lamindb 0.70.2__py3-none-any.whl → 0.70.3__py3-none-any.whl

lamindb/__init__.py

@@ -40,7 +40,7 @@ Modules & settings:
 
 """
 
-__version__ = "0.70.2"  # denote a release candidate for 0.1.0 with 0.1rc1
+__version__ = "0.70.3"  # denote a release candidate for 0.1.0 with 0.1rc1
 
 import os as _os
 

lamindb/_artifact.py

@@ -1,7 +1,7 @@
 from __future__ import annotations
 
 from pathlib import Path, PurePath, PurePosixPath
-from typing import TYPE_CHECKING, Any
+from typing import TYPE_CHECKING, Any, Mapping
 
 import fsspec
 import lamindb_setup as ln_setup
@@ -26,7 +26,7 @@ from lnschema_core.types import (
 )
 
 from lamindb._utils import attach_func_to_class_method
-from lamindb.core._data import _track_run_input
+from lamindb.core._data import Data, _track_run_input
 from lamindb.core._settings import settings
 from lamindb.core.storage import (
     LocalPathClasses,
@@ -53,6 +53,14 @@ from .core._data import (
 from .core.storage.objects import _mudata_is_installed
 from .core.storage.paths import AUTO_KEY_PREFIX
 
+try:
+    from .core.storage._zarr import zarr_is_adata
+except ImportError:
+
+    def zarr_is_adata(storepath):  # type: ignore
+        raise ImportError("Please install zarr: pip install zarr")
+
+
 if TYPE_CHECKING:
     from lamindb_setup.core.types import UPathStr
     from mudata import MuData
@@ -329,16 +337,17 @@ def get_artifact_kwargs_from_data(
         using_key=using_key,
     )
     if isinstance(stat_or_artifact, Artifact):
+        artifact = stat_or_artifact
         # update the run of the existing artifact
         if run is not None:
             # save the information that this artifact was previously
             # produced by another run
-            if stat_or_artifact.run is not None:
-                stat_or_artifact.run.replicated_output_artifacts.add(stat_or_artifact)
+            if artifact.run is not None:
+                artifact.run.replicated_output_artifacts.add(artifact)
             # update the run of the artifact with the latest run
             stat_or_artifact.run = run
             stat_or_artifact.transform = run.transform
-        return stat_or_artifact, None
+        return artifact, None
     else:
         size, hash, hash_type, n_objects = stat_or_artifact
 
@@ -446,12 +455,19 @@ def data_is_anndata(data: AnnData | UPathStr):
     if isinstance(data, AnnData):
         return True
     if isinstance(data, (str, Path, UPath)):
-        if Path(data).suffix == ".h5ad":
+        data_path = UPath(data)
+        if data_path.suffix == ".h5ad":
             return True
-        elif Path(data).suffix == ".zarr":
-            raise NotImplementedError(
-                "auto-detecting AnnData from Zarr is not yet supported"
-            )
+        elif data_path.suffix == ".zarr":
+            # ".anndata.zarr" is a valid suffix (core.storage._valid_suffixes)
+            if ".anndata" in data_path.suffixes:
+                return True
+            # check only for local, expensive for cloud
+            if fsspec.utils.get_protocol(data_path) == "file":
+                return zarr_is_adata(data_path)
+            else:
+                logger.warning("We do not check if cloud zarr is AnnData or not.")
+                return False
     return False
 
 
@@ -461,27 +477,39 @@ def data_is_mudata(data: MuData | UPathStr):
 
         if isinstance(data, MuData):
             return True
-    if isinstance(data, (str, Path, UPath)):
-        return Path(data).suffix in {".h5mu"}
+    if isinstance(data, (str, Path)):
+        return UPath(data).suffix in {".h5mu"}
     return False
 
 
 def _check_accessor_artifact(data: Any, accessor: str | None = None):
-    if accessor is None and not isinstance(data, (str, Path, UPath)):
+    if accessor is None:
         if isinstance(data, pd.DataFrame):
             logger.warning("data is a DataFrame, please use .from_df()")
             accessor = "DataFrame"
-        elif data_is_anndata(data):
-            logger.warning("data is an AnnData, please use .from_anndata()")
+            return accessor
+
+        data_is_path = isinstance(data, (str, Path))
+        if data_is_anndata(data):
+            if not data_is_path:
+                logger.warning("data is an AnnData, please use .from_anndata()")
             accessor = "AnnData"
         elif data_is_mudata(data):
-            logger.warning("data is a MuData, please use .from_mudata()")
+            if not data_is_path:
+                logger.warning("data is a MuData, please use .from_mudata()")
             accessor = "MuData"
-        else:
+        elif not data_is_path:  # UPath is a subclass of Path
             raise TypeError("data has to be a string, Path, UPath")
     return accessor
 
