lamindb 0.77.3__py3-none-any.whl → 1.0rc1__py3-none-any.whl

lamindb/base/fields.py

@@ -0,0 +1,281 @@
+from django.db import models
+
+
+class CharField(models.CharField):
+    """Custom `CharField` with default values for `blank`, `default`, and `max_length`.
+
+    Django default values for `CharField` are `blank=False`, `default=""`, undefined `max_length`.
+    """
+
+    def __init__(self, max_length: int = 255, **kwargs):
+        kwargs["max_length"] = max_length  # Set max_length in kwargs
+        kwargs.setdefault("blank", True)
+        kwargs.setdefault("default", None)
+        super().__init__(**kwargs)  # Pass all arguments as kwargs
+
+
+class TextField(models.TextField):
+    """Custom `TextField` with default values for `blank` and `default`.
+
+    Django default values for `TextField` are `blank=False`, `default=''`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        kwargs.setdefault("default", None)
+        super().__init__(*args, **kwargs)
+
+
+class ForeignKey(models.ForeignKey):
+    """Custom `ForeignKey` with default values for `blank`.
+
+    Django default value for `ForeignKey` `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class BooleanField(models.BooleanField):
+    """Custom `BooleanField` with default values for `blank` and `default`.
+
+    Django default values for `BooleanField` are `blank=False`, `default=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        kwargs.setdefault("default", None)
+        super().__init__(*args, **kwargs)
+
+
+class DateField(models.DateField):
+    """Custom `DateField` with default values for `blank`.
+
+    Django default values for `DateField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class DateTimeField(models.DateTimeField):
+    """Custom `DateTimeField` with default values for `blank`.
+
+    Django default values for `DateTimeField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class BigIntegerField(models.BigIntegerField):
+    """Custom `BigIntegerField` with default values for `blank`.
+
+    Django default values for `BigIntegerField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        kwargs.setdefault("default", None)
+        super().__init__(*args, **kwargs)
+
+
+class IntegerField(models.IntegerField):
+    """Custom `IntegerField` with default values for `blank`.
+
+    Django default values for `IntegerField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class OneToOneField(models.OneToOneField):
+    """Custom `OneToOneField` with default values for `blank`.
+
+    Django default values for `OneToOneField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class FloatField(models.FloatField):
+    """Custom `FloatField` with default values for `blank`.
+
+    Django default values for `FloatField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class DecimalField(models.DecimalField):
+    """Custom `DecimalField` with default values for `blank`.
+
+    Django default values for `DecimalField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class JSONField(models.JSONField):
+    """Custom `JSONField` with default values for `blank`.
+
+    Django default values for `JSONField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class DurationField(models.DurationField):
+    """Custom `DurationField` with default values for `blank`.
+
+    Django default values for `DurationField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class URLField(models.URLField):
+    """Custom `URLField` with default values for `blank`.
+
+    Django default values for `URLField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class EmailField(models.EmailField):
+    """Custom `EmailField` with default values for `blank`.
+
+    Django default values for `EmailField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class TimeField(models.TimeField):
+    """Custom `TimeField` with default values for `blank`.
+
+    Django default values for `TimeField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class SlugField(models.SlugField):
+    """Custom `SlugField` with default values for `blank`.
+
+    Django default values for `SlugField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class UUIDField(models.UUIDField):
+    """Custom `UUIDField` with default values for `blank`.
+
+    Django default values for `UUIDField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class PositiveIntegerField(models.PositiveIntegerField):
+    """Custom `PositiveIntegerField` with default values for `blank`.
+
+    Django default values for `PositiveIntegerField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class PositiveSmallIntegerField(models.PositiveSmallIntegerField):
+    """Custom `PositiveSmallIntegerField` with default values for `blank`.
+
+    Django default values for `PositiveSmallIntegerField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class SmallIntegerField(models.SmallIntegerField):
+    """Custom `SmallIntegerField` with default values for `blank`.
+
+    Django default values for `SmallIntegerField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class BinaryField(models.BinaryField):
+    """Custom `BinaryField` with default values for `blank`.
+
+    Django default values for `BinaryField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class GenericIPAddressField(models.GenericIPAddressField):
+    """Custom `GenericIPAddressField` with default values for `blank`.
+
+    Django default values for `GenericIPAddressField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class FileField(models.FileField):
+    """Custom `FileField` with default values for `blank`.
+
+    Django default values for `FileField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)
+
+
+class ImageField(models.ImageField):
+    """Custom `ImageField` with default values for `blank`.
+
+    Django default values for `ImageField` are `blank=False`.
+    """
+
+    def __init__(self, *args, **kwargs):
+        kwargs.setdefault("blank", True)
+        super().__init__(*args, **kwargs)

