This is an automated email from the ASF dual-hosted git repository.
kevinjqliu pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/iceberg-python.git
The following commit(s) were added to refs/heads/main by this push:
new b63278bc Add RemovePartitionSpecsUpdate event (#2198)
b63278bc is described below
commit b63278bcae13f959a3c7d1dbb7b6caa3c7edf7c8
Author: Alex Stephen <[email protected]>
AuthorDate: Thu Sep 25 20:10:02 2025 -0700
Add RemovePartitionSpecsUpdate event (#2198)
<!--
Thanks for opening a pull request!
-->
<!-- In the case this PR will resolve an issue, please replace
${GITHUB_ISSUE_ID} below with the actual Github issue id. -->
<!-- Closes #${GITHUB_ISSUE_ID} -->
# Rationale for this change
This adds RemovePartitionSpecsUpdate, another missing TableMetadata
update event.
It also ends up testing AddPartitionSpecUpdate, which it looks like it
untested.
[REST API Spec
Reference](https://github.com/apache/iceberg/blob/09140e52836048b112c42c9cfe721295bd57048b/open-api/rest-catalog-open-api.yaml#L3006-L3018)
# Are these changes tested?
Unit tests added.
# Are there any user-facing changes?
- Added RemovePartitionSpecsUpdate event
<!-- In the case of user-facing changes, please add the changelog label.
-->
---
pyiceberg/table/update/__init__.py | 21 +++++++++++++++++++++
tests/table/test_init.py | 34 ++++++++++++++++++++++++++++++++++
2 files changed, 55 insertions(+)
diff --git a/pyiceberg/table/update/__init__.py
b/pyiceberg/table/update/__init__.py
index 30315b0c..038b952b 100644
--- a/pyiceberg/table/update/__init__.py
+++ b/pyiceberg/table/update/__init__.py
@@ -193,6 +193,11 @@ class RemoveStatisticsUpdate(IcebergBaseModel):
snapshot_id: int = Field(alias="snapshot-id")
+class RemovePartitionSpecsUpdate(IcebergBaseModel):
+ action: Literal["remove-partition-specs"] =
Field(default="remove-partition-specs")
+ spec_ids: List[int] = Field(alias="spec-ids")
+
+
class RemoveSchemasUpdate(IcebergBaseModel):
action: Literal["remove-schemas"] = Field(default="remove-schemas")
schema_ids: List[int] = Field(alias="schema-ids")
@@ -227,6 +232,7 @@ TableUpdate = Annotated[
RemovePropertiesUpdate,
SetStatisticsUpdate,
RemoveStatisticsUpdate,
+ RemovePartitionSpecsUpdate,
RemoveSchemasUpdate,
SetPartitionStatisticsUpdate,
RemovePartitionStatisticsUpdate,
@@ -595,6 +601,21 @@ def _(update: RemoveStatisticsUpdate, base_metadata:
TableMetadata, context: _Ta
return base_metadata.model_copy(update={"statistics": statistics})
+@_apply_table_update.register(RemovePartitionSpecsUpdate)
+def _(update: RemovePartitionSpecsUpdate, base_metadata: TableMetadata,
context: _TableMetadataUpdateContext) -> TableMetadata:
+ for remove_spec_id in update.spec_ids:
+ if not any(spec.spec_id == remove_spec_id for spec in
base_metadata.partition_specs):
+ raise ValueError(f"Partition spec with id {remove_spec_id} does
not exist")
+
+ if base_metadata.default_spec_id in update.spec_ids:
+ raise ValueError(f"Cannot remove default partition spec:
{base_metadata.default_spec_id}")
+
+ partition_specs = [spec for spec in base_metadata.partition_specs if
spec.spec_id not in update.spec_ids]
+
+ context.add_update(update)
+ return base_metadata.model_copy(update={"partition_specs":
partition_specs})
+
+
@_apply_table_update.register(RemoveSchemasUpdate)
def _(update: RemoveSchemasUpdate, base_metadata: TableMetadata, context:
_TableMetadataUpdateContext) -> TableMetadata:
# This method should error if any schemas do not exist.
diff --git a/tests/table/test_init.py b/tests/table/test_init.py
index cd81df4d..5f64738d 100644
--- a/tests/table/test_init.py
+++ b/tests/table/test_init.py
@@ -66,6 +66,7 @@ from pyiceberg.table.sorting import (
)
from pyiceberg.table.statistics import BlobMetadata, PartitionStatisticsFile,
StatisticsFile
from pyiceberg.table.update import (
+ AddPartitionSpecUpdate,
AddSnapshotUpdate,
AddSortOrderUpdate,
AssertCreate,
@@ -76,6 +77,7 @@ from pyiceberg.table.update import (
AssertLastAssignedPartitionId,
AssertRefSnapshotId,
AssertTableUUID,
+ RemovePartitionSpecsUpdate,
RemovePartitionStatisticsUpdate,
RemovePropertiesUpdate,
RemoveSchemasUpdate,
@@ -1294,6 +1296,38 @@ def test_update_metadata_log_overflow(table_v2: Table)
-> None:
assert len(new_metadata.metadata_log) == 1
+def test_remove_partition_spec_update(table_v2: Table) -> None:
+ base_metadata = table_v2.metadata
+ new_spec = PartitionSpec(PartitionField(source_id=2, field_id=1001,
transform=IdentityTransform(), name="y"), spec_id=1)
+ metadata_with_new_spec = update_table_metadata(base_metadata,
(AddPartitionSpecUpdate(spec=new_spec),))
+
+ assert len(metadata_with_new_spec.partition_specs) == 2
+
+ update = RemovePartitionSpecsUpdate(spec_ids=[1])
+ updated_metadata = update_table_metadata(
+ metadata_with_new_spec,
+ (update,),
+ )
+
+ assert len(updated_metadata.partition_specs) == 1
+
+
+def test_remove_partition_spec_update_spec_does_not_exist(table_v2: Table) ->
None:
+ update = RemovePartitionSpecsUpdate(
+ spec_ids=[123],
+ )
+ with pytest.raises(ValueError, match="Partition spec with id 123 does not
exist"):
+ update_table_metadata(table_v2.metadata, (update,))
+
+
+def test_remove_partition_spec_update_default_spec(table_v2: Table) -> None:
+ update = RemovePartitionSpecsUpdate(
+ spec_ids=[0],
+ )
+ with pytest.raises(ValueError, match="Cannot remove default partition
spec: 0"):
+ update_table_metadata(table_v2.metadata, (update,))
+
+
def test_remove_schemas_update(table_v2: Table) -> None:
base_metadata = table_v2.metadata
assert len(base_metadata.schemas) == 2