 
+def update_attributes(data: Data, attributes: Mapping[str, str]):
+    for key, value in attributes.items():
+        if getattr(data, key) != value:
+            logger.warning(f"updated {key} from {getattr(data, key)} to {value}")
+            setattr(data, key, value)
+
+
 def __init__(artifact: Artifact, *args, **kwargs):
     # Below checks for the Django-internal call in from_db()
     # it'd be better if we could avoid this, but not being able to create a Artifact
@@ -558,8 +586,13 @@ def __init__(artifact: Artifact, *args, **kwargs):
     if isinstance(kwargs_or_artifact, Artifact):
         from ._registry import init_self_from_db
 
-        # kwargs_or_artifact is an existing file
         init_self_from_db(artifact, kwargs_or_artifact)
+        # adding "key" here is dangerous because key might be auto-populated
+        update_attributes(artifact, {"description": description})
+        if artifact.key != key:
+            logger.warning(
+                f"key {artifact.key} on existing artifact differs from passed key {key}"
+            )
         return None
     else:
         kwargs = kwargs_or_artifact

lamindb/_collection.py

@@ -18,13 +18,13 @@ from lamindb_setup.core.hashing import hash_set
 from lnschema_core.models import Collection, CollectionArtifact, FeatureSet
 from lnschema_core.types import VisibilityChoice
 
+from lamindb._artifact import update_attributes
 from lamindb._utils import attach_func_to_class_method
 from lamindb.core._data import _track_run_input
 from lamindb.core._mapped_collection import MappedCollection
 from lamindb.core.versioning import get_uid_from_old_version, init_uid
 
 from . import Artifact, Run
-from ._artifact import data_is_anndata
 from ._registry import init_self_from_db
 from .core._data import (
     add_transform_to_kwargs,
@@ -128,6 +128,7 @@ def __init__(
             existing_collection.run = run
             existing_collection.transform = run.transform
         init_self_from_db(collection, existing_collection)
+        update_attributes(collection, {"description": description, "name": name})
         for slot, feature_set in collection.features._feature_set_by_slot.items():
             if slot in feature_sets:
                 if not feature_sets[slot] == feature_set:

lamindb/core/_run_context.py

@@ -306,6 +306,11 @@ class run_context:
             if not is_tracked:
                 raise_transform_settings_error()
         else:
+            if transform.type in {"notebook", "script"}:
+                raise ValueError(
+                    "Use ln.track() without passing transform in a notebook or script"
+                    " - metadata is automatically parsed"
+                )
             transform_exists = None
             if transform.id is not None:
                 # transform has an id but unclear whether already saved

lamindb/core/storage/_zarr.py

@@ -7,28 +7,44 @@ import scipy.sparse as sparse
 import zarr
 from anndata._io import read_zarr
 from anndata._io.specs import write_elem
+from anndata._io.specs.registry import get_spec
+from fsspec.implementations.local import LocalFileSystem
 from lamindb_setup.core.upath import create_mapper, infer_filesystem
 
 from ._anndata_sizes import _size_elem, _size_raw, size_adata
 
 if TYPE_CHECKING:
     from anndata import AnnData
-
-
-def read_adata_zarr(storepath) -> AnnData:
-    fs, storepath = infer_filesystem(storepath)
-    store = create_mapper(fs, storepath, check=True)
-
-    adata = read_zarr(store)
-
+    from lamindb_setup.core.types import UPathStr
+
+
+def zarr_is_adata(storepath: UPathStr) -> bool:
+    fs, storepath_str = infer_filesystem(storepath)
+    if isinstance(fs, LocalFileSystem):
+        # this is faster than through an fsspec mapper for local
+        open_obj = storepath_str
+    else:
+        open_obj = create_mapper(fs, storepath_str, check=True)
+    storage = zarr.open(open_obj, mode="r")
+    return get_spec(storage).encoding_type == "anndata"
+
+
+def read_adata_zarr(storepath: UPathStr) -> AnnData:
+    fs, storepath_str = infer_filesystem(storepath)
+    if isinstance(fs, LocalFileSystem):
+        # this is faster than through an fsspec mapper for local
+        open_obj = storepath_str
+    else:
+        open_obj = create_mapper(fs, storepath_str, check=True)
+    adata = read_zarr(open_obj)
     return adata
 
