Feature: Write to branches

pyiceberg/cli/console.py

@@ -33,7 +33,7 @@
 from pyiceberg.cli.output import ConsoleOutput, JsonOutput, Output
 from pyiceberg.exceptions import NoSuchNamespaceError, NoSuchPropertyException, NoSuchTableError
 from pyiceberg.table import TableProperties
-from pyiceberg.table.refs import SnapshotRef
+from pyiceberg.table.refs import SnapshotRef, SnapshotRefType
 from pyiceberg.utils.deprecated import deprecated
 from pyiceberg.utils.properties import property_as_int
 
@@ -444,7 +444,7 @@ def list_refs(ctx: Context, identifier: str, type: str, verbose: bool) -> None:
     refs = table.refs()
     if type:
         type = type.lower()
-        if type not in {"branch", "tag"}:
+        if type not in {SnapshotRefType.BRANCH, SnapshotRefType.TAG}:
             raise ValueError(f"Type must be either branch or tag, got: {type}")
 
     relevant_refs = [
@@ -458,7 +458,7 @@ def list_refs(ctx: Context, identifier: str, type: str, verbose: bool) -> None:
 
 def _retention_properties(ref: SnapshotRef, table_properties: Dict[str, str]) -> Dict[str, str]:
     retention_properties = {}
-    if ref.snapshot_ref_type == "branch":
+    if ref.snapshot_ref_type == SnapshotRefType.BRANCH:
         default_min_snapshots_to_keep = property_as_int(
             table_properties,
             TableProperties.MIN_SNAPSHOTS_TO_KEEP,

pyiceberg/table/__init__.py

@@ -80,7 +80,7 @@
 from pyiceberg.table.name_mapping import (
     NameMapping,
 )
-from pyiceberg.table.refs import SnapshotRef
+from pyiceberg.table.refs import MAIN_BRANCH, SnapshotRef
 from pyiceberg.table.snapshots import (
     Snapshot,
     SnapshotLogEntry,
@@ -341,6 +341,7 @@ def _set_ref_snapshot(
             AssertRefSnapshotId(
                 snapshot_id=self.table_metadata.refs[ref_name].snapshot_id if ref_name in self.table_metadata.refs else None,
                 ref=ref_name,
+                ref_type=type,
             ),
         )
 
@@ -363,21 +364,22 @@ def update_schema(self, allow_incompatible_changes: bool = False, case_sensitive
             name_mapping=self.table_metadata.name_mapping(),
         )
 
-    def update_snapshot(self, snapshot_properties: Dict[str, str] = EMPTY_DICT) -> UpdateSnapshot:
+    def update_snapshot(self, snapshot_properties: Dict[str, str] = EMPTY_DICT, branch: str = MAIN_BRANCH) -> UpdateSnapshot:
         """Create a new UpdateSnapshot to produce a new snapshot for the table.
 
         Returns:
             A new UpdateSnapshot
         """
-        return UpdateSnapshot(self, io=self._table.io, snapshot_properties=snapshot_properties)
+        return UpdateSnapshot(self, io=self._table.io, branch=branch, snapshot_properties=snapshot_properties)
 
-    def append(self, df: pa.Table, snapshot_properties: Dict[str, str] = EMPTY_DICT) -> None:
+    def append(self, df: pa.Table, snapshot_properties: Dict[str, str] = EMPTY_DICT, branch: str = MAIN_BRANCH) -> None:
         """
         Shorthand API for appending a PyArrow table to a table transaction.
 
         Args:
             df: The Arrow dataframe that will be appended to overwrite the table
             snapshot_properties: Custom properties to be added to the snapshot summary
+            branch: Branch Reference to run the append operation
         """
         try:
             import pyarrow as pa
@@ -405,7 +407,7 @@ def append(self, df: pa.Table, snapshot_properties: Dict[str, str] = EMPTY_DICT)
             TableProperties.MANIFEST_MERGE_ENABLED,
             TableProperties.MANIFEST_MERGE_ENABLED_DEFAULT,
         )
-        update_snapshot = self.update_snapshot(snapshot_properties=snapshot_properties)
+        update_snapshot = self.update_snapshot(branch=branch, snapshot_properties=snapshot_properties)
         append_method = update_snapshot.merge_append if manifest_merge_enabled else update_snapshot.fast_append
 
         with append_method() as append_files:
@@ -422,6 +424,7 @@ def overwrite(
         df: pa.Table,
         overwrite_filter: Union[BooleanExpression, str] = ALWAYS_TRUE,
         snapshot_properties: Dict[str, str] = EMPTY_DICT,
+        branch: str = MAIN_BRANCH,
     ) -> None:
         """
         Shorthand for adding a table overwrite with a PyArrow table to the transaction.
@@ -437,6 +440,7 @@ def overwrite(
             overwrite_filter: ALWAYS_TRUE when you overwrite all the data,
                               or a boolean expression in case of a partial overwrite
             snapshot_properties: Custom properties to be added to the snapshot summary
+            branch: Branch Reference to run the overwrite operation
         """
         try:
             import pyarrow as pa
@@ -459,9 +463,9 @@ def overwrite(
             self.table_metadata.schema(), provided_schema=df.schema, downcast_ns_timestamp_to_us=downcast_ns_timestamp_to_us
         )
 
