HonahX commented on code in PR #8622:
URL: https://github.com/apache/iceberg/pull/8622#discussion_r1338188490


##########
python/pyiceberg/manifest.py:
##########
@@ -265,6 +306,64 @@ def __init__(self, *data: Any, **named_data: Any) -> None:
         super().__init__(*data, **{"struct": PARTITION_FIELD_SUMMARY_TYPE, 
**named_data})
 
 
+class PartitionFieldStats:
+    _type: PrimitiveType
+    _contains_null: bool
+    _contains_nan: bool
+    _min: Optional[Any]
+    _max: Optional[Any]
+
+    def __init__(self, iceberg_type: IcebergType) -> None:
+        if not isinstance(iceberg_type, PrimitiveType):
+            raise ValueError(f"Expected a primitive type for the partition 
field, got {iceberg_type}")
+        self._type = iceberg_type
+        self._contains_null = False
+        self._contains_nan = False
+        self._min = None
+        self._max = None
+
+    def to_summary(self) -> PartitionFieldSummary:
+        return PartitionFieldSummary(
+            contains_null=self._contains_null,
+            contains_nan=self._contains_nan,
+            lower_bound=to_bytes(self._type, self._min) if self._min is not 
None else None,
+            upper_bound=to_bytes(self._type, self._max) if self._max is not 
None else None,
+        )
+
+    def update(self, value: Any) -> None:
+        if value is None:
+            self._contains_null = True
+        elif math.isnan(value):
+            self._contains_nan = True
+        else:
+            if self._min is None:
+                self._min = value
+                self._max = value
+            else:
+                self._max = max(self._max, value)
+                self._min = min(self._min, value)
+
+
+class PartitionSummary:
+    _field_stats: List[PartitionFieldStats]
+    _types: List[IcebergType]
+
+    def __init__(self, spec: PartitionSpec, schema: Schema):
+        self._types = [field.field_type for field in 
spec.partition_type(schema).fields]
+        self._field_stats = [PartitionFieldStats(field_type) for field_type in 
self._types]
+
+    def summaries(self) -> List[PartitionFieldSummary]:
+        return [field.to_summary() for field in self._field_stats]
+
+    def update(self, partition_keys: Record) -> PartitionSummary:
+        for i, field_type in enumerate(self._types):
+            if not isinstance(field_type, PrimitiveType):
+                raise ValueError(f"Expected a primitive type for the partition 
field, got {field_type}")
+            partition_key = partition_keys[i]
+            
self._field_stats[i].update(conversions.partition_to_py(field_type, 
partition_key))

Review Comment:
   Thanks for the explanation. Confirmed that the values in 
`data_file.partition` should be ready to compare out-of-box. We do not need 
`partition_to_py`. 



-- 
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

Reply via email to