pixeltable 0.3.9__tar.gz → 0.3.10__tar.gz

{pixeltable-0.3.9 → pixeltable-0.3.10}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: pixeltable
-Version: 0.3.9
+Version: 0.3.10
 Summary: AI Data Infrastructure: Declarative, Multimodal, and Incremental
 License: Apache-2.0
 Keywords: data-science,machine-learning,database,ai,computer-vision,chatbot,ml,artificial-intelligence,feature-engineering,multimodal,mlops,feature-store,vector-database,llm,genai

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/__init__.py

@@ -3,8 +3,7 @@
 from .__version__ import __version__, __version_tuple__
 from .catalog import Column, InsertableTable, Table, UpdateStatus, View
 from .dataframe import DataFrame
-from .exceptions import Error, PixeltableWarning
-from .exprs import RELATIVE_PATH_ROOT
+from .exceptions import Error, ExprEvalError, PixeltableWarning
 from .func import Aggregator, Function, expr_udf, query, uda, udf
 from .globals import (
     array,

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/__version__.py

@@ -1,3 +1,3 @@
 # These version placeholders will be replaced during build.
-__version__ = '0.3.9'
-__version_tuple__ = (0, 3, 9)
+__version__ = '0.3.10'
+__version_tuple__ = (0, 3, 10)

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/catalog/column.py

@@ -202,6 +202,11 @@ class Column:
         assert self.tbl is not None
         return self.tbl.get().media_validation
 
+    @property
+    def is_required_for_insert(self) -> bool:
+        """Returns True if column is required when inserting rows."""
+        return not self.col_type.nullable and not self.is_computed
+
     def source(self) -> None:
         """
         If this is a computed col and the top-level expr is a function call, print the source, if possible.

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/catalog/globals.py

@@ -40,6 +40,22 @@ class UpdateStatus:
         self.cols_with_excs = list(dict.fromkeys(self.cols_with_excs + other.cols_with_excs))
         return self
 
+    @property
+    def insert_msg(self) -> str:
+        """Return a message describing the results of an insert operation."""
+        if self.num_excs == 0:
+            cols_with_excs_str = ''
+        else:
+            cols_with_excs_str = (
+                f' across {len(self.cols_with_excs)} column{"" if len(self.cols_with_excs) == 1 else "s"}'
+            )
+            cols_with_excs_str += f' ({", ".join(self.cols_with_excs)})'
+        msg = (
+            f'Inserted {self.num_rows} row{"" if self.num_rows == 1 else "s"} '
+            f'with {self.num_excs} error{"" if self.num_excs == 1 else "s"}{cols_with_excs_str}.'
+        )
+        return msg
+
 
 class MediaValidation(enum.Enum):
     ON_READ = 0

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/catalog/insertable_table.py

@@ -1,7 +1,8 @@
 from __future__ import annotations
 
+import enum
 import logging
-from typing import Any, Iterable, Literal, Optional, overload
+from typing import TYPE_CHECKING, Any, Iterable, Literal, Optional, overload
 from uuid import UUID
 
 import pixeltable as pxt
@@ -16,9 +17,36 @@ from .table_version import TableVersion
 from .table_version_handle import TableVersionHandle
 from .table_version_path import TableVersionPath
 
+if TYPE_CHECKING:
+    import datasets  # type: ignore[import-untyped]
+
+    from pixeltable.globals import RowData, TableDataSource
+    from pixeltable.io.table_data_conduit import TableDataConduit
+
 _logger = logging.getLogger('pixeltable')
 
 
+class OnErrorParameter(enum.Enum):
+    """Supported values for the on_error parameter"""
+
+    ABORT = 'abort'
+    IGNORE = 'ignore'
+
+    @classmethod
+    def is_valid(cls, v: Any) -> bool:
+        if isinstance(v, str):
+            return v.lower() in [c.value for c in cls]
+        return False
+
+    @classmethod
+    def fail_on_exception(cls, v: Any) -> bool:
+        if not cls.is_valid(v):
+            raise ValueError(f'Invalid value for on_error: {v}')
+        if isinstance(v, str):
+            return v.lower() != cls.IGNORE.value
+        return True
+
+
 class InsertableTable(Table):
     """A `Table` that allows inserting and deleting rows."""
 