lamindb/base/ids.py

@@ -0,0 +1,103 @@
+"""IDs.
+
+Base generators:
+
+.. autosummary::
+   :toctree: .
+
+   base26
+   base62
+   base64
+
+8 base62 characters:
+
+======= ===========
+n       p_collision
+======= ===========
+100k    2e-05
+1M      2e-03
+======= ===========
+
+12 base62 characters:
+
+======= ===========
+n       p_collision
+======= ===========
+100M    2e-06
+1B      2e-04
+======= ===========
+
+20 base62 characters (62**20=7e+35) roughly matches UUID (2*122=5e+36):
+
+======= ===========
+n       p_collision
+======= ===========
+3e15    1e-6
+======= ===========
+
+"""
+
+import secrets
+import string
+
+
+def base64(n_char: int) -> str:
+    """Random Base64 string."""
+    alphabet = string.digits + string.ascii_letters.swapcase() + "_" + "-"
+    id = "".join(secrets.choice(alphabet) for i in range(n_char))
+    return id
+
+
+def base62(n_char: int) -> str:
+    """Random Base62 string."""
+    alphabet = string.digits + string.ascii_letters.swapcase()
+    id = "".join(secrets.choice(alphabet) for i in range(n_char))
+    return id
+
+
+# the following cannot be serialized by Django
+# class Base62:
+#     def __init__(self, n_char: int):
+#         self.n_char = n_char
+
+#     def __call__(self):
+#         return base62(self.n_char)
+
+
+def base26(n_char: int):
+    """ASCII lowercase."""
+    alphabet = string.ascii_lowercase
+    id = "".join(secrets.choice(alphabet) for i in range(n_char))
+    return id
+
+
+def base62_4() -> str:
+    return base62(4)
+
+
+def base62_8() -> str:
+    return base62(8)
+
+
+def base62_12() -> str:
+    return base62(12)
+
+
+def base62_14() -> str:
+    return base62(14)
+
+
+def base62_16() -> str:
+    return base62(16)
+
+
+def base62_18() -> str:
+    return base62(18)
+
+
+def base62_20() -> str:
+    return base62(20)
+
+
+def base62_24() -> str:
+    return base62(24)

lamindb/base/types.py

@@ -0,0 +1,51 @@
+"""Types.
+
+Central object types.
+
+.. autosummary::
+   :toctree: .
+
+   ArtifactKind
+   TransformType
+   FeatureDtype
+
+Basic types.
+
+.. autosummary::
+   :toctree: .
+
+   UPathStr
+   StrField
+   ListLike
+   FieldAttr
+"""
+
+from __future__ import annotations
+
+from typing import Literal, Union
+
+import numpy as np
+import pandas as pd
+from django.db.models.query_utils import DeferredAttribute as FieldAttr
+from lamindb_setup.core.types import UPathStr  # noqa: F401
+
+# need to use Union because __future__.annotations doesn't do the job here <3.10
+# typing.TypeAlias, >3.10 on but already deprecated
+ListLike = Union[list[str], pd.Series, np.array]
+StrField = Union[str, FieldAttr]  # typing.TypeAlias
+
+TransformType = Literal[
+    "pipeline", "notebook", "upload", "script", "function", "linker"
+]
+ArtifactKind = Literal["dataset", "model"]
+FeatureDtype = Literal[
+    "cat",  # categorical variables
+    "num",  # numerical variables
+    "str",  # string variables
+    "int",  # integer variables
+    "float",  # float variables
+    "bool",  # boolean variables
+    "date",  # date variables
+    "datetime",  # datetime variables
+    "object",  # this is a pandas type, we're only using it for complicated types, not for strings
+]

lamindb/base/users.py

@@ -0,0 +1,30 @@
+user_id_cache = {}
+
+
+def current_user_id() -> int:
+    import lamindb_setup as ln_setup
+    from lamindb_setup import settings
+    from lamindb_setup._init_instance import register_user
+
+    from lamindb.models import User
+
+    def query_user_id():
+        if ln_setup.core.django.IS_MIGRATING:
+            return 1
+        else:
+            exc_attr = (
+                "DoesNotExist" if hasattr(User, "DoesNotExist") else "_DoesNotExist"
+            )
+            try:
+                user_id = User.objects.get(uid=settings.user.uid).id
+            except getattr(User, exc_attr):
+                register_user(settings.user)
+                user_id = User.objects.get(uid=settings.user.uid).id
+            return user_id
+
+    if settings._instance_exists:
+        if settings.instance.slug not in user_id_cache:
+            user_id_cache[settings.instance.slug] = query_user_id()
+        return user_id_cache[settings.instance.slug]
+    else:
+        return query_user_id()

