDs assigned based on name mapping
+fn apply_name_mapping_to_arrow_schema(
Review Comment:
Opened https://github.com/apache/iceberg-rust/issues/1845
--
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]
