datachain 0.24.1__py3-none-any.whl → 0.24.3__py3-none-any.whl

datachain/catalog/catalog.py

@@ -1098,9 +1098,18 @@ class Catalog:
     ) -> DatasetRecord:
         from datachain.lib.listing import is_listing_dataset
 
+        project = project or self.metastore.default_project
+
         if is_listing_dataset(name):
             project = self.metastore.listing_project
-        return self.metastore.get_dataset(name, project.id if project else None)
+
+        try:
+            return self.metastore.get_dataset(name, project.id if project else None)
+        except DatasetNotFoundError:
+            raise DatasetNotFoundError(
+                f"Dataset {name} not found in namespace {project.namespace.name}"
+                f" and project {project.name}"
+            ) from None
 
     def get_dataset_with_remote_fallback(
         self,
@@ -1124,7 +1133,7 @@ class Catalog:
             raise DatasetNotFoundError(
                 f"Dataset {name}"
                 + (f" version {version} " if version else " ")
-                + "not found"
+                + f"not found in namespace {namespace_name} and project {project_name}"
             )
 
         if pull_dataset:

datachain/data_storage/metastore.py

@@ -1194,14 +1194,16 @@ class AbstractDBMetastore(AbstractMetastore):
         Gets a single dataset in project by dataset name.
         """
         project_id = project_id or self.default_project.id
+
         d = self._datasets
         query = self._base_dataset_query()
         query = query.where(d.c.name == name, d.c.project_id == project_id)  # type: ignore [attr-defined]
         ds = self._parse_dataset(self.db.execute(query, conn=conn))
         if not ds:
             raise DatasetNotFoundError(
-                f"Dataset {name} not found in project {project_id}"
+                f"Dataset {name} not found in project with id {project_id}"
             )
+
         return ds
 
     def remove_dataset_version(

datachain/data_storage/sqlite.py

@@ -774,7 +774,15 @@ class SQLiteWarehouse(AbstractWarehouse):
         query: Select,
         progress_cb: Optional[Callable[[int], None]] = None,
     ) -> None:
-        if len(query._group_by_clause) > 0:
+        col_id = (
+            query.selected_columns.sys__id
+            if "sys__id" in query.selected_columns
+            else None
+        )
+
+        # If there is no sys__id column, we cannot copy the table in batches,
+        # and we need to copy all rows at once. Same if there is a group by clause.
+        if col_id is None or len(query._group_by_clause) > 0:
             select_q = query.with_only_columns(
                 *[c for c in query.selected_columns if c.name != "sys__id"]
             )
@@ -782,12 +790,7 @@ class SQLiteWarehouse(AbstractWarehouse):
             self.db.execute(q)
             return
 
-        if "sys__id" in query.selected_columns:
-            col_id = query.selected_columns.sys__id
-        else:
-            col_id = sqlalchemy.column("sys__id")
         select_ids = query.with_only_columns(col_id)
-
         ids = self.db.execute(select_ids).fetchall()
 
         select_q = (

datachain/data_storage/warehouse.py

@@ -218,7 +218,7 @@ class AbstractWarehouse(ABC, Serializable):
         limit = query._limit
         paginated_query = query.limit(page_size)
 
-        offset = 0
+        offset = query._offset or 0
         num_yielded = 0
 
         # Ensure we're using a thread-local connection
@@ -234,13 +234,13 @@ class AbstractWarehouse(ABC, Serializable):
                 # Cursor results are not thread-safe, so we convert them to a list
                 results = list(wh.dataset_rows_select(paginated_query.offset(offset)))
 
-                processed = False
+                processed = 0
                 for row in results:
-                    processed = True
+                    processed += 1
                     yield row
                     num_yielded += 1
 
-                if not processed:
+                if processed < page_size:
                     break  # no more results
                 offset += page_size
 
@@ -343,6 +343,8 @@ class AbstractWarehouse(ABC, Serializable):
         if (id_col := get_query_id_column(query)) is None:
             raise RuntimeError("sys__id column not found in query")
 
