GH-49351: [C++] Check TZDIR environment variable in vendored date library

cpp/src/arrow/public_api_test.cc

@@ -15,6 +15,7 @@
 // specific language governing permissions and limitations
 // under the License.
 
+#include <filesystem>
 #include <sstream>
 #include <string>
 
@@ -122,6 +123,28 @@ TEST(Misc, BuildInfo) {
   ASSERT_THAT(info.full_so_version, ::testing::HasSubstr(info.so_version));
 }
 
+#ifndef _WIN32
+TEST(Misc, TZDIREnvironmentVariable) {
+  // Find a valid zoneinfo directory
+  std::string tz_dir;
+  const char* env_tzdir = std::getenv("TZDIR");
+  if (env_tzdir != nullptr && std::filesystem::is_directory(env_tzdir)) {
+    tz_dir = env_tzdir;
+  } else if (std::filesystem::is_directory("/usr/share/zoneinfo")) {
+    tz_dir = "/usr/share/zoneinfo";
+  } else {
+    GTEST_SKIP() << "No system zoneinfo directory found";
+  }
+
+  // Set TZDIR and verify timezone resolution works
+  EnvVarGuard guard("TZDIR", tz_dir);
+  auto arr = ArrayFromJSON(timestamp(TimeUnit::SECOND, "UTC"), "[0]");
+  ASSERT_OK_AND_ASSIGN(
+      auto result, compute::Cast(arr, timestamp(TimeUnit::SECOND, "America/New_York")));
+  ASSERT_NE(result.make_array(), nullptr);
+}
+#endif
+
 TEST(Misc, SetTimezoneConfig) {
 #ifndef _WIN32
   GTEST_SKIP() << "Can only set the Timezone database on Windows";

cpp/src/arrow/vendored/datetime/README.md

@@ -25,7 +25,10 @@ The following changes are made:
 - enclose the `date` namespace inside the `arrow_vendored` namespace
 - fix 4 declarations like `CONSTCD11 date::day  operator "" _d(unsigned long long d) NOEXCEPT;`
   to not have offending whitespace for modern clang:
-  `CONSTCD11 date::day  operator ""_d(unsigned long long d) NOEXCEPT;` 
+  `CONSTCD11 date::day  operator ""_d(unsigned long long d) NOEXCEPT;`
+- check the TZDIR environment variable in `discover_tz_dir()` (tz.cpp) before
+  falling back to hardcoded paths, for compatibility with non-FHS Linux
+  distributions (e.g. NixOS)
 
 ## How to update
 

cpp/src/arrow/vendored/datetime/tz.cpp

@@ -486,6 +486,12 @@ discover_tz_dir()
 {
     struct stat sb;
     using namespace std;
+    // Check TZDIR environment variable first (POSIX standard)
+    const char* tz_dir_env = std::getenv("TZDIR");
+    if (tz_dir_env != nullptr
+        && stat(tz_dir_env, &sb) == 0
+        && S_ISDIR(sb.st_mode))
+        return tz_dir_env;
 #  if defined(ANDROID) || defined(__ANDROID__)
     CONSTDATA auto tz_dir_default = "/apex/com.android.tzdata/etc/tz";
     CONSTDATA auto tz_dir_fallback = "/system/usr/share/zoneinfo";

python/pyarrow/conftest.py

@@ -111,7 +111,8 @@
 if sys.platform == "win32":
     defaults['timezone_data'] = windows_has_tzdata()
 elif sys.platform == "emscripten":
-    defaults['timezone_data'] = os.path.exists("/usr/share/zoneinfo")
+    tz_dir = os.environ.get("TZDIR", "/usr/share/zoneinfo")
+    defaults['timezone_data'] = os.path.exists(tz_dir)
 
 try:
     import cython  # noqa