Burn down the coord, protobuf and forwarding-mismatch diagnostics

dascore/core/attrs.py

@@ -3,7 +3,7 @@
 from __future__ import annotations
 
 from collections.abc import Mapping
-from typing import Annotated, Any
+from typing import Annotated, Any, cast
 
 from pydantic import ConfigDict, Field, PlainValidator, model_validator
 from typing_extensions import Self
@@ -141,7 +141,9 @@ def from_dict(
         if isinstance(out, Mapping):
             out = dict(out)
             out.pop("dims", None)
-        return cls(**out)
+        # Anything else may still be unpackable -- a pandas Series, say --
+        # and the constructor has always been what rejects the rest.
+        return cls(**cast("Mapping[str, Any]", out))
 
     def update(self, **kwargs) -> Self:
         """Update an attribute in the model, return new model."""

dascore/core/coordmanager.py

@@ -425,7 +425,7 @@ def snap(
         assert out.shape == self.shape
         return out, array
 
-    def new(self, dims=None, coord_map=None, dim_map=None) -> Self:
+    def new(self, dims=None, coord_map=None, dim_map=None, **kwargs) -> Self:
         """
         Return a new coordmanager with specified attributes replaced.
 
@@ -442,12 +442,13 @@ def new(self, dims=None, coord_map=None, dim_map=None) -> Self:
             dims=dims if dims is not None else self.dims,
             coord_map=coord_map if coord_map is not None else self.coord_map,
             dim_map=dim_map if dim_map is not None else self.dim_map,
+            **kwargs,
         )
         return out
 
     def drop_coords(
         self,
-        *coords: str | Sequence[str],
+        *coords: str,
         array: MaybeArray = None,
     ) -> tuple[Self, MaybeArray]:
         """
@@ -1133,7 +1134,7 @@ def _flip_coord(coord, axis):
 
 def get_coord_manager(
     coords: CoordManagerInput | CoordManager | None = None,
-    dims: tuple[str, ...] | None = None,
+    dims: Sequence[str] | None = None,
     shape=None,
 ) -> CoordManager:
     """
@@ -1175,6 +1176,8 @@ def get_coord_manager(
     >>> cm = get_coord_manager(coords=coords, dims=dims)
     """
     # return coords if we already have a coord manager.
+    # A list of dims would never compare equal to a CoordManager's tuple.
+    dims = None if dims is None else tuple(dims)
     if isinstance(coords, CoordManager):
         # maybe try to rename dims.
         if dims is not None and dims != coords.dims:

dascore/core/coords.py

@@ -341,13 +341,13 @@ def check_time_units(cls, data: Any) -> Any:
 
     @field_validator("shape", mode="before")
     @classmethod
-    def _validate_nullish_to_nan(cls, value):
+    def _validate_shape_to_tuple(cls, value):
         """Ensure shape is a tuple."""
         # This also allows shape to be an int.
         return tuple(iterate(value))
 
     @abc.abstractmethod
-    def convert_units(self, unit) -> Self:
+    def convert_units(self, units) -> Self:
         """Convert from one unit to another. Set units if None are set."""
 
     def _get_value_index(self, coord_array, values_to_find):
@@ -433,16 +433,16 @@ def _select_by_samples(self, arg):
 
     @abc.abstractmethod
     def select(
-        self, arg, relative=False, samples=False
-    ) -> tuple[Self, slice | ArrayLike]:
+        self, args, relative=False, samples=False
+    ) -> tuple[BaseCoord, slice | ArrayLike]:
         """
         Returns an entity that can be used in a list for numpy indexing
         and selected coord.
         """
 
     def order(
         self, array, relative=False, samples=False
-    ) -> tuple[Self, slice | ArrayLike]:
+    ) -> tuple[BaseCoord, slice | ArrayLike]:
         """
         Order coordinate according to array values or samples.
 
@@ -700,7 +700,7 @@ def coord_range(self, extend: bool = True):
     def sort(self, reverse=False) -> tuple[BaseCoord, slice | ArrayLike]:
         """Sort the contents of the coord. Return new coord and slice for sorting."""
 
-    def snap(self) -> CoordRange:
+    def snap(self) -> BaseCoord:
         """
         Snap the coordinates to evenly sampled grid points.
 