+        query = query._clone().offset(None).limit(None).order_by(None)
+
         if is_batched:
             for batch in ids:
                 yield list(self.dataset_rows_select(query.where(id_col.in_(batch))))

datachain/listing.py

@@ -65,10 +65,17 @@ class Listing:
 
     @cached_property
     def dataset(self) -> "DatasetRecord":
+        from datachain.error import DatasetNotFoundError
+
         assert self.dataset_name
-        return self.metastore.get_dataset(
-            self.dataset_name, self.metastore.listing_project.id
-        )
+        project = self.metastore.listing_project
+        try:
+            return self.metastore.get_dataset(self.dataset_name, project.id)
+        except DatasetNotFoundError:
+            raise DatasetNotFoundError(
+                f"Dataset {self.dataset_name} not found in namespace"
+                f" {project.namespace.name} and project {project.name}"
+            ) from None
 
     @cached_property
     def dataset_rows(self):

datachain/query/dataset.py

@@ -11,6 +11,7 @@ from collections.abc import Generator, Iterable, Iterator, Sequence
 from copy import copy
 from functools import wraps
 from secrets import token_hex
+from types import GeneratorType
 from typing import (
     TYPE_CHECKING,
     Any,
@@ -557,8 +558,8 @@ class UDFStep(Step, ABC):
         """
         assert self.partition_by is not None
 
-        if isinstance(self.partition_by, Sequence):
-            list_partition_by = self.partition_by
+        if isinstance(self.partition_by, (list, tuple, GeneratorType)):
+            list_partition_by = list(self.partition_by)
         else:
             list_partition_by = [self.partition_by]
 
@@ -575,7 +576,10 @@ class UDFStep(Step, ABC):
             f.dense_rank().over(order_by=partition_by).label(PARTITION_COLUMN_ID),
         ]
         self.catalog.warehouse.db.execute(
-            tbl.insert().from_select(cols, query.with_only_columns(*cols))
+            tbl.insert().from_select(
+                cols,
+                query.offset(None).limit(None).with_only_columns(*cols),
+            )
         )
 
         return tbl
@@ -601,13 +605,10 @@ class UDFStep(Step, ABC):
         if self.partition_by is not None:
             partition_tbl = self.create_partitions_table(query)
             temp_tables.append(partition_tbl.name)
-
-            subq = query.subquery()
-            query = (
-                sqlalchemy.select(*subq.c)
-                .outerjoin(partition_tbl, partition_tbl.c.sys__id == subq.c.sys__id)
-                .add_columns(*partition_columns())
-            )
+            query = query.outerjoin(
+                partition_tbl,
+                partition_tbl.c.sys__id == query.selected_columns.sys__id,
+            ).add_columns(*partition_columns())
 
         query, tables = self.process_input_query(query)
         temp_tables.extend(t.name for t in tables)

{datachain-0.24.1.dist-info → datachain-0.24.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: datachain
-Version: 0.24.1
+Version: 0.24.3
 Summary: Wrangle unstructured AI data at scale
 Author-email: Dmitry Petrov <support@dvc.org>
 License-Expression: Apache-2.0