@@ -86,62 +114,75 @@ class InsertableTable(Table):
     @overload
     def insert(
         self,
-        rows: Iterable[dict[str, Any]],
+        source: Optional[TableDataSource] = None,
         /,
         *,
-        print_stats: bool = False,
+        source_format: Optional[Literal['csv', 'excel', 'parquet', 'json']] = None,
+        schema_overrides: Optional[dict[str, ts.ColumnType]] = None,
         on_error: Literal['abort', 'ignore'] = 'abort',
+        print_stats: bool = False,
+        **kwargs: Any,
     ) -> UpdateStatus: ...
 
     @overload
     def insert(
-        self, *, print_stats: bool = False, on_error: Literal['abort', 'ignore'] = 'abort', **kwargs: Any
+        self, /, *, on_error: Literal['abort', 'ignore'] = 'abort', print_stats: bool = False, **kwargs: Any
     ) -> UpdateStatus: ...
 
-    def insert(  # type: ignore[misc]
+    def insert(
         self,
-        rows: Optional[Iterable[dict[str, Any]]] = None,
+        source: Optional[TableDataSource] = None,
         /,
         *,
-        print_stats: bool = False,
+        source_format: Optional[Literal['csv', 'excel', 'parquet', 'json']] = None,
+        schema_overrides: Optional[dict[str, ts.ColumnType]] = None,
         on_error: Literal['abort', 'ignore'] = 'abort',
+        print_stats: bool = False,
         **kwargs: Any,
     ) -> UpdateStatus:
-        if rows is None:
-            rows = [kwargs]
-        else:
-            rows = list(rows)
-            if len(kwargs) > 0:
-                raise excs.Error('`kwargs` cannot be specified unless `rows is None`.')
-
-        fail_on_exception = on_error == 'abort'
-
-        if not isinstance(rows, list):
-            raise excs.Error('rows must be a list of dictionaries')
-        if len(rows) == 0:
-            raise excs.Error('rows must not be empty')
-        for row in rows:
-            if not isinstance(row, dict):
-                raise excs.Error('rows must be a list of dictionaries')
-        self._validate_input_rows(rows)
-        with Env.get().begin_xact():
-            status = self._tbl_version.get().insert(
-                rows, None, print_stats=print_stats, fail_on_exception=fail_on_exception
-            )
-
-        if status.num_excs == 0:
-            cols_with_excs_str = ''
-        else:
-            cols_with_excs_str = (
-                f' across {len(status.cols_with_excs)} column{"" if len(status.cols_with_excs) == 1 else "s"}'
-            )
-            cols_with_excs_str += f' ({", ".join(status.cols_with_excs)})'
-        msg = (
-            f'Inserted {status.num_rows} row{"" if status.num_rows == 1 else "s"} '
-            f'with {status.num_excs} error{"" if status.num_excs == 1 else "s"}{cols_with_excs_str}.'
+        from pixeltable.io.table_data_conduit import UnkTableDataConduit
+
+        table = self
+        if source is None:
+            source = [kwargs]
+            kwargs = None
+
+        tds = UnkTableDataConduit(
+            source, source_format=source_format, src_schema_overrides=schema_overrides, extra_fields=kwargs
         )
-        Env.get().console_logger.info(msg)
-        _logger.info(f'InsertableTable {self._name}: {msg}')
+        data_source = tds.specialize()
+        if data_source.source_column_map is None:
+            data_source.src_pk = []
+
+        assert isinstance(table, Table)
+        data_source.add_table_info(table)
+        data_source.prepare_for_insert_into_table()
+
+        fail_on_exception = OnErrorParameter.fail_on_exception(on_error)
+        return table.insert_table_data_source(
+            data_source=data_source, fail_on_exception=fail_on_exception, print_stats=print_stats
+        )
+
+    def insert_table_data_source(
+        self, data_source: TableDataConduit, fail_on_exception: bool, print_stats: bool = False
+    ) -> pxt.UpdateStatus:
+        """Insert row batches into this table from a `TableDataConduit`."""
+        from pixeltable.io.table_data_conduit import DFTableDataConduit, TableDataConduit
+
+        status = pxt.UpdateStatus()
+        with Env.get().begin_xact():
+            if isinstance(data_source, DFTableDataConduit):
+                status += self._tbl_version.get().insert(
+                    rows=None, df=data_source.pxt_df, print_stats=print_stats, fail_on_exception=fail_on_exception
+                )
+            else:
+                for row_batch in data_source.valid_row_batch():
+                    status += self._tbl_version.get().insert(
+                        rows=row_batch, df=None, print_stats=print_stats, fail_on_exception=fail_on_exception
+                    )
+
+        Env.get().console_logger.info(status.insert_msg)
+
         FileCache.get().emit_eviction_warnings()
         return status
 

