_format_api_dates: materialize iterable inputs (Copilot #4)

Previously `time`/`datetime`/`last_modified`/`begin`/`end` were typed
as `str | Iterable[str] | None`, but the implementation used `len(...)`
and subscripting — generators and other non-Sequence iterables would
have raised at runtime, contradicting the annotation.

Add a single `list(...)` materialization line right after the str
wrap, so any iterable (pandas.Series, numpy.ndarray, generators,
sets) flows through cleanly. The half-bounded NaT/None range form
is preserved.

Verified by passing list / tuple / Series / generator / `[pd.NaT, ...]`
through and getting the expected formatted output in each case.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: thodson-usgs

dataretrieval/waterdata/utils.py

@@ -230,6 +230,10 @@ def _format_api_dates(
     # Convert single string to list for uniform processing
     if isinstance(datetime_input, str):
         datetime_input = [datetime_input]
+    elif not isinstance(datetime_input, (list, tuple)):
+        # Materialize any other iterable (pandas.Series, numpy.ndarray,
+        # generator, ...) so the len()/subscript operations below work.
+        datetime_input = list(datetime_input)
 
     # Check for null or all NA and return None
     if all(pd.isna(dt) or dt == "" or dt is None for dt in datetime_input):