syun64 commented on code in PR #921:
URL: https://github.com/apache/iceberg-python/pull/921#discussion_r1676566490
##########
pyiceberg/io/pyarrow.py:
##########
@@ -2079,36 +2083,63 @@ def _check_schema_compatible(table_schema: Schema,
other_schema: pa.Schema, down
Raises:
ValueError: If the schemas are not compatible.
"""
- name_mapping = table_schema.name_mapping
- try:
- task_schema = pyarrow_to_schema(
- other_schema, name_mapping=name_mapping,
downcast_ns_timestamp_to_us=downcast_ns_timestamp_to_us
- )
- except ValueError as e:
- other_schema = _pyarrow_to_schema_without_ids(other_schema,
downcast_ns_timestamp_to_us=downcast_ns_timestamp_to_us)
- additional_names = set(other_schema.column_names) -
set(table_schema.column_names)
- raise ValueError(
- f"PyArrow table contains more columns: {',
'.join(sorted(additional_names))}. Update the schema first (hint, use
union_by_name)."
- ) from e
-
- if table_schema.as_struct() != task_schema.as_struct():
- from rich.console import Console
- from rich.table import Table as RichTable
+ task_schema = assign_fresh_schema_ids(
+ _pyarrow_to_schema_without_ids(other_schema,
downcast_ns_timestamp_to_us=downcast_ns_timestamp_to_us)
+ )
- console = Console(record=True)
+ extra_fields = task_schema.field_names - table_schema.field_names
+ missing_fields = table_schema.field_names - task_schema.field_names
+ fields_in_both =
task_schema.field_names.intersection(table_schema.field_names)
+
+ from rich.console import Console
+ from rich.table import Table as RichTable
+
+ console = Console(record=True)
+
+ rich_table = RichTable(show_header=True, header_style="bold")
+ rich_table.add_column("Field Name")
+ rich_table.add_column("Category")
+ rich_table.add_column("Table field")
+ rich_table.add_column("Dataframe field")
+
+ def print_nullability(required: bool) -> str:
+ return "required" if required else "optional"
+
+ for field_name in fields_in_both:
+ lhs = table_schema.find_field(field_name)
+ rhs = task_schema.find_field(field_name)
+ # Check nullability
+ if lhs.required != rhs.required:
+ rich_table.add_row(
+ field_name,
+ "Nullability",
+ f"{print_nullability(lhs.required)} {str(lhs.field_type)}",
+ f"{print_nullability(rhs.required)} {str(rhs.field_type)}",
+ )
+ # Check if type is consistent
+ if any(
+ (isinstance(lhs.field_type, container_type) and
isinstance(rhs.field_type, container_type))
+ for container_type in {StructType, MapType, ListType}
+ ):
+ continue
+ elif lhs.field_type != rhs.field_type:
+ rich_table.add_row(
+ field_name,
+ "Type",
+ f"{print_nullability(lhs.required)} {str(lhs.field_type)}",
+ f"{print_nullability(rhs.required)} {str(rhs.field_type)}",
+ )
Review Comment:
Hi @HonahX I think that's a great suggestion! Thank you for pointing that
out. I think it'll actually be a very simple change that addresses my question
above:
> Question: is it correct to compare both as Iceberg Schema? or do we want
to allow a more permissive range of pyarrow types to be supported for writes?
>
> For example, do we want to support writing pa.int32() into LongType()? -
maybe we could support this in a subsequent PR?
--
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: issues-unsubscr...@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org
For additional commands, e-mail: issues-h...@iceberg.apache.org
