Imports decouple (#505) (#558)

HonahX · ndrluis · web-flow · commit a4a475a61471 · 2024-03-29T07:45:00.000+01:00
Backport to 0.6.1

Co-authored-by: Andre Luis Anastacio &lt;andreluisanastacio@gmail.com&gt;
diff --git a/pyiceberg/catalog/glue.py b/pyiceberg/catalog/glue.py
@@ -46,8 +46,6 @@
     PREVIOUS_METADATA_LOCATION,
     TABLE_TYPE,
     Catalog,
-    Identifier,
-    Properties,
     PropertiesUpdateSummary,
 )
 from pyiceberg.exceptions import (
@@ -67,7 +65,7 @@
 from pyiceberg.table import CommitTableRequest, CommitTableResponse, Table, update_table_metadata
 from pyiceberg.table.metadata import TableMetadata, new_table_metadata
 from pyiceberg.table.sorting import UNSORTED_SORT_ORDER, SortOrder
-from pyiceberg.typedef import EMPTY_DICT
+from pyiceberg.typedef import EMPTY_DICT, Identifier, Properties
 from pyiceberg.types import (
     BinaryType,
     BooleanType,
diff --git a/pyiceberg/catalog/hive.py b/pyiceberg/catalog/hive.py
@@ -59,8 +59,6 @@
     METADATA_LOCATION,
     TABLE_TYPE,
     Catalog,
-    Identifier,
-    Properties,
     PropertiesUpdateSummary,
 )
 from pyiceberg.exceptions import (
@@ -79,7 +77,7 @@
 from pyiceberg.table import CommitTableRequest, CommitTableResponse, Table, TableProperties, update_table_metadata
 from pyiceberg.table.metadata import new_table_metadata
 from pyiceberg.table.sorting import UNSORTED_SORT_ORDER, SortOrder
-from pyiceberg.typedef import EMPTY_DICT
+from pyiceberg.typedef import EMPTY_DICT, Identifier, Properties
 from pyiceberg.types import (
     BinaryType,
     BooleanType,
diff --git a/pyiceberg/catalog/noop.py b/pyiceberg/catalog/noop.py
@@ -28,10 +28,9 @@
 from pyiceberg.table import (
     CommitTableRequest,
     CommitTableResponse,
-    SortOrder,
     Table,
 )
-from pyiceberg.table.sorting import UNSORTED_SORT_ORDER
+from pyiceberg.table.sorting import UNSORTED_SORT_ORDER, SortOrder
 from pyiceberg.typedef import EMPTY_DICT, Identifier, Properties
 
 if TYPE_CHECKING:
diff --git a/pyiceberg/catalog/rest.py b/pyiceberg/catalog/rest.py
@@ -64,8 +64,8 @@
     CommitTableResponse,
     Table,
     TableIdentifier,
-    TableMetadata,
 )
+from pyiceberg.table.metadata import TableMetadata
 from pyiceberg.table.sorting import UNSORTED_SORT_ORDER, SortOrder, assign_fresh_sort_order_ids
 from pyiceberg.typedef import EMPTY_DICT, UTF8, IcebergBaseModel
 
diff --git a/pyiceberg/catalog/sql.py b/pyiceberg/catalog/sql.py
@@ -44,8 +44,6 @@
 from pyiceberg.catalog import (
     METADATA_LOCATION,
     Catalog,
-    Identifier,
-    Properties,
     PropertiesUpdateSummary,
 )
 from pyiceberg.exceptions import (
@@ -64,7 +62,7 @@
 from pyiceberg.table import CommitTableRequest, CommitTableResponse, Table, update_table_metadata
 from pyiceberg.table.metadata import new_table_metadata
 from pyiceberg.table.sorting import UNSORTED_SORT_ORDER, SortOrder
-from pyiceberg.typedef import EMPTY_DICT
+from pyiceberg.typedef import EMPTY_DICT, Identifier, Properties
 
 if TYPE_CHECKING:
     import pyarrow as pa
diff --git a/pyiceberg/expressions/visitors.py b/pyiceberg/expressions/visitors.py
@@ -54,7 +54,6 @@
     BoundStartsWith,
     BoundTerm,
     BoundUnaryPredicate,
-    L,
     Not,
     Or,
     UnboundPredicate,
@@ -63,7 +62,7 @@
 from pyiceberg.manifest import DataFile, ManifestFile, PartitionFieldSummary
 from pyiceberg.partitioning import PartitionSpec
 from pyiceberg.schema import Schema
-from pyiceberg.typedef import EMPTY_DICT, StructProtocol
+from pyiceberg.typedef import EMPTY_DICT, L, StructProtocol
 from pyiceberg.types import (
     DoubleType,
     FloatType,
diff --git a/pyiceberg/io/pyarrow.py b/pyiceberg/io/pyarrow.py
@@ -75,8 +75,8 @@
     AlwaysTrue,
     BooleanExpression,
     BoundTerm,
-    Literal,
 )
+from pyiceberg.expressions.literals import Literal
 from pyiceberg.expressions.visitors import (
     BoundBooleanExpressionVisitor,
     bind,
diff --git a/pyiceberg/table/__init__.py b/pyiceberg/table/__init__.py
@@ -45,16 +45,15 @@
 from sortedcontainers import SortedList
 from typing_extensions import Annotated
 
+import pyiceberg.expressions.parser as parser
+import pyiceberg.expressions.visitors as visitors
 from pyiceberg.exceptions import CommitFailedException, ResolveError, ValidationError
 from pyiceberg.expressions import (
     AlwaysTrue,
     And,
     BooleanExpression,
     EqualTo,
-    parser,
-    visitors,
 )
-from pyiceberg.expressions.visitors import _InclusiveMetricsEvaluator, inclusive_projection
 from pyiceberg.io import FileIO, load_file_io
 from pyiceberg.manifest import (
     POSITIONAL_DELETE_SCHEMA,
@@ -1293,7 +1292,9 @@ def _match_deletes_to_data_file(data_entry: ManifestEntry, positional_delete_ent
     relevant_entries = positional_delete_entries[positional_delete_entries.bisect_right(data_entry) :]
 
     if len(relevant_entries) > 0:
-        evaluator = _InclusiveMetricsEvaluator(POSITIONAL_DELETE_SCHEMA, EqualTo("file_path", data_entry.data_file.file_path))
+        evaluator = visitors._InclusiveMetricsEvaluator(
+            POSITIONAL_DELETE_SCHEMA, EqualTo("file_path", data_entry.data_file.file_path)
+        )
         return {
             positional_delete_entry.data_file
             for positional_delete_entry in relevant_entries
@@ -1317,7 +1318,7 @@ def __init__(
         super().__init__(table, row_filter, selected_fields, case_sensitive, snapshot_id, options, limit)
 
     def _build_partition_projection(self, spec_id: int) -> BooleanExpression:
-        project = inclusive_projection(self.table.schema(), self.table.specs()[spec_id])
+        project = visitors.inclusive_projection(self.table.schema(), self.table.specs()[spec_id])
         return project(self.row_filter)
 
     @cached_property
@@ -1384,7 +1385,7 @@ def plan_files(self) -> Iterable[FileScanTask]:
         # this filter depends on the partition spec used to write the manifest file
 
         partition_evaluators: Dict[int, Callable[[DataFile], bool]] = KeyDefaultDict(self._build_partition_evaluator)
-        metrics_evaluator = _InclusiveMetricsEvaluator(
+        metrics_evaluator = visitors._InclusiveMetricsEvaluator(
             self.table.schema(), self.row_filter, self.case_sensitive, self.options.get("include_empty_files") == "true"
         ).eval
 
diff --git a/tests/catalog/test_base.py b/tests/catalog/test_base.py
@@ -30,8 +30,6 @@
 
 from pyiceberg.catalog import (
     Catalog,
-    Identifier,
-    Properties,
     PropertiesUpdateSummary,
 )
 from pyiceberg.exceptions import (
@@ -56,7 +54,7 @@
 from pyiceberg.table.metadata import TableMetadata, TableMetadataV1, new_table_metadata
 from pyiceberg.table.sorting import UNSORTED_SORT_ORDER, SortOrder
 from pyiceberg.transforms import IdentityTransform
-from pyiceberg.typedef import EMPTY_DICT
+from pyiceberg.typedef import EMPTY_DICT, Identifier, Properties
 from pyiceberg.types import IntegerType, LongType, NestedField
 
 
diff --git a/tests/catalog/test_rest.py b/tests/catalog/test_rest.py
@@ -23,7 +23,7 @@
 from requests_mock import Mocker
 
 import pyiceberg
-from pyiceberg.catalog import PropertiesUpdateSummary, Table, load_catalog
+from pyiceberg.catalog import PropertiesUpdateSummary, load_catalog
 from pyiceberg.catalog.rest import AUTH_URL, RestCatalog
 from pyiceberg.exceptions import (
     NamespaceAlreadyExistsError,
@@ -35,6 +35,7 @@
 from pyiceberg.io import load_file_io
 from pyiceberg.partitioning import PartitionField, PartitionSpec
 from pyiceberg.schema import Schema
+from pyiceberg.table import Table
 from pyiceberg.table.metadata import TableMetadataV1
 from pyiceberg.table.sorting import SortField, SortOrder
 from pyiceberg.transforms import IdentityTransform, TruncateTransform
diff --git a/tests/catalog/test_sql.py b/tests/catalog/test_sql.py
@@ -24,7 +24,6 @@
 from pytest_lazyfixture import lazy_fixture
 from sqlalchemy.exc import ArgumentError, IntegrityError
 
