Upgrade C++ Arrow to 21.0.0 (#548)

Author: adamreeve

.github/workflows/ci.yml

@@ -115,7 +115,7 @@ jobs:
       if: runner.os == 'Linux' && matrix.arch == 'x64'
       run: |
         docker run -d --name centos --entrypoint tail -v $PWD:$PWD -v $VCPKG_INSTALLATION_ROOT:$VCPKG_INSTALLATION_ROOT quay.io/pypa/manylinux2014_x86_64 -f /dev/null
-        docker exec centos sh -c "yum install -y devtoolset-7 rh-git227 httpd24-curl flex bison perl-Data-Dumper perl-IPC-Cmd && \
+        docker exec centos sh -c "yum install -y devtoolset-10 rh-git227 httpd24-curl flex bison perl-Data-Dumper perl-IPC-Cmd && \
                                   curl -fsSL -o /tmp/cmake.sh https://github.com/Kitware/CMake/releases/download/v${{ steps.cmake-info.outputs.version }}/cmake-${{ steps.cmake-info.outputs.version }}-linux-x86_64.sh && \
                                   sh /tmp/cmake.sh --skip-license --prefix=/usr/local && \
                                   rm /tmp/cmake.sh"
@@ -150,7 +150,7 @@ jobs:
       if: runner.os == 'Linux' || runner.os == 'macOS'
       run: |
         if [ "${{ runner.os }}" == "Linux" ] && [ "${{ matrix.arch }}" == "x64" ]; then
-          exec="docker exec -w $PWD -e GITHUB_ACTIONS -e ACTIONS_CACHE_URL -e ACTIONS_RUNTIME_TOKEN -e VCPKG_BINARY_SOURCES -e VCPKG_INSTALLATION_ROOT centos scl enable devtoolset-7 rh-git227 httpd24 --"
+          exec="docker exec -w $PWD -e GITHUB_ACTIONS -e ACTIONS_CACHE_URL -e ACTIONS_RUNTIME_TOKEN -e VCPKG_BINARY_SOURCES -e VCPKG_INSTALLATION_ROOT centos scl enable devtoolset-10 rh-git227 httpd24 --"
         fi
         $exec ./build_unix.sh ${{ matrix.arch }}
       env:

README.md

@@ -270,7 +270,7 @@ dotnet jb cleanupcode "csharp" "csharp.test" "csharp.benchmark" --profile="Built
 Building ParquetSharp natively requires the following dependencies:
 - A modern C++ compiler toolchain
 - .NET SDK 8.0
-- Apache Arrow (20.0.0)
+- Apache Arrow (21.0.0)
 
 For building Arrow (including Parquet) and its dependencies, we recommend using Microsoft's [vcpkg](https://vcpkg.io).
 The build scripts will use an existing vcpkg installation if either of the `VCPKG_INSTALLATION_ROOT` or `VCPKG_ROOT` environment variables are defined, otherwise vcpkg will be downloaded into the build directory.

cpp/ColumnDecryptionPropertiesBuilder.cpp

@@ -16,7 +16,7 @@ extern "C"
 
     PARQUETSHARP_EXPORT ExceptionInfo* ColumnDecryptionPropertiesBuilder_Create_From_Column_Path(const std::shared_ptr<schema::ColumnPath>* path, ColumnDecryptionProperties::Builder** builder)
     {
-        TRYCATCH(*builder = new ColumnDecryptionProperties::Builder(*path);)
+        TRYCATCH(*builder = new ColumnDecryptionProperties::Builder(**path);)
     }
 
     PARQUETSHARP_EXPORT void ColumnDecryptionPropertiesBuilder_Free(ColumnDecryptionProperties::Builder* builder)

cpp/ColumnEncryptionPropertiesBuilder.cpp

@@ -16,7 +16,7 @@ extern "C"
 
     PARQUETSHARP_EXPORT ExceptionInfo* ColumnEncryptionPropertiesBuilder_Create_From_Column_Path(const std::shared_ptr<schema::ColumnPath>* path, ColumnEncryptionProperties::Builder** builder)
     {
-        TRYCATCH(*builder = new ColumnEncryptionProperties::Builder(*path);)
+        TRYCATCH(*builder = new ColumnEncryptionProperties::Builder(**path);)
     }
 
