chore!: require Python >=3.10; remove the deprecated variable_info metadata property (#332)

Ahead of the breaking 1.2.0 release.

Require Python >=3.10
- 3.9 support was already effectively broken: `waterdata`'s `anyio` dependency
  and the test stack require 3.10+, and the `waterdata` test modules already
  skipped on <3.10. `requires-python = ">=3.10"`; ruff `target-version = "py310"`;
  mypy `python_version = "3.10"` (drops the `mypy<2` pin that only existed to keep
  targeting 3.9); declare `anyio>=4.0` as a direct dependency (imported directly
  by `waterdata`); CI matrix `["3.10", "3.13", "3.14"]`; fill in per-version Trove
  classifiers.
- The py310 ruff target newly enables `B905` (zip `strict=`) and `UP036`
  (now-dead `sys.version_info < (3, 10)` skip guards): add explicit `strict=False`
  to the remaining `zip()` calls (`nwis`, `ogc/planning`, `waterdata/nearest`) and
  drop the dead version-skip guards in the four `waterdata` test modules.
  Behavior-identical.

Remove the deprecated `variable_info` metadata property entirely
- The `NWIS_Metadata` override only warned and returned `None` (it relied on the
  defunct `get_pmcodes`). Nothing else implemented it, so the `BaseMetadata`
  abstract is removed too — accessing `.variable_info` now raises `AttributeError`.
- `site_info` is kept: `NWIS_Metadata`/`WQP_Metadata` implement it, and the modern
  `waterdata` metadata intentionally leaves it abstract (use
  `waterdata.get_monitoring_locations()` for site descriptions).

Fix the live-API flaky-rerun marker
- A live test that hits an SSL/connection timeout surfaces as a typed
  `NetworkError`, but the `flaky_api` rerun patterns matched only the raw httpx
  exception names. `pytest-rerunfailures` matches the crash line (the
  `NetworkError`), not the chained `ConnectTimeout`, so the transient was never
  retried and failed CI. Add `NetworkError` to `_TRANSIENT_RERUN_PATTERNS`.

BREAKING CHANGE: Python 3.9 is no longer supported; `pip install
dataretrieval>=1.2.0` requires Python >=3.10. The `variable_info` metadata
property is removed (it always returned `None`); use `site_info` or
`waterdata.get_monitoring_locations()`.


Claude-Session: https://claude.ai/code/session_01Sjb14HkwuCydKSKMsaXsgd

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

Author: thodson-usgs

.github/workflows/python-package.yml

@@ -49,7 +49,7 @@ jobs:
       fail-fast: false
       matrix:
         os: [ubuntu-latest, windows-latest]
-        python-version: ["3.9", "3.13", "3.14"]
+        python-version: ["3.10", "3.13", "3.14"]
 
     steps:
       - uses: actions/checkout@v6

NEWS.md

@@ -1,3 +1,5 @@
+**06/23/2026:** **Breaking change (1.2.0):** the minimum supported Python is now **3.10** (`requires-python = ">=3.10"`). 3.9 support was already effectively broken — the `waterdata` module's dependencies (`anyio`, the test stack) require 3.10+, and the `waterdata` test modules already skipped on <3.10. `anyio` is now declared as a direct dependency (it is imported directly by `waterdata`), and the CI/ruff/mypy targets move to 3.10. Also fully removed the deprecated `variable_info` metadata property: the `NWIS_Metadata` override only warned and returned `None` (it relied on the defunct `get_pmcodes`), and the `BaseMetadata` abstract is gone too since nothing implemented it — accessing `.variable_info` now raises `AttributeError`. `site_info` is unaffected.
+
 **06/23/2026:** **Breaking change (1.2.0):** removed the `nadp` module and the deprecated `samples` module ahead of the 1.2.0 release. `nadp` was deprecated on 05/01/2026 — NADP is not a USGS data source, so retrieve NADP data directly from https://nadp.slh.wisc.edu/. The `samples.get_usgs_samples` shim (a deprecated forward to the modern getter) is gone; use `waterdata.get_samples()` instead. `import dataretrieval.nadp` / `import dataretrieval.samples` now raise `ModuleNotFoundError`.
 