@@ -766,7 +766,7 @@ def get_discontinuities(self, kind="all", tolerance=None) -> pd.DataFrame:
         return pd.DataFrame(columns=columns)
 
     @abc.abstractmethod
-    def update_limits(self, min=None, max=None, step=None, **kwargs) -> Self:
+    def update_limits(self, min=None, max=None, step=None, **kwargs) -> BaseCoord:
         """
         Update the limits or sampling of the coordinates.
 
@@ -796,7 +796,7 @@ def update_data(
         data: ArrayLike | np.ndarray | None = None,
         values: ArrayLike | np.ndarray | None = None,
         **kwargs,
-    ) -> Self:
+    ) -> BaseCoord:
         """
         Update the data of the coordinate.
 
@@ -903,7 +903,7 @@ def _get_relative_values(self, value):
             out = self.min() + value if pos else self.max() + value
         return out
 
-    def empty(self, axes=None) -> Self:
+    def empty(self, axes=None) -> BaseCoord:
         """
         Empty out the coordinate.
 
@@ -922,7 +922,7 @@ def empty(self, axes=None) -> Self:
         data = np.empty(tuple(new_shape), dtype=self.dtype)
         return get_coord(data=data)
 
-    def index(self, indexer, axis: int | None = None) -> Self:
+    def index(self, indexer, axis: int | None = None) -> BaseCoord:
         """
         Index the coordinate and return new coordinate.
 
@@ -1240,9 +1240,15 @@ def update(self, **kwargs):
         """No values to change so update can just call new."""
         return self.new(**kwargs)
 
-    # Other operations that normally modify data do not in this case.
-    update_limits = update
-    set_units = update
+    # update_limits is spelled out rather than aliased to update so it keeps
+    # the signature its base declares. It must forward only what the caller
+    # supplied: a None reaching _validate_nullish_to_nan would overwrite the
+    # stored start, stop or step with nan.
+    def update_limits(self, min=None, max=None, step=None, **kwargs) -> BaseCoord:
+        """No values to limit, so only what was passed is applied."""
+        limits = {"min": min, "max": max, "step": step}
+        passed = {i: v for i, v in limits.items() if v is not None}
+        return self.update(**passed, **kwargs)
 
     def convert_units(self, units) -> Self:
         """Convert scalar metadata units, or set units if none exist."""
@@ -1307,7 +1313,7 @@ def select(
     @compose_docstring(doc=get_docstring(BaseCoord.order))
     def order(
         self, array, relative=False, samples=False
-    ) -> tuple[Self, slice | ArrayLike]:
+    ) -> tuple[BaseCoord, slice | ArrayLike]:
         """
         {doc}.
         """
@@ -1616,7 +1622,7 @@ def _get_index(self, value, forward=True):
         return fraction.astype(np.int64)
 
     @compose_docstring(doc=get_docstring(BaseCoord.update_limits))
-    def update_limits(self, min=None, max=None, step=None, **kwargs) -> Self:
+    def update_limits(self, min=None, max=None, step=None, **kwargs) -> BaseCoord:
         """{doc}."""
         if all(x is not None for x in [min, max, step]):
             msg = "At most two parameters can be specified in update_limits."
@@ -1730,7 +1736,7 @@ def convert_units(self, units) -> Self:
 
     def select(
         self, args, relative=False, samples=False
-    ) -> tuple[Self, slice | ArrayLike]:
+    ) -> tuple[BaseCoord, slice | ArrayLike]:
         """Apply select, return selected coords and index for selecting data."""
         if is_array(args):
             return self._select_by_array(args, relative=relative, samples=samples)
@@ -1800,7 +1806,7 @@ def snap(self):
         return out.change_length(len(self))
 
     @compose_docstring(doc=get_docstring(BaseCoord.update_limits))
-    def update_limits(self, min=None, max=None, step=None, **kwargs) -> Self:
+    def update_limits(self, min=None, max=None, step=None, **kwargs) -> BaseCoord:
         """{doc}."""
         if sum(x is not None for x in [min, max, step]) > 1:
             msg = "At most one parameter can be specified in update_limits."
@@ -1863,7 +1869,7 @@ class CoordMonotonicArray(CoordArray):
 
     def select(
         self, args, relative=False, samples=False
-    ) -> tuple[Self, slice | ArrayLike]:
+    ) -> tuple[BaseCoord, slice | ArrayLike]:
         """Apply select, return selected coords and index for selecting data."""
         if is_array(args):
             return self._select_by_array(args, relative=relative, samples=samples)
@@ -2332,7 +2338,7 @@ def sort(self, reverse=False) -> tuple[BaseCoord, slice | ArrayLike]:
         return self.new(segments=segments), slice(None, None, -1)
 
