From 550051959f550b849e879726c53605146489dea3 Mon Sep 17 00:00:00 2001
From: Animesh Kumar <kanimesh187@gmail.com>
Date: Tue, 6 Oct 2026 04:18:46 +0530
Subject: [PATCH] GH-51779: [C++][IPC] Read a DictionaryEncoding without
 indexType as int32 indices

---
 cpp/src/arrow/ipc/message_internal_test.cc | 24 ++++++++++++++++++++++
 cpp/src/arrow/ipc/metadata_internal.cc     |  8 ++++++--
 2 files changed, 30 insertions(+), 2 deletions(-)

diff --git a/cpp/src/arrow/ipc/message_internal_test.cc b/cpp/src/arrow/ipc/message_internal_test.cc
index 90b091c6b671..5d7fcb899b3d 100644
--- a/cpp/src/arrow/ipc/message_internal_test.cc
+++ b/cpp/src/arrow/ipc/message_internal_test.cc
@@ -19,6 +19,7 @@
 #include <iosfwd>
 #include <memory>
 #include <sstream>
+#include <string>
 
 #include <flatbuffers/flatbuffers.h>
 
@@ -114,6 +115,29 @@ TEST(TestMessageInternal, TestEndiannessRoundtrip) {
   }
 }
 
+// GH-51779: Schema.fbs allows DictionaryEncoding.indexType to be omitted, in
+// which case the indices are signed int32.
+TEST(TestMessageInternal, DictionaryEncodingWithoutIndexType) {
+  FBB fbb;
+  auto name = fbb.CreateString("f0");
+  auto value_type = flatbuf::CreateUtf8(fbb);
+  auto encoding = flatbuf::CreateDictionaryEncoding(fbb, /*id=*/0, /*indexType=*/0);
+  auto fb_field = flatbuf::CreateField(fbb, name, /*nullable=*/true, flatbuf::Type_Utf8,
+                                       value_type.Union(), encoding);
+  auto fb_schema = flatbuf::CreateSchema(fbb, flatbuf::Endianness_Little,
+                                         fbb.CreateVector({fb_field}));
+  fbb.Finish(flatbuf::CreateMessage(fbb, flatbuf::MetadataVersion_V5,
+                                    flatbuf::MessageHeader_Schema, fb_schema.Union()));
+  auto metadata = Buffer::FromString(
+      std::string(reinterpret_cast<const char*>(fbb.GetBufferPointer()), fbb.GetSize()));
+
+  ASSERT_OK_AND_ASSIGN(auto message, Message::Open(metadata, /*body=*/nullptr));
+  DictionaryMemo memo;
+  ASSERT_OK_AND_ASSIGN(auto schema, ReadSchema(*message, &memo));
+  AssertSchemaEqual(*::arrow::schema({field("f0", dictionary(int32(), utf8()))}),
+                    *schema);
+}
+
 struct SampleMessageParams {
   std::shared_ptr<const KeyValueMetadata> custom_metadata = {};
   int64_t body_length = 0;
diff --git a/cpp/src/arrow/ipc/metadata_internal.cc b/cpp/src/arrow/ipc/metadata_internal.cc
index 94a956916162..7ac5990513a6 100644
--- a/cpp/src/arrow/ipc/metadata_internal.cc
+++ b/cpp/src/arrow/ipc/metadata_internal.cc
@@ -898,8 +898,12 @@ Status FieldFromFlatbuffer(const flatbuf::Field* field, FieldPosition field_pos,
     // dictionary_memo
     std::shared_ptr<DataType> index_type;
     auto int_data = encoding->indexType();
-    CHECK_FLATBUFFERS_NOT_NULL(int_data, "DictionaryEncoding.indexType");
-    RETURN_NOT_OK(IntFromFlatbuffer(int_data, &index_type));
+    if (int_data == nullptr) {
+      // Schema.fbs: "If this field is null, the indices must be signed int32."
+      index_type = int32();
+    } else {
+      RETURN_NOT_OK(IntFromFlatbuffer(int_data, &index_type));
+    }
     dict_value_type = type;
     ARROW_ASSIGN_OR_RAISE(type,
                           DictionaryType::Make(index_type, type, encoding->isOrdered()));