 **06/03/2026:** The request-error hierarchy is now unified. Every module (`nwis`, `wqp`, `nldi`, `waterdata`, `nadp`, `streamstats`) raises a subclass of `dataretrieval.DataRetrievalError` on a failed request, so a single `except dataretrieval.DataRetrievalError` spans them all. An HTTP error status surfaces as an `HTTPError` carrying `.status_code` (inspect it to branch on a specific code); the retryable 429/5xx subset is `TransientError` (`RateLimited` / `ServiceUnavailable`, carrying `.retry_after`); and a request too large to satisfy is a `RequestTooLarge` (`URLTooLong` for an over-long single request, `Unchunkable` when the Water Data chunker cannot split a call small enough). Connection-level failures (timeouts, DNS, refused connections) are wrapped as a `NetworkError`, with the underlying `httpx` exception on `__cause__`. Every `DataRetrievalError` also exposes `.status_code` (`None` when there is no HTTP status), `.retry_after`, and `.retryable`, so a single `except dataretrieval.DataRetrievalError as e` clause can branch on the status or retry transient failures without knowing the concrete subclass. **Breaking change:** these exceptions no longer multiply-inherit a built-in — code that caught request failures with `except ValueError` or `except RuntimeError` should switch to `except dataretrieval.DataRetrievalError` (or a specific subclass). A no-data result is **not** an error: the modern getters (`waterdata`, `wqp`, `nldi`) return an empty DataFrame when nothing matches. Only the deprecated `nwis` (waterservices) path still raises `NoSitesError` on no data.

dataretrieval/nwis.py

@@ -1033,11 +1033,15 @@ def _read_json(json: dict[str, Any]) -> pd.DataFrame:
     # for example, [0, 21, 22] would be the first and last indices
     index_list = [0]
     index_list.extend(
-        [i + 1 for i, (a, b) in enumerate(zip(site_list[:-1], site_list[1:])) if a != b]
+        [
+            i + 1
+            for i, (a, b) in enumerate(zip(site_list[:-1], site_list[1:], strict=False))
+            if a != b
+        ]
     )
     index_list.append(len(site_list))
 
-    for start, end in zip(index_list[:-1], index_list[1:]):
+    for start, end in zip(index_list[:-1], index_list[1:], strict=False):
         # grab a block containing timeseries 0:21,
         # which are all from the same site
         site_block = json["value"]["timeSeries"][start:end]
@@ -1133,8 +1137,8 @@ class NWIS_Metadata(BaseMetadata):
 
     Notes
     -----
-    ``site_info`` and ``variable_info`` are exposed as properties (documented
-    below) rather than plain attributes.
+    ``site_info`` is exposed as a property (documented below) rather than a
+    plain attribute.
 
     """
 
@@ -1196,17 +1200,3 @@ def site_info(self) -> tuple[pd.DataFrame, BaseMetadata] | None:
 
         else:
             return None  # don't set metadata site_info attribute
-
-    @property
-    def variable_info(self) -> None:
-        """
-        Deprecated. Accessing variable_info via NWIS_Metadata is deprecated.
-        Returns None.
-        """
-        warnings.warn(
-            "Accessing variable_info via NWIS_Metadata is deprecated as "
-            "it relies on the defunct get_pmcodes function.",
-            DeprecationWarning,
-            stacklevel=2,
-        )
-        return None

dataretrieval/ogc/planning.py

@@ -497,7 +497,7 @@ def iter_sub_args(self) -> Iterator[dict[str, Any]]:
         chunk_lists = [self.chunks[ax.arg_key] for ax in self.axes]
         for combo in itertools.product(*chunk_lists):
             sub_args = dict(self.args)
-            for axis, chunk in zip(self.axes, combo):
+            for axis, chunk in zip(self.axes, combo, strict=False):
                 sub_args[axis.arg_key] = axis.render(chunk)
             yield sub_args
 

dataretrieval/utils.py

@@ -252,19 +252,15 @@ def __init__(self, response: httpx.Response) -> None:
         # # disclaimer seems to be only part of importWaterML1
         # self.disclaimer = None
 
-    # These properties are to be set by `nwis` or `wqp`-specific metadata classes.
+    # ``site_info`` is set by ``nwis`` / ``wqp``-specific metadata classes; the
+    # modern ``waterdata`` metadata leaves it unimplemented (use
+    # ``waterdata.get_monitoring_locations`` to retrieve site descriptions).
     @property
     def site_info(self) -> Any:
         raise NotImplementedError(
             "site_info must be implemented by utils.BaseMetadata children"
         )
 
-    @property
-    def variable_info(self) -> Any:
-        raise NotImplementedError(
-            "variable_info must be implemented by utils.BaseMetadata children"
-        )
-
     def __repr__(self) -> str:
         return f"{type(self).__name__}(url={self.url})"
 

dataretrieval/waterdata/nearest.py

@@ -197,7 +197,8 @@ def _build_window_or_filter(targets: pd.DatetimeIndex, window_td: pd.Timedelta)
     lowers = (targets - window_td).strftime(fmt)
     uppers = (targets + window_td).strftime(fmt)
     return " OR ".join(
-        f"(time >= '{lo}' AND time <= '{up}')" for lo, up in zip(lowers, uppers)
+        f"(time >= '{lo}' AND time <= '{up}')"
+        for lo, up in zip(lowers, uppers, strict=False)
     )
 
