HuaHuaY commented on code in PR #198:
URL: https://github.com/apache/iceberg-cpp/pull/198#discussion_r2315608562
##########
test/parquet_test.cc:
##########
@@ -17,38 +17,121 @@
* under the License.
*/
+#include <optional>
+
#include <arrow/array.h>
#include <arrow/c/bridge.h>
#include <arrow/json/from_string.h>
#include <arrow/record_batch.h>
#include <arrow/table.h>
+#include <arrow/type.h>
#include <arrow/util/key_value_metadata.h>
#include <parquet/arrow/reader.h>
#include <parquet/arrow/writer.h>
#include <parquet/metadata.h>
+#include "iceberg/arrow/arrow_error_transform_internal.h"
#include "iceberg/arrow/arrow_fs_file_io_internal.h"
-#include "iceberg/parquet/parquet_reader.h"
+#include "iceberg/file_reader.h"
+#include "iceberg/file_writer.h"
#include "iceberg/parquet/parquet_register.h"
+#include "iceberg/result.h"
#include "iceberg/schema.h"
+#include "iceberg/schema_field.h"
+#include "iceberg/schema_internal.h"
#include "iceberg/type.h"
#include "iceberg/util/checked_cast.h"
+#include "iceberg/util/macros.h"
#include "matchers.h"
-#include "temp_file_test_base.h"
namespace iceberg::parquet {
-class ParquetReaderTest : public TempFileTestBase {
+namespace {
+
+Status WriteTableInner(Writer& writer, std::shared_ptr<::arrow::Array> data) {
+ ArrowArray arr;
+ ICEBERG_ARROW_RETURN_NOT_OK(::arrow::ExportArray(*data, &arr));
+ ICEBERG_RETURN_UNEXPECTED(writer.Write(arr));
+ return writer.Close();
+}
+
+Status WriteTable(std::shared_ptr<::arrow::Array> data,
+ const WriterOptions& writer_options) {
+ ICEBERG_ASSIGN_OR_RAISE(
+ auto writer, WriterFactoryRegistry::Open(FileFormatType::kParquet,
writer_options));
+ return WriteTableInner(*writer, data);
+}
+
+Status ReadTable(std::shared_ptr<::arrow::Array>& out,
+ const ReaderOptions& reader_options) {
+ ICEBERG_ASSIGN_OR_RAISE(
+ auto reader, ReaderFactoryRegistry::Open(FileFormatType::kParquet,
reader_options));
+ ICEBERG_ASSIGN_OR_RAISE(auto read_data, reader->Next());
+
+ if (!read_data.has_value()) {
+ out = nullptr;
+ return {};
+ }
+ auto arrow_c_array = read_data.value();
+
+ ArrowSchema arrow_schema;
+ ICEBERG_RETURN_UNEXPECTED(ToArrowSchema(*reader_options.projection,
&arrow_schema));
+ ICEBERG_ARROW_ASSIGN_OR_RETURN(out,
+ ::arrow::ImportArray(&arrow_c_array,
&arrow_schema));
+ return {};
+}
+
+void DoRoundtrip(std::shared_ptr<::arrow::Array> data, std::shared_ptr<Schema>
schema,
+ std::shared_ptr<::arrow::Array>& out) {
+ std::shared_ptr<FileIO> file_io =
arrow::ArrowFileSystemFileIO::MakeMockFileIO();
+ const std::string basePath = "base.parquet";
+
+ auto writer_data = WriterFactoryRegistry::Open(
+ FileFormatType::kParquet, {.path = basePath, .schema = schema, .io =
file_io});
+ ASSERT_THAT(writer_data, IsOk())
+ << "Failed to create writer: " << writer_data.error().message;
+ auto writer = std::move(writer_data.value());
+ ASSERT_THAT(WriteTableInner(*writer, data), IsOk());
+
+ ASSERT_THAT(ReadTable(out, {.path = basePath,
+ .length = writer->length(),
+ .io = file_io,
+ .projection = schema}),
+ IsOk());
+
+ ASSERT_TRUE(out != nullptr) << "Reader.Next() returned no data";
+}
+
+} // namespace
+
+class ParquetReaderTest : public ::testing::Test {
protected:
static void SetUpTestSuite() { parquet::RegisterAll(); }
void SetUp() override {
- TempFileTestBase::SetUp();
- file_io_ = arrow::ArrowFileSystemFileIO::MakeLocalFileIO();
- temp_parquet_file_ = CreateNewTempFilePathWithSuffix(".parquet");
+ file_io_ = arrow::ArrowFileSystemFileIO::MakeMockFileIO();
+ temp_parquet_file_ = "parquet_reader_test.parquet";
}
void CreateSimpleParquetFile() {
+ auto schema = std::make_shared<Schema>(
+ std::vector<SchemaField>{SchemaField::MakeRequired(1, "id", int32()),
+ SchemaField::MakeOptional(2, "name",
string())});
+
+ ArrowSchema arrow_c_schema;
+ ASSERT_THAT(ToArrowSchema(*schema, &arrow_c_schema), IsOk());
+ auto arrow_schema = ::arrow::ImportType(&arrow_c_schema).ValueOrDie();
Review Comment:
`::arrow::ImportSchema` converts a `ArrowSchema` to
`std::shared_ptr<Schema>` and it doesn't seems to be helpful here.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