     @compose_docstring(doc=get_docstring(BaseCoord.update_limits))
-    def update_limits(self, min=None, max=None, step=None, **kwargs) -> Self:
+    def update_limits(self, min=None, max=None, step=None, **kwargs) -> BaseCoord:
         """{doc}."""
         if step is not None:
             msg = (
@@ -2708,9 +2714,9 @@ def _validate_values(cls, values):
         values["step"] = None
         return values
 
-    def convert_units(self, unit) -> Self:
+    def convert_units(self, units) -> Self:
         """String coordinates cannot be converted between units."""
-        if unit not in (None, ""):
+        if units not in (None, ""):
             _raise_string_coord_error("unit conversion")
         return self
 
@@ -2730,7 +2736,7 @@ def _get_compatible_value(self, value, relative=False):
 
     def select(
         self, args, relative=False, samples=False
-    ) -> tuple[Self, slice | ArrayLike]:
+    ) -> tuple[BaseCoord, slice | ArrayLike]:
         """Select by exact values, wildcard patterns, regexes, samples, or masks."""
         if relative:
             _raise_string_coord_error("relative selection")
@@ -2779,7 +2785,7 @@ def reverse_sorted(self) -> bool:
             return False
         return bool(np.all(values[:-1] >= values[1:]))
 
-    def update_limits(self, min=None, max=None, step=None, **kwargs) -> Self:
+    def update_limits(self, min=None, max=None, step=None, **kwargs) -> BaseCoord:
         """Reject numeric limit updates on string coords."""
         # Deliberately match BaseCoord/CoordRange parameter names for API parity.
         unsupported_kwargs = set(kwargs) - {"data"}

dascore/core/patch.py

@@ -4,7 +4,7 @@
 
 from collections.abc import Mapping, Sequence
 from functools import cached_property
-from typing import Final
+from typing import Any, Final
 from uuid import uuid4
 
 import numpy as np
@@ -17,7 +17,6 @@
 from dascore.compat import DataArray, array
 from dascore.core.attrs import PatchAttrs
 from dascore.core.coordmanager import CoordManager, get_coord_manager
-from dascore.core.coords import BaseCoord
 from dascore.core.summary import PatchSummary
 from dascore.utils.array import (
     PatchUFunc,
@@ -77,7 +76,7 @@ class Patch(NamespaceOwner):
     def __init__(
         self,
         data: ArrayLike | DataArray | None = None,
-        coords: Mapping[str, ArrayLike | BaseCoord] | CoordManager | None = None,
+        coords: Mapping[str, Any] | CoordManager | None = None,
         dims: Sequence[str] | None = None,
         attrs: Mapping | PatchAttrs | None = None,
     ):

dascore/core/summary.py

@@ -84,7 +84,7 @@ def _flatten_coord_summary(
     exclude: set[str] | None = None,
 ) -> dict[str, Any]:
     """Flatten a single coord summary into scan/index-style fields."""
-    exclude = set() if exclude is None else exclude
+    exclude = set[str]() if exclude is None else exclude
     summary_dict = _coord_summary_to_dict(summary)
     out = {}
     if dim_tuple and coord_name not in exclude:
@@ -261,7 +261,7 @@ def dump_structured(self) -> dict[str, Any]:
 
     def flat_dump(self, dim_tuple: bool = False, exclude=None) -> dict[str, Any]:
         """Return a flat dict suitable for indexing/dataframes."""
-        exclude = set(() if exclude is None else exclude)
+        exclude = set[str](() if exclude is None else exclude)
         # Build flattened attrs first, then overlay coord summaries so coord-
         # derived fields win over any attr using the same simplified key.
         out = self.attrs.flat_dump(exclude=exclude)

dascore/io/core.py

@@ -694,7 +694,7 @@ def _yield_extensions(self, extension, input_type=None):
 