 

pyproject.toml

@@ -6,7 +6,7 @@ build-backend = "setuptools.build_meta"
 name = "dataretrieval"
 description = "Discover and retrieve water data from U.S. federal hydrologic web services."
 readme = "README.md"
-requires-python = ">=3.9"
+requires-python = ">=3.10"
 keywords = ["USGS", "water data"]
 license = "CC0-1.0"
 license-files = ["LICENSE.md"]
@@ -17,11 +17,22 @@ maintainers = [
   {name = "Elise Hinman", email = "ehinman@usgs.gov"},
 ]
 classifiers = [
+    "Development Status :: 5 - Production/Stable",
+    "Intended Audience :: Science/Research",
+    "Topic :: Scientific/Engineering :: Hydrology",
     "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
+    "Programming Language :: Python :: 3.14",
 ]
 dependencies = [
     "httpx",
     "pandas>=2.0.0,<4.0.0",
+    # Directly imported by ``waterdata`` (``anyio.from_thread.start_blocking_portal``),
+    # so declared here rather than relied on transitively via httpx.
+    "anyio>=4.0",
 ]
 dynamic = ["version"]
 
@@ -35,7 +46,7 @@ dataretrieval = ["py.typed"]
 # Minimal set the CI ``type-check`` job installs — just mypy + the package,
 # not the whole test stack.
 type-check = [
-  "mypy<2",  # <2 so it can still target Python 3.9 (the project's floor)
+  "mypy",
 ]
 test = [
   "pytest > 5.0.0",
@@ -77,7 +88,7 @@ repository = "https://github.com/DOI-USGS/dataretrieval-python.git"
 write_to = "dataretrieval/_version.py"
 
 [tool.ruff]
-target-version = "py39"
+target-version = "py310"
 extend-exclude = ["demos"]
 
 [tool.ruff.lint]
@@ -115,7 +126,7 @@ docstring-code-line-length = 72
 # libraries (pandas, geopandas) are treated as ``Any`` instead of
 # requiring stub packages. Dropping that — via pandas-stubs/types-requests and
 # per-module overrides — can follow.
-python_version = "3.9"  # the project's minimum supported version
+python_version = "3.10"  # the project's minimum supported version
 files = ["dataretrieval"]
 strict = true
 ignore_missing_imports = true

tests/conftest.py

@@ -26,6 +26,11 @@
     r"(?:QuotaExhausted|ServiceInterrupted):",
     r"Connect(ion)?Error",  # requests' ConnectionError + httpx' ConnectError
     r"ReadTimeout|ConnectTimeout|Timeout",
+    # ``dataretrieval`` wraps connection-level failures (timeout / DNS / refused)
+    # in a typed ``NetworkError``; rerunfailures matches the crash line (the
+    # ``NetworkError``), not the chained raw httpx exception, so match the
+    # wrapper too -- otherwise a transient SSL/handshake timeout fails CI.
+    r"NetworkError",
 ]
 
 #: Apply to a test module (``pytestmark = flaky_api``) or class (``@flaky_api``)

tests/ngwmn_test.py

@@ -6,14 +6,8 @@
 behavior change still fails on the first run.
 """
 
-import sys
-
-import pytest
 from pandas import DataFrame
 
-if sys.version_info < (3, 10):
-    pytest.skip("Skip entire module on Python < 3.10", allow_module_level=True)
-
 from dataretrieval import ngwmn
 from dataretrieval.utils import BaseMetadata
 from tests.conftest import flaky_api

tests/nwis_test.py

@@ -359,7 +359,7 @@ class TestMetaData:
     -----
 
     - Originally based on GitHub Issue #73.
-    - Modified to site_info and variable_info as properties, not callables.
+    - Modified to expose site_info as a property, not a callable.
     """
 
     def test_set_metadata_info_site(self):
@@ -416,17 +416,6 @@ def test_set_metadata_info_countyCd(self):
         # assert that site_info is implemented
         assert md.site_info
 
-    def test_variable_info_deprecated(self):
-        """Test that variable_info raises a DeprecationWarning and returns None."""
-        response = mock.MagicMock()
-        md = NWIS_Metadata(response)
-        with pytest.warns(
-            DeprecationWarning,
-            match="Accessing variable_info via NWIS_Metadata is deprecated",
-        ):
-            result = md.variable_info
-        assert result is None
-
 
 class TestReadRdb:
     """Tests for the NWIS-specific _read_rdb wrapper.