smftools 0.2.3__py3-none-any.whl → 0.2.5__py3-none-any.whl

smftools/constants.py

@@ -0,0 +1,37 @@
+from __future__ import annotations
+
+from types import MappingProxyType
+from typing import Any, Dict, Final, Mapping
+
+
+## Helpers ##
+def _deep_freeze(obj: Any) -> Any:
+    """Recursively freeze common containers. Use for constant exports."""
+    if isinstance(obj, dict):
+        return MappingProxyType({k: _deep_freeze(v) for k, v in obj.items()})
+    if isinstance(obj, (list, tuple)):
+        return tuple(_deep_freeze(v) for v in obj)
+    if isinstance(obj, set):
+        return frozenset(_deep_freeze(v) for v in obj)
+    return obj  # ints/strs/tuples (already immutable)
+
+
+## Constants ##
+BAM_SUFFIX: Final[str] = ".bam"
+BARCODE_BOTH_ENDS: Final[bool] = False
+REF_COL: Final[str] = "Reference_strand"
+SAMPLE_COL: Final[str] = "Experiment_name_and_barcode"
+SPLIT_DIR: Final[str] = "demultiplexed_BAMs"
+TRIM: Final[bool] = False
+
+_private_conversions = ["unconverted"]
+CONVERSIONS: Final[list[str]] = _deep_freeze(_private_conversions)
+
+_private_mod_list = ("5mC_5hmC", "6mA")
+MOD_LIST: Final[tuple[str, ...]] = _deep_freeze(_private_mod_list)
+
+_private_mod_map: Dict[str, str] = {"6mA": "6mA", "5mC_5hmC": "5mC"}
+MOD_MAP: Final[Mapping[str, str]] = _deep_freeze(_private_mod_map)
+
+_private_strands = ("bottom", "top")
+STRANDS: Final[tuple[str, ...]] = _deep_freeze(_private_strands)

smftools/datasets/__init__.py

@@ -1,9 +1,3 @@
-from .datasets import (
-    dCas9_kinetics,
-    Kissiov_and_McKenna_2025
-)
+from .datasets import Kissiov_and_McKenna_2025, dCas9_kinetics
 
-__all__ = [
-    "dCas9_kinetics",
-    "Kissiov_and_McKenna_2025"
-]
+__all__ = ["dCas9_kinetics", "Kissiov_and_McKenna_2025"]

smftools/datasets/datasets.py

@@ -1,28 +1,42 @@
-## datasets
+"""Dataset helpers for bundled SMF datasets."""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    import anndata as ad
 
-def import_HERE():
-    """
-    Imports HERE for loading datasets
-    """
-    from pathlib import Path
-    from .._settings import settings
-    HERE = Path(__file__).parent
-    return HERE
 
-def dCas9_kinetics():
+def import_HERE() -> Path:
+    """Resolve the local dataset directory.
+
+    Returns:
+        Path: Path to the datasets directory.
     """
-    in vitro Hia5 dCas9 kinetics SMF dataset. Nanopore HAC m6A modcalls.
+    return Path(__file__).parent
+
+
+def dCas9_kinetics() -> "ad.AnnData":
+    """Load the in vitro Hia5 dCas9 kinetics SMF dataset.
+
+    Returns:
+        anndata.AnnData: Annotated dataset with Nanopore HAC m6A modcalls.
     """
     import anndata as ad
-    HERE = import_HERE()
-    filepath = HERE / "dCas9_m6A_invitro_kinetics.h5ad.gz"
+
+    filepath = import_HERE() / "dCas9_m6A_invitro_kinetics.h5ad.gz"
     return ad.read_h5ad(filepath)
 
-def Kissiov_and_McKenna_2025():
-    """
-    F1 Hybrid M.CviPI natural killer cell SMF. Nanopore canonical calls of NEB EMseq converted SMF gDNA.
+
+def Kissiov_and_McKenna_2025() -> "ad.AnnData":
+    """Load the F1 Hybrid M.CviPI natural killer cell SMF dataset.
+
+    Returns:
+        anndata.AnnData: Annotated dataset with canonical calls of NEB EMseq converted SMF gDNA.
     """
     import anndata as ad
-    HERE = import_HERE()
-    filepath = HERE / "F1_hybrid_NKG2A_enhander_promoter_GpC_conversion_SMF.h5ad.gz"
+
+    filepath = import_HERE() / "F1_hybrid_NKG2A_enhander_promoter_GpC_conversion_SMF.h5ad.gz"
     return ad.read_h5ad(filepath)