 
 def write_adata_zarr(
-    adata: AnnData, storepath, callback=None, chunks=None, **dataset_kwargs
+    adata: AnnData, storepath: UPathStr, callback=None, chunks=None, **dataset_kwargs
 ):
-    fs, storepath = infer_filesystem(storepath)
-    store = create_mapper(fs, storepath, create=True)
+    fs, storepath_str = infer_filesystem(storepath)
+    store = create_mapper(fs, storepath_str, create=True)
 
     f = zarr.open(store, mode="w")
 

lamindb/core/storage/paths.py

@@ -29,7 +29,7 @@ try:
     from ._zarr import read_adata_zarr
 except ImportError:
 
-    def read_adata_zarr(filepath):  # type: ignore
+    def read_adata_zarr(storepath):  # type: ignore
         raise ImportError("Please install zarr: pip install zarr")
 
 
@@ -114,7 +114,7 @@ def store_file_or_folder(local_path: UPathStr, storage_path: UPath) -> None:
     local_path = Path(local_path)
     if not isinstance(storage_path, LocalPathClasses):
         # this uploads files and directories
-        storage_path.upload_from(local_path, recursive=True, print_progress=True)
+        storage_path.upload_from(local_path, dir_inplace=True, print_progress=True)
     else:  # storage path is local
         storage_path.parent.mkdir(parents=True, exist_ok=True)
         if local_path.is_file():

{lamindb-0.70.2.dist-info → lamindb-0.70.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: lamindb
-Version: 0.70.2
+Version: 0.70.3
 Summary: A data framework for biology.
 Author-email: Lamin Labs <open-source@lamin.ai>
 Requires-Python: >=3.8
@@ -10,7 +10,7 @@ Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Requires-Dist: lnschema_core==0.65.1
-Requires-Dist: lamindb_setup==0.69.4
+Requires-Dist: lamindb_setup==0.69.5
 Requires-Dist: lamin_utils==0.13.2
 Requires-Dist: lamin_cli==0.12.3
 Requires-Dist: rapidfuzz