     def _get_format(
         self,
-        path: str | Path | IOResourceManager,
+        path: path_types | IOResourceManager,
         file_format: str | None = None,
         file_version: str | None = None,
         fiber_io_hint: dict[str, FiberIO] | None = None,

dascore/io/index/indexer.py

@@ -33,7 +33,11 @@
 )
 from dascore.io.index.schema import SPOOL_HIDDEN_COLUMNS
 from dascore.utils.misc import _iter_filesystem
-from dascore.utils.paths import directory_writable, requires_local_directory
+from dascore.utils.paths import (
+    coerce_to_local_path,
+    directory_writable,
+    requires_local_directory,
+)
 
 
 def _path_digest(path) -> str:
@@ -132,7 +136,7 @@ def __init__(
     ):
         path = UPath(path).absolute() if isinstance(path, UPath) else Path(path)
         requires_local_directory(path, label="DBDirectoryIndexer")
-        self.path = Path(path).absolute()
+        self.path = Path(coerce_to_local_path(path)).absolute()
         self.index_path = Path(self._find_index_path(index_path))
         try:
             self._backend = get_backend(self.index_path)

dascore/io/index/planned.py

@@ -330,7 +330,7 @@ def __init__(
         # informational only: the directory/file the plan derived from
         self.origin_path = origin_path
 
-    def live_entries(self) -> Mapping[str, dc.Patch]:
+    def live_entries(self) -> dict[str, dc.Patch]:
         """Expose the loader's live registry (for absorption/transfer)."""
         return self.loader.live_entries()
 

dascore/io/sintela/protobuf_utils.py

@@ -43,7 +43,7 @@
 from __future__ import annotations
 
 import struct
-from collections.abc import Iterator
+from collections.abc import Iterable, Iterator
 from functools import cache
 from typing import Any
 
@@ -464,7 +464,7 @@ def _common_header_time(common_header) -> np.datetime64 | None:
 
 
 def _parse_records(
-    records: list[EnvelopeRecord], *, scan_mode: bool = False
+    records: Iterable[EnvelopeRecord], *, scan_mode: bool = False
 ) -> tuple[list[Any], ParsedMeta]:
     """Decode protobuf payloads and return messages plus selected META."""
     messages = _get_proto_messages(include_sample_fields=not scan_mode)
@@ -530,8 +530,14 @@ def _get_distance_coord(start_channel: int, spacing: float, count: int, step: in
     )
 
 
-def _get_times(times: list[np.datetime64]):
-    """Build a time coordinate from packet timestamps."""
+def _get_times(times: list[np.datetime64 | None]):
+    """
+    Build a time coordinate from packet timestamps.
+
+    Callers reject a packet with no header time before getting here; the
+    None in the signature is what the list comprehension produces, not a
+    supported input.
+    """
     return get_coord(data=np.asarray(times, dtype="datetime64[ns]"))
 
 
@@ -557,7 +563,7 @@ def _base_attrs(
     Each packet family supplies its own ``data_type``/``data_units`` via
     ``extra``; the fields below are shared across all families.
     """
-    attrs = dict(
+    attrs = SintelaProtobufAttrs(
         data_category="DAS",
         packet_type=packet_type,
         recorder_namespace=meta.recorder_namespace,
@@ -572,9 +578,7 @@ def _base_attrs(
         start_channel=int(getattr(common_header, "start_channel", 0)),
         channel_step=None,
     )
-    if extra:
-        attrs.update(extra)
-    return SintelaProtobufAttrs(**attrs)
+    return attrs.new(**extra) if extra else attrs
 
 
 def _get_band_attr_data_type(band_def: tuple[tuple[Any, ...], ...]) -> tuple[str, str]:

dascore/io/terra15/utils.py

@@ -179,8 +179,7 @@ def _get_time_coord(data_node, snap_dims=True):
     """Get the time coordinate."""
     t_min, t_max, _time_len, d_time = _get_scanned_time_info(data_node)
     if snap_dims:
-        kwargs = dict(start=t_min, stop=t_max + d_time, step=d_time, units="s")
-        time_coord = get_coord(**kwargs)
+        time_coord = get_coord(start=t_min, stop=t_max + d_time, step=d_time, units="s")
     else:
         time_coord = _get_raw_time_coord(data_node)
     return time_coord