{datachain-0.24.1.dist-info → datachain-0.24.3.dist-info}/RECORD

@@ -7,7 +7,7 @@ datachain/dataset.py,sha256=wDrukmkDnYP0X8bAGY-7O1NDE3DWCFqrH8VVDpXM9Ok,25263
 datachain/delta.py,sha256=fTEhCedseUsHuH_Ek52NXFhFPyFD_6MioEH5sCilNgo,9897
 datachain/error.py,sha256=OWwWMkzZYJrkcoEDGhJHMf7SfKvxcsOLRF94mjPf29I,1609
 datachain/job.py,sha256=x5PB6d5sqx00hePNNkirESlOVAvnmkEM5ygUgQmAhsk,1262
-datachain/listing.py,sha256=T4bCgdCRuFW7bsPUG2PSl5om2nfJL6fzB84m7mCO8cA,7136
+datachain/listing.py,sha256=GuUlj3oYJQOIbLYr3IqNCSeuiU5pfOtRAQQl9uBEFQU,7458
 datachain/namespace.py,sha256=r7YwpXFc7LdMS2jhyLMTrAHffALFG2bAIiYUSuZa5Rc,1786
 datachain/node.py,sha256=KWDT0ClYXB7FYI-QOvzAa-UDkLJErUI2eWm5FBteYuU,5577
 datachain/nodes_fetcher.py,sha256=_wgaKyqEjkqdwJ_Hj6D8vUYz7hnU7g6xhm0H6ZnYxmE,1095
@@ -21,7 +21,7 @@ datachain/studio.py,sha256=bLok-eJNFRHQScEyAyA_Fas52dmijd5r-73KudWxV4k,13337
 datachain/telemetry.py,sha256=0A4IOPPp9VlP5pyW9eBfaTK3YhHGzHl7dQudQjUAx9A,994
 datachain/utils.py,sha256=DNqOi-Ydb7InyWvD9m7_yailxz6-YGpZzh00biQaHNo,15305
 datachain/catalog/__init__.py,sha256=cMZzSz3VoUi-6qXSVaHYN-agxQuAcz2XSqnEPZ55crE,353
-datachain/catalog/catalog.py,sha256=z4GbRMHeW0YA20Sjh7QuPy1Rj4RkX547WN9Pp5wAD6o,65277
+datachain/catalog/catalog.py,sha256=7N4QmG6mpj8WaWcJSp7LijFQlVHbz_B6DLUnMKXmx6k,65644
 datachain/catalog/datasource.py,sha256=IkGMh0Ttg6Q-9DWfU_H05WUnZepbGa28HYleECi6K7I,1353
 datachain/catalog/loader.py,sha256=UXjYD6BNRoupPvkiz3-b04jepXhtLHCA4gzKFnXxOtQ,5987
 datachain/cli/__init__.py,sha256=WvBqnwjG8Wp9xGCn-4eqfoZ3n7Sj1HJemCi4MayJh_c,8221
@@ -49,11 +49,11 @@ datachain/client/s3.py,sha256=6DNVGLg-woPS1DVlYVX2rIlunNblsuxyOnI1rSzhW3k,7515
 datachain/data_storage/__init__.py,sha256=9Wit-oe5P46V7CJQTD0BJ5MhOa2Y9h3ddJ4VWTe-Lec,273
 datachain/data_storage/db_engine.py,sha256=n8ojCbvVMPY2e3SG8fUaaD0b9GkVfpl_Naa_6EiHfWg,3788
 datachain/data_storage/job.py,sha256=9r0OGwh22bHNIvLHqg8_-eJSP1YYB-BN5HOla5TdCxw,402
-datachain/data_storage/metastore.py,sha256=9mWYOKK3AoHeKPGFm-WBfPrmnYHhwYeXx5MOueKTe7I,52657
+datachain/data_storage/metastore.py,sha256=af7TsMHbANXmWKIu-LzQmsZpNxM6-hhzZfInWx7MQXI,52667
 datachain/data_storage/schema.py,sha256=o3JbURKXRg3IJyIVA4QjHHkn6byRuz7avbydU2FlvNY,9897
 datachain/data_storage/serializer.py,sha256=6G2YtOFqqDzJf1KbvZraKGXl2XHZyVml2krunWUum5o,927
-datachain/data_storage/sqlite.py,sha256=tT_soVi6l_pFSKaDktA1t4qW_vmPvXnvYSf4TZTKZYk,30067
-datachain/data_storage/warehouse.py,sha256=_7btARw-kd-Nx19S0qW6JqdF3VYyypQXFzsXq68SWKI,32327