     PARQUETSHARP_EXPORT void ColumnEncryptionPropertiesBuilder_Free(ColumnEncryptionProperties::Builder* builder)

cpp/Enums.cpp

@@ -53,7 +53,10 @@ namespace
 		static_assert(LogicalType::Type::BSON == 13);
 		static_assert(LogicalType::Type::UUID == 14);
 		static_assert(LogicalType::Type::FLOAT16 == 15);
-		static_assert(LogicalType::Type::NONE == 16);
+		static_assert(LogicalType::Type::GEOMETRY == 16);
+		static_assert(LogicalType::Type::GEOGRAPHY == 17);
+		static_assert(LogicalType::Type::VARIANT == 18);
+		static_assert(LogicalType::Type::NONE == 19);
 
 		static_assert(ParquetCipher::AES_GCM_V1 == 0);
 		static_assert(ParquetCipher::AES_GCM_CTR_V1 == 1);

cpp/FileDecryptionPropertiesBuilder.cpp

@@ -37,7 +37,7 @@ extern "C"
 	            m.insert(std::make_pair((*column_decryption_properties[i])->column_path(), (*column_decryption_properties[i])));
 	        }
 
-	        builder->column_keys(m);
+	        builder->column_keys(std::move(m));
         )
     }
 

cpp/FileEncryptionPropertiesBuilder.cpp

@@ -60,7 +60,7 @@ extern "C"
                 m.insert(std::make_pair((*column_encryption_properties[i])->column_path(), (*column_encryption_properties[i])));
 			}
 
-            builder->encrypted_columns(m);
+            builder->encrypted_columns(std::move(m));
         )
     }
 

csharp.test/TestPhysicalTypeRoundtrip.cs

@@ -163,7 +163,7 @@ private static void AssertReadRoundtrip(ResizableBuffer buffer, ExpectedColumn[]
 
             var numRows = expectedColumns.First().Values.Length;
 
-            Assert.AreEqual("parquet-cpp-arrow version 20.0.0", fileMetaData.CreatedBy);
+            Assert.AreEqual("parquet-cpp-arrow version 21.0.0", fileMetaData.CreatedBy);
             Assert.AreEqual(new Dictionary<string, string> {{"case", "Test"}, {"Awesome", "true"}}, fileMetaData.KeyValueMetadata);
             Assert.AreEqual(expectedColumns.Length, fileMetaData.NumColumns);
             Assert.AreEqual(numRows, fileMetaData.NumRows);
@@ -174,7 +174,7 @@ private static void AssertReadRoundtrip(ResizableBuffer buffer, ExpectedColumn[]
             // The parquet format only stores an integer file version (1 or 2) and
             // 2 gets mapped to the latest 2.x version.
             Assert.AreEqual(ParquetVersion.PARQUET_2_6, fileMetaData.Version);
-            Assert.AreEqual("parquet-cpp-arrow version 20.0.0", fileMetaData.WriterVersion.ToString());
+            Assert.AreEqual("parquet-cpp-arrow version 21.0.0", fileMetaData.WriterVersion.ToString());
 
             using var rowGroupReader = fileReader.RowGroup(0);
             var rowGroupMetaData = rowGroupReader.MetaData;

csharp.test/TestWriterProperties.cs

@@ -14,7 +14,7 @@ public static void TestDefaultProperties()
         {
             var p = WriterProperties.GetDefaultWriterProperties();
 
-            Assert.AreEqual("parquet-cpp-arrow version 20.0.0", p.CreatedBy);
+            Assert.AreEqual("parquet-cpp-arrow version 21.0.0", p.CreatedBy);
             Assert.AreEqual(Compression.Uncompressed, p.Compression(new ColumnPath("anypath")));
             Assert.AreEqual(int.MinValue, p.CompressionLevel(new ColumnPath("anypath")));
             Assert.AreEqual(1024 * 1024, p.DataPageSize);

csharp/LogicalType.cs

@@ -22,6 +22,9 @@ public enum LogicalTypeEnum
         Uuid = 14,
         None = 15,
         Float16 = 16,
+        Geometry = 17,
+        Geography = 18,
+        Variant = 19,
     }
 
     /// <summary>
@@ -47,7 +50,10 @@ internal enum CppLogicalTypeEnum
         Bson = 13,
         Uuid = 14,
         Float16 = 15,
-        None = 16,
+        Geometry = 16,
+        Geography = 17,
+        Variant = 18,
+        None = 19,
     }
 
     internal static class CppLogicalTypeEnumExtensions
@@ -72,6 +78,9 @@ public static LogicalTypeEnum ToPublicEnum(this CppLogicalTypeEnum enumValue)
                 CppLogicalTypeEnum.Bson => LogicalTypeEnum.Bson,
                 CppLogicalTypeEnum.Uuid => LogicalTypeEnum.Uuid,
                 CppLogicalTypeEnum.Float16 => LogicalTypeEnum.Float16,
+                CppLogicalTypeEnum.Geometry => LogicalTypeEnum.Geometry,
+                CppLogicalTypeEnum.Geography => LogicalTypeEnum.Geography,
+                CppLogicalTypeEnum.Variant => LogicalTypeEnum.Variant,
                 CppLogicalTypeEnum.None => LogicalTypeEnum.None,
                 _ => throw new ArgumentOutOfRangeException(nameof(enumValue), enumValue, null)
             };