From d5e5e5dbeb9fb23e895b9ac654fafb764a164c0b Mon Sep 17 00:00:00 2001
From: Peter Schafhalter <pschafhalter@berkeley.edu>
Date: Wed, 16 May 2018 17:00:11 -0700
Subject: [PATCH] Fix python2 integer serialization bug

---
 cpp/src/arrow/python/arrow_to_python.cc | 10 ++++++++--
 cpp/src/arrow/python/python_to_arrow.cc | 11 ++++++++++-
 2 files changed, 18 insertions(+), 3 deletions(-)

diff --git a/cpp/src/arrow/python/arrow_to_python.cc b/cpp/src/arrow/python/arrow_to_python.cc
index 57884e58c374c..518a560b89483 100644
--- a/cpp/src/arrow/python/arrow_to_python.cc
+++ b/cpp/src/arrow/python/arrow_to_python.cc
@@ -112,9 +112,15 @@ Status GetValue(PyObject* context, const UnionArray& parent, const Array& arr,
     case Type::BOOL:
       *result = PyBool_FromLong(checked_cast<const BooleanArray&>(arr).Value(index));
       return Status::OK();
-    case Type::INT64:
-      *result = PyLong_FromSsize_t(checked_cast<const Int64Array&>(arr).Value(index));
+    case Type::INT64: {
+      const std::string& child_name = parent.type()->child(type)->name();
+      if (child_name == "py2_int") {
+        *result = PyInt_FromSsize_t(checked_cast<const Int64Array&>(arr).Value(index));
+      } else {
+        *result = PyLong_FromSsize_t(checked_cast<const Int64Array&>(arr).Value(index));
+      }
       return Status::OK();
+    }
     case Type::BINARY: {
       int32_t nchars;
       const uint8_t* str = checked_cast<const BinaryArray&>(arr).GetValue(index, &nchars);
diff --git a/cpp/src/arrow/python/python_to_arrow.cc b/cpp/src/arrow/python/python_to_arrow.cc
index 64cf2b4c168c0..23445ee26d0e7 100644
--- a/cpp/src/arrow/python/python_to_arrow.cc
+++ b/cpp/src/arrow/python/python_to_arrow.cc
@@ -61,6 +61,7 @@ class SequenceBuilder {
         nones_(pool),
         bools_(::arrow::boolean(), pool),
         ints_(::arrow::int64(), pool),
+        py2_ints_(::arrow::int64(), pool),
         bytes_(::arrow::binary(), pool),
         strings_(pool),
         half_floats_(::arrow::float16(), pool),
@@ -103,6 +104,11 @@ class SequenceBuilder {
     return AppendPrimitive(data, &bool_tag_, &bools_);
   }
 
+  /// Appending a python 2 int64_t to the sequence
+  Status AppendPy2Int64(const int64_t data) {
+    return AppendPrimitive(data, &py2_int_tag_, &py2_ints_);
+  }
+
   /// Appending an int64_t to the sequence
   Status AppendInt64(const int64_t data) {
     return AppendPrimitive(data, &int_tag_, &ints_);
@@ -250,6 +256,7 @@ class SequenceBuilder {
 
     RETURN_NOT_OK(AddElement(bool_tag_, &bools_));
     RETURN_NOT_OK(AddElement(int_tag_, &ints_));
+    RETURN_NOT_OK(AddElement(py2_int_tag_, &py2_ints_, "py2_int"));
     RETURN_NOT_OK(AddElement(string_tag_, &strings_));
     RETURN_NOT_OK(AddElement(bytes_tag_, &bytes_));
     RETURN_NOT_OK(AddElement(half_float_tag_, &half_floats_));
@@ -280,6 +287,7 @@ class SequenceBuilder {
   NullBuilder nones_;
   BooleanBuilder bools_;
   Int64Builder ints_;
+  Int64Builder py2_ints_;
   BinaryBuilder bytes_;
   StringBuilder strings_;
   HalfFloatBuilder half_floats_;
@@ -302,6 +310,7 @@ class SequenceBuilder {
   // happens in the UPDATE macro in sequence.cc.
   int8_t bool_tag_ = -1;
   int8_t int_tag_ = -1;
+  int8_t py2_int_tag_ = -1;
   int8_t string_tag_ = -1;
   int8_t bytes_tag_ = -1;
   int8_t half_float_tag_ = -1;
@@ -481,7 +490,7 @@ Status Append(PyObject* context, PyObject* elem, SequenceBuilder* builder,
     }
 #if PY_MAJOR_VERSION < 3
   } else if (PyInt_Check(elem)) {
-    RETURN_NOT_OK(builder->AppendInt64(static_cast<int64_t>(PyInt_AS_LONG(elem))));
+    RETURN_NOT_OK(builder->AppendPy2Int64(static_cast<int64_t>(PyInt_AS_LONG(elem))));
 #endif
   } else if (PyBytes_Check(elem)) {
     auto data = reinterpret_cast<uint8_t*>(PyBytes_AS_STRING(elem));