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/catalog/table.py

@@ -8,6 +8,7 @@ from pathlib import Path
 from typing import TYPE_CHECKING, Any, Iterable, Literal, Optional, Union, overload
 
 from typing import _GenericAlias  # type: ignore[attr-defined]  # isort: skip
+from keyword import iskeyword as is_python_keyword
 from uuid import UUID
 
 import pandas as pd
@@ -42,9 +43,11 @@ from .table_version_handle import TableVersionHandle
 from .table_version_path import TableVersionPath
 
 if TYPE_CHECKING:
+    import datasets  # type: ignore[import-untyped]
     import torch.utils.data
 
     import pixeltable.plan
+    from pixeltable.globals import RowData, TableDataSource
 
 _logger = logging.getLogger('pixeltable')
 
@@ -720,13 +723,18 @@ class Table(SchemaObject):
             columns.append(column)
         return columns
 
+    @classmethod
+    def validate_column_name(cls, name: str) -> None:
+        """Check that a name is usable as a pixeltalbe column name"""
+        if is_system_column_name(name) or is_python_keyword(name):
+            raise excs.Error(f'{name!r} is a reserved name in Pixeltable; please choose a different column name.')
+        if not is_valid_identifier(name):
+            raise excs.Error(f'Invalid column name: {name!r}')
+
     @classmethod
     def _verify_column(cls, col: Column) -> None:
         """Check integrity of user-supplied Column and supply defaults"""
-        if is_system_column_name(col.name):
-            raise excs.Error(f'{col.name!r} is a reserved name in Pixeltable; please choose a different column name.')
-        if not is_valid_identifier(col.name):
-            raise excs.Error(f'Invalid column name: {col.name!r}')
+        cls.validate_column_name(col.name)
         if col.stored is False and not col.is_computed:
             raise excs.Error(f'Column {col.name!r}: stored={col.stored} only applies to computed columns')
         if col.stored is False and col.has_window_fn_call():
@@ -745,16 +753,6 @@ class Table(SchemaObject):
             cls._verify_column(col)
             column_names.add(col.name)
 
-    def __check_column_name_exists(self, column_name: str, include_bases: bool = False) -> None:
-        col = self._tbl_version_path.get_column(column_name, include_bases)
-        if col is None:
-            raise excs.Error(f'Column {column_name!r} unknown')
-
-    def __check_column_ref_exists(self, col_ref: ColumnRef, include_bases: bool = False) -> None:
-        exists = self._tbl_version_path.has_column(col_ref.col, include_bases)
-        if not exists:
-            raise excs.Error(f'Unknown column: {col_ref.col.qualified_name}')
-
     def drop_column(self, column: Union[str, ColumnRef], if_not_exists: Literal['error', 'ignore'] = 'error') -> None:
         """Drop a column from the table.
 