-        self.delete(delete_filter=overwrite_filter, snapshot_properties=snapshot_properties)
+        self.delete(delete_filter=overwrite_filter, snapshot_properties=snapshot_properties, branch=branch)
 
-        with self.update_snapshot(snapshot_properties=snapshot_properties).fast_append() as update_snapshot:
+        with self.update_snapshot(branch=branch, snapshot_properties=snapshot_properties).fast_append() as update_snapshot:
             # skip writing data files if the dataframe is empty
             if df.shape[0] > 0:
                 data_files = _dataframe_to_data_files(
@@ -470,7 +474,12 @@ def overwrite(
                 for data_file in data_files:
                     update_snapshot.append_data_file(data_file)
 
-    def delete(self, delete_filter: Union[str, BooleanExpression], snapshot_properties: Dict[str, str] = EMPTY_DICT) -> None:
+    def delete(
+        self,
+        delete_filter: Union[str, BooleanExpression],
+        snapshot_properties: Dict[str, str] = EMPTY_DICT,
+        branch: str = MAIN_BRANCH,
+    ) -> None:
         """
         Shorthand for deleting record from a table.
 
@@ -482,6 +491,7 @@ def delete(self, delete_filter: Union[str, BooleanExpression], snapshot_properti
         Args:
             delete_filter: A boolean expression to delete rows from a table
             snapshot_properties: Custom properties to be added to the snapshot summary
+            branch: Branch Reference to run the delete operation
         """
         from pyiceberg.io.pyarrow import (
             ArrowScan,
@@ -498,15 +508,15 @@ def delete(self, delete_filter: Union[str, BooleanExpression], snapshot_properti
         if isinstance(delete_filter, str):
             delete_filter = _parse_row_filter(delete_filter)
 
-        with self.update_snapshot(snapshot_properties=snapshot_properties).delete() as delete_snapshot:
+        with self.update_snapshot(branch=branch, snapshot_properties=snapshot_properties).delete() as delete_snapshot:
             delete_snapshot.delete_by_predicate(delete_filter)
 
         # Check if there are any files that require an actual rewrite of a data file
         if delete_snapshot.rewrites_needed is True:
             bound_delete_filter = bind(self.table_metadata.schema(), delete_filter, case_sensitive=True)
             preserve_row_filter = _expression_to_complementary_pyarrow(bound_delete_filter)
 
-            files = self._scan(row_filter=delete_filter).plan_files()
+            files = self._scan(row_filter=delete_filter).use_ref(branch).plan_files()
 
             commit_uuid = uuid.uuid4()
             counter = itertools.count(0)
@@ -546,7 +556,7 @@ def delete(self, delete_filter: Union[str, BooleanExpression], snapshot_properti
                     ))
 
             if len(replaced_files) > 0:
-                with self.update_snapshot(snapshot_properties=snapshot_properties).overwrite(
+                with self.update_snapshot(branch=branch, snapshot_properties=snapshot_properties).overwrite(
                     commit_uuid=commit_uuid
                 ) as overwrite_snapshot:
                     for original_data_file, replaced_data_files in replaced_files:
@@ -972,22 +982,24 @@ def name_mapping(self) -> Optional[NameMapping]:
         """Return the table's field-id NameMapping."""
         return self.metadata.name_mapping()
 
-    def append(self, df: pa.Table, snapshot_properties: Dict[str, str] = EMPTY_DICT) -> None:
+    def append(self, df: pa.Table, snapshot_properties: Dict[str, str] = EMPTY_DICT, branch: str = MAIN_BRANCH) -> None:
         """
         Shorthand API for appending a PyArrow table to the table.
 