{lamindb-0.70.2.dist-info → lamindb-0.70.3.dist-info}/RECORD

@@ -1,8 +1,8 @@
-lamindb/__init__.py,sha256=JHrhBHkihUcLRbreh9xyafsWang75NJxz9CVIWx6Bc0,2163
+lamindb/__init__.py,sha256=y-amn3quDKMdf4F0-sm8mFvjJButkNAFPpwMP5bZNj0,2163
 lamindb/_annotate.py,sha256=B0KSvo5S2kJPeMMqy2SSFkqRJCS2QRC4NtI0_vWEZMs,43080
-lamindb/_artifact.py,sha256=BUl_3WYwrZ28P93Pb9AiwriVFBSLvBUEQjYEyYCrkZ0,37307
+lamindb/_artifact.py,sha256=SoGsS-uZP7AdrlEWuMC0l50EkNYSvtzrEiXzU5R6NWY,38557
 lamindb/_can_validate.py,sha256=nvoZG-35n3HofkY4Xc6hBv9AV54_RDan7Hzp5TuqY9I,14709
-lamindb/_collection.py,sha256=wf7ClfiD3vsetts_iSUk4UihWsHq0IdOF8LdIIHS7JU,14536
+lamindb/_collection.py,sha256=SDM35R_5WHrgLKjVb14Q8-Rz_gn5hdZLJobPcanm4PM,14627
 lamindb/_feature.py,sha256=srAKchY7gqD-h-cWlEiAWuHlpFKFwv0PWIA-JX0Go8c,6758
 lamindb/_feature_set.py,sha256=AzjOcHzQajpeikPOAic-aj0z_C5b7VpHVegg3ThRSLw,9045
 lamindb/_filter.py,sha256=xnjJzjF3Zj4dK_Kfymvhgczk27MhhXz5ZYc7XINbgHY,1331
@@ -25,7 +25,7 @@ lamindb/core/_data.py,sha256=En3v29eiJARy5l7nSsttAsDsqDLTZ4-xM8fCNyVzExI,17465
 lamindb/core/_feature_manager.py,sha256=LlYgU71AoTnrseWFCq-oZkUAYWITtRR7BNFm0AhHe-c,15773
 lamindb/core/_label_manager.py,sha256=0RtegYnK3zIisOnd970EobOrHMpp7OCH-mEoPrPXw2c,9075
 lamindb/core/_mapped_collection.py,sha256=_OwFZh5SePDUD70XIK5kngv3we_Z5-YdGHNfpUSatSQ,19469
-lamindb/core/_run_context.py,sha256=xcsPhpabm2a-PnwH-vDJ4Mx4qQLdFsn4ZUpWMWKZgoM,17472
+lamindb/core/_run_context.py,sha256=tqKPNkryy4yc7vtYSIfGjUu_pJSBQt1Kx8Cbq9vwXK8,17726
 lamindb/core/_settings.py,sha256=r9si7wJb31tI4vfz9dUN4iXe6QQU7FjnqAEsHy2UDzM,5727
 lamindb/core/_sync_git.py,sha256=IlTqw55inPp_RZbN_YScaCeKza7LeF9mClQw55W3_d4,3921
 lamindb/core/_track_environment.py,sha256=xLZ6kgzxWS6MWZ5LQ_wkbJX99vmYOT8iQ-Fz4OHCgWw,754
@@ -42,14 +42,14 @@ lamindb/core/storage/__init__.py,sha256=6jnbFj-eBV3xZt04qP-kTsMWoP8YwpM50wlnnxDY
 lamindb/core/storage/_anndata_sizes.py,sha256=aXO3OB--tF5MChenSsigW6Q-RuE8YJJOUTVukkLrv9A,1029
 lamindb/core/storage/_backed_access.py,sha256=eManrLsu3pSSQAyAKy47FDBm-iHgjaNfHA-zLy59uDs,24536
 lamindb/core/storage/_valid_suffixes.py,sha256=sewRRU3I6fJ-Jd5ACNcco_o3hic9zmqTs8BuZui-450,133
-lamindb/core/storage/_zarr.py,sha256=0i9-cJPjieIsp5UpK-IyRPkHAF-iKkWgpkWviSni2MM,2900
+lamindb/core/storage/_zarr.py,sha256=5ceEz6YIvgvUnVVNWhK5Z4W0WfrvyvY82Yna5jSX1_E,3661
 lamindb/core/storage/objects.py,sha256=5LbBeZVKuOOB8DceSE-PN8elKY0N9OhFXZPQJE4lK48,1538
-lamindb/core/storage/paths.py,sha256=XWfSHK5b3_TFiK-IMvH-srvxO0bZStzA_rwjNaTxQU4,7725
+lamindb/core/storage/paths.py,sha256=kvu4Xi4dvreXpg4iuskN_nd2yyGmEdCmoIfi3nCrTyo,7728
 lamindb/integrations/__init__.py,sha256=aH2PmO2m4-vwIifMYTB0Fyyr_gZWtVnV71jT0tVWSw0,123
 lamindb/integrations/_vitessce.py,sha256=Ii2YhGwXH_tNDS9MXzxNekthWoDmDGpgGxAOVcTIbB4,2550
 lamindb/setup/__init__.py,sha256=OwZpZzPDv5lPPGXZP7-zK6UdO4FHvvuBh439yZvIp3A,410
 lamindb/setup/core/__init__.py,sha256=SevlVrc2AZWL3uALbE5sopxBnIZPWZ1IB0NBDudiAL8,167
-lamindb-0.70.2.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-lamindb-0.70.2.dist-info/WHEEL,sha256=EZbGkh7Ie4PoZfRQ8I0ZuP9VklN_TvcZ6DSE5Uar4z4,81
-lamindb-0.70.2.dist-info/METADATA,sha256=KqXBjKMhKMBbx7VYeaWWfP0Xot_OyRRjpFO9vZotb7c,2835
-lamindb-0.70.2.dist-info/RECORD,,
+lamindb-0.70.3.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+lamindb-0.70.3.dist-info/WHEEL,sha256=EZbGkh7Ie4PoZfRQ8I0ZuP9VklN_TvcZ6DSE5Uar4z4,81
+lamindb-0.70.3.dist-info/METADATA,sha256=CYkk_Pk8Xqup7tDkwgd2qiuibHZqAlMG3we02vXTIV8,2835
+lamindb-0.70.3.dist-info/RECORD,,