-from pyiceberg.catalog import Identifier
 from pyiceberg.catalog.sql import SqlCatalog
 from pyiceberg.exceptions import (
     CommitFailedException,
@@ -47,6 +46,7 @@
     SortOrder,
 )
 from pyiceberg.transforms import IdentityTransform
+from pyiceberg.typedef import Identifier
 from pyiceberg.types import IntegerType
 
 
diff --git a/tests/expressions/test_parser.py b/tests/expressions/test_parser.py
@@ -17,6 +17,7 @@
 import pytest
 from pyparsing import ParseException
 
+import pyiceberg.expressions.parser as parser
 from pyiceberg.expressions import (
     AlwaysFalse,
     AlwaysTrue,
@@ -37,7 +38,6 @@
     NotStartsWith,
     Or,
     StartsWith,
-    parser,
 )
 
 
diff --git a/tests/integration/test_writes.py b/tests/integration/test_writes.py
@@ -31,10 +31,12 @@
 from pyspark.sql import SparkSession
 from pytest_mock.plugin import MockerFixture
 
-from pyiceberg.catalog import Catalog, Properties, Table, load_catalog
+from pyiceberg.catalog import Catalog, load_catalog
 from pyiceberg.catalog.sql import SqlCatalog
 from pyiceberg.exceptions import NamespaceAlreadyExistsError, NoSuchTableError
 from pyiceberg.schema import Schema
+from pyiceberg.table import Table
+from pyiceberg.typedef import Properties
 from pyiceberg.types import (
     BinaryType,
     BooleanType,
diff --git a/tests/io/test_pyarrow.py b/tests/io/test_pyarrow.py
@@ -53,8 +53,8 @@
     GreaterThan,
     Not,
     Or,
-    literal,
 )
+from pyiceberg.expressions.literals import literal
 from pyiceberg.io import InputStream, OutputStream, load_file_io
 from pyiceberg.io.pyarrow import (
     ICEBERG_SCHEMA,
diff --git a/tests/table/test_init.py b/tests/table/test_init.py
@@ -53,7 +53,6 @@
     RemovePropertiesUpdate,
     SetPropertiesUpdate,
     SetSnapshotRefUpdate,
-    SnapshotRef,
     StaticTable,
     Table,
     UpdateSchema,
@@ -64,6 +63,7 @@
     update_table_metadata,
 )
 from pyiceberg.table.metadata import INITIAL_SEQUENCE_NUMBER, TableMetadataUtil, TableMetadataV2
+from pyiceberg.table.refs import SnapshotRef
 from pyiceberg.table.snapshots import (
     Operation,
     Snapshot,
diff --git a/tests/table/test_metadata.py b/tests/table/test_metadata.py
@@ -29,15 +29,14 @@
 from pyiceberg.partitioning import PartitionField, PartitionSpec
 from pyiceberg.schema import Schema
 from pyiceberg.serializers import FromByteStream
-from pyiceberg.table import SortOrder
 from pyiceberg.table.metadata import (
     TableMetadataUtil,
     TableMetadataV1,
     TableMetadataV2,
     new_table_metadata,
 )
 from pyiceberg.table.refs import SnapshotRef, SnapshotRefType
-from pyiceberg.table.sorting import NullOrder, SortDirection, SortField
+from pyiceberg.table.sorting import NullOrder, SortDirection, SortField, SortOrder
 from pyiceberg.transforms import IdentityTransform
 from pyiceberg.typedef import UTF8
 from pyiceberg.types import (
diff --git a/tests/test_schema.py b/tests/test_schema.py
@@ -23,8 +23,8 @@
 
 from pyiceberg import schema
 from pyiceberg.exceptions import ResolveError, ValidationError
-from pyiceberg.expressions import Accessor
 from pyiceberg.schema import (
+    Accessor,
     Schema,
     build_position_accessors,
     promote,
diff --git a/tests/utils/test_manifest.py b/tests/utils/test_manifest.py
@@ -37,8 +37,7 @@
 )
 from pyiceberg.partitioning import PartitionField, PartitionSpec
 from pyiceberg.schema import Schema
-from pyiceberg.table import Snapshot
-from pyiceberg.table.snapshots import Operation, Summary
+from pyiceberg.table.snapshots import Operation, Snapshot, Summary
 from pyiceberg.transforms import IdentityTransform
 from pyiceberg.typedef import Record
 from pyiceberg.types import IntegerType, NestedField

Original file line number	Diff line number	Diff line change
`@@ -64,8 +64,8 @@`
`64`	`64`	`CommitTableResponse,`
`65`	`65`	`Table,`
`66`	`66`	`TableIdentifier,`
`67`		`- TableMetadata,`
`68`	`67`	`)`
	`68`	`+from pyiceberg.table.metadata import TableMetadata`
`69`	`69`	`from pyiceberg.table.sorting import UNSORTED_SORT_ORDER, SortOrder, assign_fresh_sort_order_ids`
`70`	`70`	`from pyiceberg.typedef import EMPTY_DICT, UTF8, IcebergBaseModel`
`71`	`71`