+datachain/data_storage/sqlite.py,sha256=TTQjdDXUaZSr3MEaxZjDhsVIkIJqxFNA-sD25TO3m_4,30228
+datachain/data_storage/warehouse.py,sha256=2Bp2fXfcm-acwYjDWqVzGjoIQSAR4L56GPNtPcaT2gU,32418
 datachain/diff/__init__.py,sha256=-OFZzgOplqO84iWgGY7kfe60NXaWR9JRIh9T-uJboAM,9668
 datachain/fs/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 datachain/fs/reference.py,sha256=A8McpXF0CqbXPqanXuvpKu50YLB3a2ZXA3YAPxtBXSM,914
@@ -125,7 +125,7 @@ datachain/model/ultralytics/pose.py,sha256=pBlmt63Qe68FKmexHimUGlNbNOoOlMHXG4fzX
 datachain/model/ultralytics/segment.py,sha256=63bDCj43E6iZ0hFI5J6uQfksdCmjEp6sEm1XzVaE8pw,2986
 datachain/query/__init__.py,sha256=7DhEIjAA8uZJfejruAVMZVcGFmvUpffuZJwgRqNwe-c,263
 datachain/query/batch.py,sha256=-goxLpE0EUvaDHu66rstj53UnfHpYfBUGux8GSpJ93k,4306
-datachain/query/dataset.py,sha256=C60VM0pScsrWcMqLNdX-tU0HE1SnEE9lRN3TU8CfTu4,61223
+datachain/query/dataset.py,sha256=mKee4PkQHYPT96utPjM1DocURU4TghAR7AHtYkzdqwY,61292
 datachain/query/dispatch.py,sha256=A0nPxn6mEN5d9dDo6S8m16Ji_9IvJLXrgF2kqXdi4fs,15546
 datachain/query/metrics.py,sha256=DOK5HdNVaRugYPjl8qnBONvTkwjMloLqAr7Mi3TjCO0,858
 datachain/query/params.py,sha256=O_j89mjYRLOwWNhYZl-z7mi-rkdP7WyFmaDufsdTryE,863
@@ -157,9 +157,9 @@ datachain/sql/sqlite/vector.py,sha256=ncW4eu2FlJhrP_CIpsvtkUabZlQdl2D5Lgwy_cbfqR
 datachain/toolkit/__init__.py,sha256=eQ58Q5Yf_Fgv1ZG0IO5dpB4jmP90rk8YxUWmPc1M2Bo,68
 datachain/toolkit/split.py,sha256=ktGWzY4kyzjWyR86dhvzw-Zhl0lVk_LOX3NciTac6qo,2914
 datachain/torch/__init__.py,sha256=gIS74PoEPy4TB3X6vx9nLO0Y3sLJzsA8ckn8pRWihJM,579
-datachain-0.24.1.dist-info/licenses/LICENSE,sha256=8DnqK5yoPI_E50bEg_zsHKZHY2HqPy4rYN338BHQaRA,11344
-datachain-0.24.1.dist-info/METADATA,sha256=lr7Q889hnRechtjUZUnuwDSfiydAGE6wGxMMG9ICSVg,13281
-datachain-0.24.1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-datachain-0.24.1.dist-info/entry_points.txt,sha256=0GMJS6B_KWq0m3VT98vQI2YZodAMkn4uReZ_okga9R4,49
-datachain-0.24.1.dist-info/top_level.txt,sha256=lZPpdU_2jJABLNIg2kvEOBi8PtsYikbN1OdMLHk8bTg,10
-datachain-0.24.1.dist-info/RECORD,,
+datachain-0.24.3.dist-info/licenses/LICENSE,sha256=8DnqK5yoPI_E50bEg_zsHKZHY2HqPy4rYN338BHQaRA,11344
+datachain-0.24.3.dist-info/METADATA,sha256=qlqP_kAxEtkEJvV-ZzEPikT2OwW9S7Yqx4ly-iIOMBY,13281
+datachain-0.24.3.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+datachain-0.24.3.dist-info/entry_points.txt,sha256=0GMJS6B_KWq0m3VT98vQI2YZodAMkn4uReZ_okga9R4,49
+datachain-0.24.3.dist-info/top_level.txt,sha256=lZPpdU_2jJABLNIg2kvEOBi8PtsYikbN1OdMLHk8bTg,10
+datachain-0.24.3.dist-info/RECORD,,