@@ -907,7 +905,7 @@ class Table(SchemaObject):
         Args:
             column: The name of, or reference to, the column to be indexed; must be a `String` or `Image` column.
             idx_name: An optional name for the index. If not specified, a name such as `'idx0'` will be generated
-                automatically. If specified, the name must be unique for this table.
+                automatically. If specified, the name must be unique for this table and a valid pixeltable column name.
             embedding: The UDF to use for the embedding. Must be a UDF that accepts a single argument of type `String`
                 or `Image` (as appropriate for the column being indexed) and returns a fixed-size 1-dimensional
                 array of floats.
@@ -960,13 +958,7 @@ class Table(SchemaObject):
         """
         if self._tbl_version_path.is_snapshot():
             raise excs.Error('Cannot add an index to a snapshot')
-        col: Column
-        if isinstance(column, str):
-            self.__check_column_name_exists(column, include_bases=True)
-            col = self._tbl_version_path.get_column(column, include_bases=True)
-        else:
-            self.__check_column_ref_exists(column, include_bases=True)
-            col = column.col
+        col = self._resolve_column_parameter(column)
 
         with Env.get().begin_xact():
             if idx_name is not None and idx_name in self._tbl_version.get().idxs_by_name:
@@ -986,6 +978,10 @@ class Table(SchemaObject):
                 assert idx_name not in self._tbl_version.get().idxs_by_name
             from pixeltable.index import EmbeddingIndex
 
+            # idx_name must be a valid pixeltable column name
+            if idx_name is not None:
+                Table.validate_column_name(idx_name)
+
             # create the EmbeddingIndex instance to verify args
             idx = EmbeddingIndex(
                 col, metric=metric, embed=embedding, string_embed=string_embed, image_embed=image_embed
@@ -1049,17 +1045,28 @@ class Table(SchemaObject):
 
         col: Column = None
         if idx_name is None:
-            if isinstance(column, str):
-                self.__check_column_name_exists(column, include_bases=True)
-                col = self._tbl_version_path.get_column(column, include_bases=True)
-            else:
-                self.__check_column_ref_exists(column, include_bases=True)
-                col = column.col
+            col = self._resolve_column_parameter(column)
             assert col is not None
 
         with Env.get().begin_xact():
             self._drop_index(col=col, idx_name=idx_name, _idx_class=index.EmbeddingIndex, if_not_exists=if_not_exists)
 
+    def _resolve_column_parameter(self, column: Union[str, ColumnRef]) -> Column:
+        """Resolve a column parameter to a Column object"""
+        col: Column = None
+        if isinstance(column, str):
+            col = self._tbl_version_path.get_column(column, include_bases=True)
+            if col is None:
+                raise excs.Error(f'Column {column!r} unknown')
+        elif isinstance(column, ColumnRef):
+            exists = self._tbl_version_path.has_column(column.col, include_bases=True)
+            if not exists:
+                raise excs.Error(f'Unknown column: {column.col.qualified_name}')
+            col = column.col
+        else:
+            raise excs.Error(f'Invalid column parameter type: {type(column)}')
+        return col
+
     def drop_index(
         self,
         *,
@@ -1115,12 +1122,7 @@ class Table(SchemaObject):
 
         col: Column = None
         if idx_name is None:
-            if isinstance(column, str):
-                self.__check_column_name_exists(column, include_bases=True)
-                col = self._tbl_version_path.get_column(column, include_bases=True)
-            else:
-                self.__check_column_ref_exists(column, include_bases=True)
-                col = column.col
+            col = self._resolve_column_parameter(column)
             assert col is not None
 
