DEPS/TST: tzdata is optional, not required

ci/deps/actions-310.yaml

@@ -47,6 +47,7 @@ dependencies:
   - scipy
   - sqlalchemy
   - tabulate
+  - tzdata>=2022a
   - xarray
   - xlrd
   - xlsxwriter

ci/deps/actions-38-minimum_versions.yaml

@@ -49,6 +49,7 @@ dependencies:
   - scipy=1.7.1
   - sqlalchemy=1.4.16
   - tabulate=0.8.9
+  - tzdata=2022a
   - xarray=0.19.0
   - xlrd=2.0.1
   - xlsxwriter=1.4.3

ci/deps/actions-39.yaml

@@ -47,6 +47,7 @@ dependencies:
   - scipy
   - sqlalchemy
   - tabulate
+  - tzdata>=2022a
   - xarray
   - xlrd
   - xlsxwriter

doc/source/getting_started/install.rst

@@ -270,6 +270,23 @@ For example, :func:`pandas.read_hdf` requires the ``pytables`` package, while
 optional dependency is not installed, pandas will raise an ``ImportError`` when
 the method requiring that dependency is called.
 
+Timezones
+^^^^^^^^^
+
+========================= ========================= =============================================================
+Dependency                Minimum Version           Notes
+========================= ========================= =============================================================
+tzdata                    2022.1(pypi)/             Allows the use of ``zoneinfo`` timezones with pandas.
+                          2022a(for system tzdata)  **Note**: You only need to install the pypi package if your
+                                                    system does not already provide the IANA tz database.
+                                                    However, the minimum tzdata version still applies, even if it
+                                                    is not enforced through an error.
+
+                                                    If you would like to keep your system tzdata version updated,
+                                                    it is recommended to use the ``tzdata`` package from
+                                                    conda-forge.
+========================= ========================= =============================================================
+
 Visualization
 ^^^^^^^^^^^^^
 

environment.yml

@@ -48,6 +48,7 @@ dependencies:
   - scipy
   - sqlalchemy
   - tabulate
+  - tzdata>=2022a
   - xarray
   - xlrd
   - xlsxwriter

pandas/_libs/tslibs/timezones.pyx

@@ -3,6 +3,8 @@ from datetime import (
     timezone,
 )
 
+from pandas.compat._optional import import_optional_dependency
+
 try:
     # py39+
     import zoneinfo
@@ -67,6 +69,9 @@ cdef inline bint is_utc_zoneinfo(tzinfo tz):
             utc_zoneinfo = ZoneInfo("UTC")
         except zoneinfo.ZoneInfoNotFoundError:
             return False
+        # Warn if tzdata is too old, even if there is a system tzdata to alert
+        # users about the mismatch between local/system tzdata
+        import_optional_dependency("tzdata", errors="warn", min_version="2022.1")
 
     return tz is utc_zoneinfo
 

pandas/compat/_optional.py

@@ -44,6 +44,7 @@
     "xlwt": "1.3.0",
     "xlsxwriter": "1.4.3",
     "zstandard": "0.15.2",
+    "tzdata": "2022.1",
 }
 
 # A mapping from import name to package name (on PyPI) for packages where

pandas/conftest.py

@@ -83,6 +83,14 @@
     # Import "zoneinfo" could not be resolved (reportMissingImports)
     import zoneinfo  # type: ignore[no-redef]
 
+    # Although zoneinfo can be imported in Py39, it is effectively
+    # "not available" without tzdata/IANA tz data.
+    # We will set zoneinfo to not found in this case
+    try:
+        zoneinfo.ZoneInfo("UTC")  # type: ignore[attr-defined]
+    except zoneinfo.ZoneInfoNotFoundError:  # type: ignore[attr-defined]
+        zoneinfo = None
+
 # Until https://github.com/numpy/numpy/issues/19078 is sorted out, just suppress
 suppress_npdev_promotion_warning = pytest.mark.filterwarnings(
     "ignore:Promotion of numbers and bools:FutureWarning"

requirements-dev.txt

@@ -39,6 +39,7 @@ s3fs
 scipy
 sqlalchemy
 tabulate
+tzdata>=2022.1
 xarray
 xlrd
 xlsxwriter

scripts/generate_pip_deps_from_conda.py

@@ -21,6 +21,7 @@
 import yaml
 
 EXCLUDE = {"python", "c-compiler", "cxx-compiler"}
+REMAP_VERSION = {"tzdata": "2022.1"}
 RENAME = {"pytables": "tables", "geopandas-base": "geopandas", "pytorch": "torch"}
 
 
@@ -41,7 +42,8 @@ def conda_package_to_pip(package: str):
             pkg, version = package.split(compare)
             if pkg in EXCLUDE:
                 return
-
+            if pkg in REMAP_VERSION:
+                return "".join((pkg, compare, REMAP_VERSION[pkg]))
             if pkg in RENAME:
                 return "".join((RENAME[pkg], compare, version))
 

scripts/validate_min_versions_in_sync.py

@@ -21,6 +21,7 @@
     pathlib.Path("ci/deps").absolute().glob("actions-*-minimum_versions.yaml")
 )
 CODE_PATH = pathlib.Path("pandas/compat/_optional.py").resolve()
+EXCLUDE_DEPS = {"tzdata"}
 # pandas package is not available
 # in pre-commit environment
 sys.path.append("pandas/compat")
@@ -34,6 +35,8 @@
 def get_versions_from_code() -> dict[str, str]:
     install_map = _optional.INSTALL_MAPPING
     versions = _optional.VERSIONS
+    for item in EXCLUDE_DEPS:
+        versions.pop(item)
     return {
         install_map.get(k, k).casefold(): v
         for k, v in versions.items()
@@ -55,6 +58,8 @@ def get_versions_from_ci(content: list[str]) -> tuple[dict[str, str], dict[str,
         elif seen_required and line.strip():
             package, version = line.strip().split("=")
             package = package[2:]
+            if package in EXCLUDE_DEPS:
+                continue
             if not seen_optional:
                 required_deps[package] = version
             else: