GH-40633: [C++][Python] Fix ORC crash when file contains unknown time…

…zone (#45051)

### Rationale for this change

If the timezone database is present on the system, but does not contain a timezone referenced in a ORC file, the ORC reader will crash with an uncaught C++ exception.

This can happen for example on Ubuntu 24.04 where some timezone aliases have been removed from the main `tzdata` package to a `tzdata-legacy` package. If `tzdata-legacy` is not installed, trying to read a ORC file that references e.g. the "US/Pacific" timezone would crash.

Here is a backtrace excerpt:
```
#12 0x00007f1a3ce23a55 in std::terminate() () from /lib/x86_64-linux-gnu/libstdc++.so.6
#13 0x00007f1a3ce39391 in __cxa_throw () from /lib/x86_64-linux-gnu/libstdc++.so.6
#14 0x00007f1a3f4accc4 in orc::loadTZDB(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) ()
   from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
#15 0x00007f1a3f4ad392 in std::call_once<orc::LazyTimezone::getImpl() const::{lambda()#1}>(std::once_flag&, orc::LazyTimezone::getImpl() const::{lambda()#1}&&)::{lambda()#2}::_FUN() () from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
#16 0x00007f1a4298bec3 in __pthread_once_slow (once_control=0xa5ca7c8, init_routine=0x7f1a3ce69420 <__once_proxy>) at ./nptl/pthread_once.c:116
#17 0x00007f1a3f4a9ad0 in orc::LazyTimezone::getEpoch() const ()
   from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
#18 0x00007f1a3f4e76b1 in orc::TimestampColumnReader::TimestampColumnReader(orc::Type const&, orc::StripeStreams&, bool) ()
   from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
#19 0x00007f1a3f4e84ad in orc::buildReader(orc::Type const&, orc::StripeStreams&, bool, bool, bool) ()
   from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
#20 0x00007f1a3f4e8dd7 in orc::StructColumnReader::StructColumnReader(orc::Type const&, orc::StripeStreams&, bool, bool) ()
   from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
#21 0x00007f1a3f4e8532 in orc::buildReader(orc::Type const&, orc::StripeStreams&, bool, bool, bool) ()
   from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
#22 0x00007f1a3f4925e9 in orc::RowReaderImpl::startNextStripe() ()
   from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
#23 0x00007f1a3f492c9d in orc::RowReaderImpl::next(orc::ColumnVectorBatch&) ()
   from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
#24 0x00007f1a3e6b251f in arrow::adapters::orc::ORCFileReader::Impl::ReadBatch(orc::RowReaderOptions const&, std::shared_ptr<arrow::Schema> const&, long) ()
   from /tmp/arrow-HEAD.ArqTs/venv-wheel-3.12-manylinux_2_17_x86_64.manylinux2014_x86_64/lib/python3.12/site-packages/pyarrow/libarrow.so.1900
```

### What changes are included in this PR?

Catch C++ exceptions when iterating ORC batches instead of letting them slip through.

### Are these changes tested?

Yes.

### Are there any user-facing changes?

No.
* GitHub Issue: #40633

Authored-by: Antoine Pitrou <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

cpp/src/arrow/adapters/orc/adapter.cc

@@ -145,28 +145,29 @@ class OrcStripeReader : public RecordBatchReader {
 
   Status ReadNext(std::shared_ptr<RecordBatch>* out) override {
     std::unique_ptr<liborc::ColumnVectorBatch> batch;
-    ORC_CATCH_NOT_OK(batch = row_reader_->createRowBatch(batch_size_));
+    std::unique_ptr<RecordBatchBuilder> builder;
+
+    ORC_BEGIN_CATCH_NOT_OK
+    batch = row_reader_->createRowBatch(batch_size_);
 
     const liborc::Type& type = row_reader_->getSelectedType();
     if (!row_reader_->next(*batch)) {
       out->reset();
       return Status::OK();
     }
 
-    std::unique_ptr<RecordBatchBuilder> builder;
     ARROW_ASSIGN_OR_RAISE(builder,
                           RecordBatchBuilder::Make(schema_, pool_, batch->numElements));
-
     // The top-level type must be a struct to read into an arrow table
     const auto& struct_batch = checked_cast<liborc::StructVectorBatch&>(*batch);
 
     for (int i = 0; i < builder->num_fields(); i++) {
       RETURN_NOT_OK(AppendBatch(type.getSubtype(i), struct_batch.fields[i], 0,
                                 batch->numElements, builder->GetField(i)));
     }
+    ORC_END_CATCH_NOT_OK
 
-    ARROW_ASSIGN_OR_RAISE(*out, builder->Flush());
-    return Status::OK();
+    return builder->Flush().Value(out);
   }
 
  private:
@@ -470,15 +471,13 @@ class ORCFileReader::Impl {
                                                  int64_t nrows) {
     std::unique_ptr<liborc::RowReader> row_reader;
     std::unique_ptr<liborc::ColumnVectorBatch> batch;
+    std::unique_ptr<RecordBatchBuilder> builder;
 
     ORC_BEGIN_CATCH_NOT_OK
     row_reader = reader_->createRowReader(opts);
     batch = row_reader->createRowBatch(std::min(nrows, kReadRowsBatch));
-    ORC_END_CATCH_NOT_OK
 
-    std::unique_ptr<RecordBatchBuilder> builder;
     ARROW_ASSIGN_OR_RAISE(builder, RecordBatchBuilder::Make(schema, pool_, nrows));
-
     // The top-level type must be a struct to read into an arrow table
     const auto& struct_batch = checked_cast<liborc::StructVectorBatch&>(*batch);
 
@@ -489,6 +488,7 @@ class ORCFileReader::Impl {
                                   batch->numElements, builder->GetField(i)));
       }
     }
+    ORC_END_CATCH_NOT_OK
 
     return builder->Flush();
   }

python/pyarrow/tests/test_orc.py

@@ -15,9 +15,14 @@
 # specific language governing permissions and limitations
 # under the License.
 
-import pytest
 import decimal
 import datetime
+from pathlib import Path
+import shutil
+import subprocess
+import sys
+
+import pytest
 
 import pyarrow as pa
 from pyarrow import fs
@@ -140,6 +145,57 @@ def test_example_using_json(filename, datadir):
     check_example_file(path, table, need_fix=True)
 
 
+def test_timezone_database_absent(datadir):
+    # Example file relies on the timezone "US/Pacific". It should gracefully
+    # fail, not crash, if the timezone database is not found.
+    path = datadir / 'TestOrcFile.testDate1900.orc'
+    code = f"""if 1:
+        import os
+        os.environ['TZDIR'] = '/tmp/non_existent'
+
+        from pyarrow import orc
+        try:
+            orc_file = orc.ORCFile({str(path)!r})
+            orc_file.read()
+        except Exception as e:
+            assert "time zone database" in str(e).lower(), e
+        else:
+            assert False, "Should have raised exception"
+    """
+    subprocess.run([sys.executable, "-c", code], check=True)
+
+
+def test_timezone_absent(datadir, tmpdir):
+    # Example file relies on the timezone "US/Pacific". It should gracefully
+    # fail, not crash, if the timezone database is present but the timezone
+    # is not found (GH-40633).
+    source_tzdir = Path('/usr/share/zoneinfo')
+    if not source_tzdir.exists():
+        pytest.skip(f"Test needs timezone database in {source_tzdir}")
+    tzdir = Path(tmpdir / 'zoneinfo')
+    try:
+        shutil.copytree(source_tzdir, tzdir, symlinks=True)
+    except OSError as e:
+        pytest.skip(f"Failed to copy timezone database: {e}")
+    (tzdir / 'US' / 'Pacific').unlink(missing_ok=True)
+
+    path = datadir / 'TestOrcFile.testDate1900.orc'
+    code = f"""if 1:
+        import os
+        os.environ['TZDIR'] = {str(tzdir)!r}
+
+        from pyarrow import orc
+        orc_file = orc.ORCFile({str(path)!r})
+        try:
+            orc_file.read()
+        except Exception as e:
+            assert "zoneinfo/US/Pacific" in str(e), e
+        else:
+            assert False, "Should have raised exception"
+    """
+    subprocess.run([sys.executable, "-c", code], check=True)
+
+
 def test_orcfile_empty(datadir):
     from pyarrow import orc