         with Env.get().begin_xact():
@@ -1145,49 +1147,62 @@ class Table(SchemaObject):
                     raise excs.Error(f'Index {idx_name!r} does not exist')
                 assert _if_not_exists == IfNotExistsParam.IGNORE
                 return
-            idx_id = self._tbl_version.get().idxs_by_name[idx_name].id
+            idx_info = self._tbl_version.get().idxs_by_name[idx_name]
         else:
             if col.tbl.id != self._tbl_version.id:
                 raise excs.Error(
                     f'Column {col.name!r}: cannot drop index from column that belongs to base ({col.tbl.get().name}!r)'
                 )
-            idx_info = [info for info in self._tbl_version.get().idxs_by_name.values() if info.col.id == col.id]
+            idx_info_list = [info for info in self._tbl_version.get().idxs_by_name.values() if info.col.id == col.id]
             if _idx_class is not None:
-                idx_info = [info for info in idx_info if isinstance(info.idx, _idx_class)]
-            if len(idx_info) == 0:
+                idx_info_list = [info for info in idx_info_list if isinstance(info.idx, _idx_class)]
+            if len(idx_info_list) == 0:
                 _if_not_exists = IfNotExistsParam.validated(if_not_exists, 'if_not_exists')
                 if _if_not_exists == IfNotExistsParam.ERROR:
                     raise excs.Error(f'Column {col.name!r} does not have an index')
                 assert _if_not_exists == IfNotExistsParam.IGNORE
                 return
-            if len(idx_info) > 1:
+            if len(idx_info_list) > 1:
                 raise excs.Error(f"Column {col.name!r} has multiple indices; specify 'idx_name' instead")
-            idx_id = idx_info[0].id
-        self._tbl_version.get().drop_index(idx_id)
+            idx_info = idx_info_list[0]
+
+        # Find out if anything depends on this index
+        dependent_user_cols = [c for c in idx_info.val_col.dependent_cols if c.name is not None]
+        if len(dependent_user_cols) > 0:
+            raise excs.Error(
+                f'Cannot drop index because the following columns depend on it:\n'
+                f'{", ".join(c.name for c in dependent_user_cols)}'
+            )
+        self._tbl_version.get().drop_index(idx_info.id)
 
     @overload
     def insert(
         self,
-        rows: Iterable[dict[str, Any]],
+        source: TableDataSource,
         /,
         *,
-        print_stats: bool = False,
+        source_format: Optional[Literal['csv', 'excel', 'parquet', 'json']] = None,
+        schema_overrides: Optional[dict[str, ts.ColumnType]] = None,
         on_error: Literal['abort', 'ignore'] = 'abort',
+        print_stats: bool = False,
+        **kwargs: Any,
     ) -> UpdateStatus: ...
 
     @overload
     def insert(
-        self, *, print_stats: bool = False, on_error: Literal['abort', 'ignore'] = 'abort', **kwargs: Any
+        self, /, *, on_error: Literal['abort', 'ignore'] = 'abort', print_stats: bool = False, **kwargs: Any
     ) -> UpdateStatus: ...
 
-    @abc.abstractmethod  # type: ignore[misc]
+    @abc.abstractmethod
     def insert(
         self,
-        rows: Optional[Iterable[dict[str, Any]]] = None,
+        source: Optional[TableDataSource] = None,
         /,
         *,
-        print_stats: bool = False,
+        source_format: Optional[Literal['csv', 'excel', 'parquet', 'json']] = None,
+        schema_overrides: Optional[dict[str, ts.ColumnType]] = None,
         on_error: Literal['abort', 'ignore'] = 'abort',
+        print_stats: bool = False,
         **kwargs: Any,
     ) -> UpdateStatus:
         """Inserts rows into this table. There are two mutually exclusive call patterns:
@@ -1196,11 +1211,12 @@ class Table(SchemaObject):
 
         ```python
         insert(
-            rows: Iterable[dict[str, Any]],
+            source: TableSourceDataType,
             /,
             *,
+            on_error: Literal['abort', 'ignore'] = 'abort',
             print_stats: bool = False,
-            on_error: Literal['abort', 'ignore'] = 'abort'
+            **kwargs: Any,
         )```
 
         To insert just a single row, you can use the more concise syntax:
@@ -1208,23 +1224,25 @@ class Table(SchemaObject):
         ```python
         insert(
             *,
-            print_stats: bool = False,
             on_error: Literal['abort', 'ignore'] = 'abort',
+            print_stats: bool = False,
             **kwargs: Any
         )```
 
         Args:
-            rows: (if inserting multiple rows) A list of rows to insert, each of which is a dictionary mapping column
-                names to values.
+            source: A data source from which data can be imported.
             kwargs: (if inserting a single row) Keyword-argument pairs representing column names and values.
-            print_stats: If `True`, print statistics about the cost of computed columns.
+                (if inserting multiple rows) Additional keyword arguments are passed to the data source.
+            source_format: A hint about the format of the source data
+            schema_overrides: If specified, then columns in `schema_overrides` will be given the specified types
             on_error: Determines the behavior if an error occurs while evaluating a computed column or detecting an
                 invalid media file (such as a corrupt image) for one of the inserted rows.
 
                 - If `on_error='abort'`, then an exception will be raised and the rows will not be inserted.
                 - If `on_error='ignore'`, then execution will continue and the rows will be inserted. Any cells
-                  with errors will have a `None` value for that cell, with information about the error stored in the
-                  corresponding `tbl.col_name.errortype` and `tbl.col_name.errormsg` fields.
+                    with errors will have a `None` value for that cell, with information about the error stored in the
+                    corresponding `tbl.col_name.errortype` and `tbl.col_name.errormsg` fields.
+            print_stats: If `True`, print statistics about the cost of computed columns.
 
         Returns:
             An [`UpdateStatus`][pixeltable.UpdateStatus] object containing information about the update.
@@ -1236,6 +1254,7 @@ class Table(SchemaObject):
                 - The table has been dropped.
                 - One of the rows being inserted does not conform to the table schema.
                 - An error occurs during processing of computed columns, and `on_error='ignore'`.
+                - An error occurs while importing data from a source, and `on_error='abort'`.
 
         Examples:
             Insert two rows into the table `my_table` with three int columns ``a``, ``b``, and ``c``.
@@ -1247,6 +1266,10 @@ class Table(SchemaObject):
             Insert a single row using the alternative syntax:
 
             >>> tbl.insert(a=3, b=3, c=3)
+
+            Insert rows from a CSV file:
+
+            >>> tbl.insert(source='path/to/file.csv')
         """
         raise NotImplementedError
 

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/catalog/table_version.py

@@ -225,7 +225,9 @@ class TableVersion:
         # create schema.Table
         # Column.dependent_cols for existing cols is wrong at this point, but init() will set it correctly
         column_md = cls._create_column_md(cols)
+        tbl_id = uuid.uuid4()
         table_md = schema.TableMd(
+            tbl_id=str(tbl_id),
             name=name,
             user=None,
             current_version=0,
@@ -241,11 +243,12 @@ class TableVersion:
         )
         # create a schema.Table here, we need it to call our c'tor;
         # don't add it to the session yet, we might add index metadata
-        tbl_id = uuid.uuid4()
         tbl_record = schema.Table(id=tbl_id, dir_id=dir_id, md=dataclasses.asdict(table_md))
 
         # create schema.TableVersion
-        table_version_md = schema.TableVersionMd(created_at=timestamp, version=0, schema_version=0, additional_md={})
+        table_version_md = schema.TableVersionMd(
+            tbl_id=str(tbl_record.id), created_at=timestamp, version=0, schema_version=0, additional_md={}
+        )
         tbl_version_record = schema.TableVersion(
             tbl_id=tbl_record.id, version=0, md=dataclasses.asdict(table_version_md)
         )
@@ -261,6 +264,7 @@ class TableVersion:
             schema_col_md[col.id] = md
 
         schema_version_md = schema.TableSchemaVersionMd(
+            tbl_id=str(tbl_record.id),
             schema_version=0,
             preceding_schema_version=None,
             columns=schema_col_md,
@@ -1240,6 +1244,11 @@ class TableVersion:
         """Return all non-system columns"""
         return [c for c in self.cols if c.is_pk]
 
+    @property
+    def primary_key(self) -> list[str]:
+        """Return the names of the primary key columns"""
+        return [c.name for c in self.cols if c.is_pk]
+
     def get_required_col_names(self) -> list[str]:
         """Return the names of all columns for which values must be specified in insert()"""
         assert not self.is_view
@@ -1306,6 +1315,7 @@ class TableVersion:
 
     def _create_tbl_md(self) -> schema.TableMd:
         return schema.TableMd(
+            tbl_id=str(self.id),
             name=self.name,
             user=None,
             current_version=self.version,
@@ -1322,7 +1332,11 @@ class TableVersion:
 
     def _create_version_md(self, timestamp: float) -> schema.TableVersionMd:
         return schema.TableVersionMd(
-            created_at=timestamp, version=self.version, schema_version=self.schema_version, additional_md={}
+            tbl_id=str(self.id),
+            created_at=timestamp,
+            version=self.version,
+            schema_version=self.schema_version,
+            additional_md={},
         )
 
     def _create_schema_version_md(self, preceding_schema_version: int) -> schema.TableSchemaVersionMd:
@@ -1335,6 +1349,7 @@ class TableVersion:
             )
         # preceding_schema_version to be set by the caller
         return schema.TableSchemaVersionMd(
+            tbl_id=str(self.id),
             schema_version=self.schema_version,
             preceding_schema_version=preceding_schema_version,
             columns=column_md,

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/catalog/view.py

@@ -251,13 +251,20 @@ class View(Table):
         md['is_snapshot'] = self._tbl_version_path.is_snapshot()
         return md
 
+    if TYPE_CHECKING:
+        import datasets  # type: ignore[import-untyped]
+
+        from pixeltable.globals import RowData, TableDataSource
+
     def insert(
         self,
-        rows: Optional[Iterable[dict[str, Any]]] = None,
+        source: Optional[TableDataSource] = None,
         /,
         *,
-        print_stats: bool = False,
+        source_format: Optional[Literal['csv', 'excel', 'parquet', 'json']] = None,
+        schema_overrides: Optional[dict[str, ts.ColumnType]] = None,
         on_error: Literal['abort', 'ignore'] = 'abort',
+        print_stats: bool = False,
         **kwargs: Any,
     ) -> UpdateStatus:
         raise excs.Error(f'{self._display_name()} {self._name!r}: cannot insert into view')

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/env.py

@@ -567,7 +567,7 @@ class Env:
         self.__register_package('transformers')
         self.__register_package('whisper', library_name='openai-whisper')
         self.__register_package('whisperx')
-        self.__register_package('yolox', library_name='git+https://github.com/Megvii-BaseDetection/YOLOX@ac58e0a')
+        self.__register_package('yolox', library_name='pixeltable-yolox')
 
     def __register_package(self, package_name: str, library_name: Optional[str] = None) -> None:
         is_installed: bool

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/exec/exec_node.py

@@ -75,7 +75,7 @@ class ExecNode(abc.ABC):
             loop = asyncio.new_event_loop()
             asyncio.set_event_loop(loop)
 
-        if 'pytest' in sys.modules:
+        if _logger.isEnabledFor(logging.DEBUG):
             loop.set_debug(True)
 
         aiter = self.__aiter__()

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/exprs/__init__.py

@@ -16,7 +16,7 @@ from .in_predicate import InPredicate
 from .inline_expr import InlineArray, InlineDict, InlineList
 from .is_null import IsNull
 from .json_mapper import JsonMapper
-from .json_path import RELATIVE_PATH_ROOT, JsonPath
+from .json_path import JsonPath
 from .literal import Literal
 from .method_ref import MethodRef
 from .object_ref import ObjectRef
@@ -24,5 +24,6 @@ from .row_builder import ColumnSlotIdx, ExecProfile, RowBuilder
 from .rowid_ref import RowidRef
 from .similarity_expr import SimilarityExpr
 from .sql_element_cache import SqlElementCache
+from .string_op import StringOp
 from .type_cast import TypeCast
 from .variable import Variable

{pixeltable-0.3.9 → pixeltable-0.3.10}/pixeltable/exprs/arithmetic_expr.py

@@ -19,6 +19,8 @@ class ArithmeticExpr(Expr):
     Allows arithmetic exprs on json paths
     """
 
+    operator: ArithmeticOperator
+
     def __init__(self, operator: ArithmeticOperator, op1: Expr, op2: Expr):
         if op1.col_type.is_json_type() or op2.col_type.is_json_type() or operator == ArithmeticOperator.DIV:
             # we assume it's a float