apacheGH-35600: [Python] Allow setting path to timezone db through py…

…thon API (apache#37436)

### Rationale for this change

Add a function to change the path where timezone db should be found as a small wrapper around the setting of a C++ option `GlobalOptions`.

### What changes are included in this PR?

New function `configure_tzdb`.

### Are these changes tested?

### Are there any user-facing changes?

No.
* Closes: apache#35600

Lead-authored-by: AlenkaF <[email protected]>
Co-authored-by: Antoine Pitrou <[email protected]>
Co-authored-by: Alenka Frim <[email protected]>
Co-authored-by: Sutou Kouhei <[email protected]>
Signed-off-by: Joris Van den Bossche <[email protected]>

ci/appveyor-cpp-build.bat

@@ -132,6 +132,19 @@ set ARROW_HOME=%CONDA_PREFIX%\Library
 @rem ARROW-3075; pkgconfig is broken for Parquet for now
 set PARQUET_HOME=%CONDA_PREFIX%\Library
 
+@rem Download IANA Timezone Database to a non-standard location to
+@rem test the configurability of the timezone database path
+curl https://data.iana.org/time-zones/releases/tzdata2021e.tar.gz --output tzdata.tar.gz || exit /B
+mkdir %USERPROFILE%\Downloads\test\tzdata
+tar --extract --file tzdata.tar.gz --directory %USERPROFILE%\Downloads\test\tzdata
+curl https://raw.githubusercontent.com/unicode-org/cldr/master/common/supplemental/windowsZones.xml ^
+  --output %USERPROFILE%\Downloads\test\tzdata\windowsZones.xml || exit /B
+@rem Remove the database from the default location
+rmdir /s /q %USERPROFILE%\Downloads\tzdata
+@rem Set the env var for the non-standard location of the database
+@rem (only needed for testing purposes)
+set PYARROW_TZDATA_PATH=%USERPROFILE%\Downloads\test\tzdata
+
 python setup.py develop -q || exit /B
 
 set PYTHONDEVMODE=1

python/pyarrow/__init__.py

@@ -66,10 +66,10 @@ def parse_git(root, **kwargs):
 if _gc_enabled:
     _gc.enable()
 
-from pyarrow.lib import (BuildInfo, RuntimeInfo, MonthDayNano,
-                         VersionInfo, cpp_build_info, cpp_version,
-                         cpp_version_info, runtime_info, cpu_count,
-                         set_cpu_count, enable_signal_handlers,
+from pyarrow.lib import (BuildInfo, RuntimeInfo, set_timezone_db_path,
+                         MonthDayNano, VersionInfo, cpp_build_info,
+                         cpp_version, cpp_version_info, runtime_info,
+                         cpu_count, set_cpu_count, enable_signal_handlers,
                          io_thread_count, set_io_thread_count)
 
 

python/pyarrow/config.pxi

@@ -18,6 +18,7 @@
 from pyarrow.includes.libarrow cimport GetBuildInfo
 
 from collections import namedtuple
+import os
 
 
 VersionInfo = namedtuple('VersionInfo', ('major', 'minor', 'patch'))
@@ -74,3 +75,21 @@ def runtime_info():
     return RuntimeInfo(
         simd_level=frombytes(c_info.simd_level),
         detected_simd_level=frombytes(c_info.detected_simd_level))
+
+
+def set_timezone_db_path(path):
+    """
+    Configure the path to text timezone database on Windows.
+
+    Parameters
+    ----------
+    path : str
+        Path to text timezone database.
+    """
+    cdef:
+        CGlobalOptions options
+
+    if path is not None:
+        options.timezone_db_path = <c_string>tobytes(path)
+
+    check_status(Initialize(options))

python/pyarrow/includes/libarrow.pxd

@@ -80,6 +80,11 @@ cdef extern from "arrow/config.h" namespace "arrow" nogil:
 
     CRuntimeInfo GetRuntimeInfo()
 
+    cdef cppclass CGlobalOptions" arrow::GlobalOptions":
+        optional[c_string] timezone_db_path
+
+    CStatus Initialize(const CGlobalOptions& options)
+
 
 cdef extern from "arrow/util/future.h" namespace "arrow" nogil:
     cdef cppclass CFuture_Void" arrow::Future<>":

python/pyarrow/tests/conftest.py

@@ -28,6 +28,7 @@
 import hypothesis as h
 from ..conftest import groups, defaults
 
+from pyarrow import set_timezone_db_path
 from pyarrow.util import find_free_port
 
 
@@ -48,6 +49,12 @@
 os.environ['AWS_CONFIG_FILE'] = "/dev/null"
 
 
+if sys.platform == 'win32':
+    tzdata_set_path = os.environ.get('PYARROW_TZDATA_PATH', None)
+    if tzdata_set_path:
+        set_timezone_db_path(tzdata_set_path)
+
+
 def pytest_addoption(parser):
     # Create options to selectively enable test groups
     def bool_env(name, default=None):

python/pyarrow/tests/test_misc.py

@@ -22,6 +22,7 @@
 import pytest
 
 import pyarrow as pa
+from pyarrow.lib import ArrowInvalid
 
 
 def test_get_include():
@@ -116,6 +117,17 @@ def test_runtime_info():
         subprocess.check_call([sys.executable, "-c", code], env=env)
 
 
+@pytest.mark.skipif(sys.platform == "win32",
+                    reason="Path to timezone database is not configurable "
+                           "on non-Windows platforms")
+def test_set_timezone_db_path_non_windows():
+    # set_timezone_db_path raises an error on non-Windows platforms
+    with pytest.raises(ArrowInvalid,
+                       match="Arrow was set to use OS timezone "
+                             "database at compile time"):
+        pa.set_timezone_db_path("path")
+
+
 @pytest.mark.parametrize('klass', [
     pa.Field,
     pa.Schema,

python/pyarrow/tests/util.py

@@ -455,5 +455,11 @@ def windows_has_tzdata():
     This is the default location where tz.cpp will look for (until we make
     this configurable at run-time)
     """
-    tzdata_path = os.path.expandvars(r"%USERPROFILE%\Downloads\tzdata")
-    return os.path.exists(tzdata_path)
+    tzdata_bool = False
+    if "PYARROW_TZDATA_PATH" in os.environ:
+        tzdata_bool = os.path.exists(os.environ['PYARROW_TZDATA_PATH'])
+    if not tzdata_bool:
+        tzdata_path = os.path.expandvars(r"%USERPROFILE%\Downloads\tzdata")
+        tzdata_bool = os.path.exists(tzdata_path)
+
+    return tzdata_bool