lamindb/base/validation.py

@@ -0,0 +1,67 @@
+from typing import TYPE_CHECKING, Literal, Union, get_args, get_origin, get_type_hints
+
+from lamin_utils import colors
+
+if TYPE_CHECKING:
+    from .models import Record
+
+
+class FieldValidationError(SystemExit):
+    """Field validation error."""
+
+    pass
+
+
+def validate_literal_fields(record: "Record", kwargs) -> None:
+    """Validate all Literal type fields in a record.
+
+    Args:
+        record: record being validated
+
+    Raises:
+        ValidationError: If any field value is not in its Literal's allowed values
+    """
+    # check is based on string to avoid circular imports
+    if record.__class__.__name__ == "Feature":
+        # the FeatureDtype is more complicated than a simple literal
+        # because it allows constructs like cat[ULabel] etc.
+        # the User model is used at startup and throws a datetime-related error otherwise
+        # simmilar for Storage & Source
+        return None
+    try:
+        type_hints = get_type_hints(record.__class__)
+    except TypeError:
+        # for 3.9, get_type_hints errors with | in type hints
+        return
+    errors = {}
+
+    for field_name, field_type in type_hints.items():
+        # Handle both plain Literal and Union/Optional Literal types
+        origin = get_origin(field_type)
+        if origin is Union:
+            # For Optional/Union types, find the Literal type if it exists
+            literal_type = next(
+                (t for t in get_args(field_type) if get_origin(t) is Literal), None
+            )
+        else:
+            # For plain types, check if it's a Literal
+            literal_type = field_type if origin is Literal else None
+
+        # Skip if no Literal type found
+        if literal_type is None:
+            continue
+
+        value = kwargs.get(field_name)
+        if value is not None:
+            valid_values = set(get_args(literal_type))
+            if value not in valid_values:
+                errors[field_name] = (
+                    f"{field_name}: {colors.yellow(value)} is not a valid value"
+                    f"\n    → Valid values are: {colors.green(', '.join(sorted(valid_values)))}"
+                )
+
+    if errors:
+        message = "\n  "
+        for _, error in errors.items():
+            message += error + "\n  "
+        raise FieldValidationError(message)

lamindb/core/__init__.py

@@ -5,10 +5,12 @@ Registries:
 .. autosummary::
    :toctree: .
 
+   BasicRecord
    Record
    Registry
    QuerySet
    QueryManager
+   Schema
    RecordList
    FeatureManager
    ParamManager
@@ -59,7 +61,6 @@ Modules:
    loaders
    datasets
    storage
-   types
    exceptions
    subsettings
    logger
@@ -68,7 +69,21 @@ Modules:
 
 from lamin_utils import logger
 from lamin_utils._inspect import InspectResult
-from lnschema_core.models import (
+
+from lamindb._query_manager import QueryManager
+from lamindb._query_set import QuerySet, RecordList
+from lamindb.core._feature_manager import FeatureManager, ParamManager
+from lamindb.core._label_manager import LabelManager
+from lamindb.curators import (
+    AnnDataCurator,
+    BaseCurator,
+    CurateLookup,
+    DataFrameCurator,
+    MuDataCurator,
+    SOMACurator,
+)
+from lamindb.models import (
+    BasicRecord,
     CanCurate,
     FeatureValue,
     HasParents,
@@ -76,24 +91,12 @@ from lnschema_core.models import (
     ParamValue,
     Record,
     Registry,
+    Schema,
     TracksRun,
     TracksUpdates,
     ValidateFields,
 )
 
-from lamindb._curate import (
-    AnnDataCurator,
-    BaseCurator,
-    CurateLookup,
-    DataFrameCurator,
-    MuDataCurator,
-    SOMACurator,
-)
-from lamindb._query_manager import QueryManager
-from lamindb._query_set import QuerySet, RecordList
-from lamindb.core._feature_manager import FeatureManager, ParamManager
-from lamindb.core._label_manager import LabelManager
-
 from . import _data, datasets, exceptions, fields, loaders, subsettings, types
 from ._context import Context
 from ._mapped_collection import MappedCollection