         Args:
             df: The Arrow dataframe that will be appended to overwrite the table
             snapshot_properties: Custom properties to be added to the snapshot summary
+            branch: Branch Reference to run the delete operation
         """
         with self.transaction() as tx:
-            tx.append(df=df, snapshot_properties=snapshot_properties)
+            tx.append(df=df, snapshot_properties=snapshot_properties, branch=branch)
 
     def overwrite(
         self,
         df: pa.Table,
         overwrite_filter: Union[BooleanExpression, str] = ALWAYS_TRUE,
         snapshot_properties: Dict[str, str] = EMPTY_DICT,
+        branch: str = MAIN_BRANCH,
     ) -> None:
         """
         Shorthand for overwriting the table with a PyArrow table.
@@ -1003,22 +1015,27 @@ def overwrite(
             overwrite_filter: ALWAYS_TRUE when you overwrite all the data,
                               or a boolean expression in case of a partial overwrite
             snapshot_properties: Custom properties to be added to the snapshot summary
+            branch: Branch Reference to run the delete operation
         """
         with self.transaction() as tx:
-            tx.overwrite(df=df, overwrite_filter=overwrite_filter, snapshot_properties=snapshot_properties)
+            tx.overwrite(df=df, overwrite_filter=overwrite_filter, snapshot_properties=snapshot_properties, branch=branch)
 
     def delete(
-        self, delete_filter: Union[BooleanExpression, str] = ALWAYS_TRUE, snapshot_properties: Dict[str, str] = EMPTY_DICT
+        self,
+        delete_filter: Union[BooleanExpression, str] = ALWAYS_TRUE,
+        snapshot_properties: Dict[str, str] = EMPTY_DICT,
+        branch: str = MAIN_BRANCH,
     ) -> None:
         """
         Shorthand for deleting rows from the table.
 
         Args:
             delete_filter: The predicate that used to remove rows
             snapshot_properties: Custom properties to be added to the snapshot summary
+            branch: Branch Reference to run the delete operation
         """
         with self.transaction() as tx:
-            tx.delete(delete_filter=delete_filter, snapshot_properties=snapshot_properties)
+            tx.delete(delete_filter=delete_filter, snapshot_properties=snapshot_properties, branch=branch)
 
     def add_files(
         self, file_paths: List[str], snapshot_properties: Dict[str, str] = EMPTY_DICT, check_duplicate_files: bool = True

pyiceberg/table/update/__init__.py

@@ -29,7 +29,7 @@
 from pyiceberg.partitioning import PARTITION_FIELD_ID_START, PartitionSpec
 from pyiceberg.schema import Schema
 from pyiceberg.table.metadata import SUPPORTED_TABLE_FORMAT_VERSION, TableMetadata, TableMetadataUtil
-from pyiceberg.table.refs import MAIN_BRANCH, SnapshotRef
+from pyiceberg.table.refs import MAIN_BRANCH, SnapshotRef, SnapshotRefType
 from pyiceberg.table.snapshots import (
     MetadataLogEntry,
     Snapshot,
@@ -154,7 +154,7 @@ class AddSnapshotUpdate(IcebergBaseModel):
 class SetSnapshotRefUpdate(IcebergBaseModel):
     action: Literal["set-snapshot-ref"] = Field(default="set-snapshot-ref")
     ref_name: str = Field(alias="ref-name")
-    type: Literal["tag", "branch"]
+    type: Literal[SnapshotRefType.TAG, SnapshotRefType.BRANCH]
     snapshot_id: int = Field(alias="snapshot-id")
     max_ref_age_ms: Annotated[Optional[int], Field(alias="max-ref-age-ms", default=None)]
     max_snapshot_age_ms: Annotated[Optional[int], Field(alias="max-snapshot-age-ms", default=None)]
@@ -609,11 +609,14 @@ class AssertRefSnapshotId(ValidatableTableRequirement):
 
     type: Literal["assert-ref-snapshot-id"] = Field(default="assert-ref-snapshot-id")
     ref: str = Field(...)
+    ref_type: SnapshotRefType = Field(...)
     snapshot_id: Optional[int] = Field(default=None, alias="snapshot-id")
 
     def validate(self, base_metadata: Optional[TableMetadata]) -> None:
         if base_metadata is None:
             raise CommitFailedException("Requirement failed: current table metadata is missing")
+        elif len(base_metadata.snapshots) == 0 and self.ref != MAIN_BRANCH:
+            raise CommitFailedException(f"Requirement failed: No snapshot available in table for ref: {self.ref}")
         elif snapshot_ref := base_metadata.refs.get(self.ref):
             ref_type = snapshot_ref.snapshot_ref_type
             if self.snapshot_id is None:
@@ -622,6 +625,8 @@ def validate(self, base_metadata: Optional[TableMetadata]) -> None:
                 raise CommitFailedException(
                     f"Requirement failed: {ref_type} {self.ref} has changed: expected id {self.snapshot_id}, found {snapshot_ref.snapshot_id}"
                 )
+            elif ref_type != self.ref_type:
+                raise CommitFailedException(f"Requirement failed: {ref_type} {self.ref} can't be changed to type {self.ref_type}")
         elif self.snapshot_id is not None:
             raise CommitFailedException(f"Requirement failed: branch or tag {self.ref} is missing, expected {self.snapshot_id}")