mapleFU commented on code in PR #33897:
URL: https://github.com/apache/arrow/pull/33897#discussion_r1114369542
##########
cpp/src/parquet/column_writer_test.cc:
##########
@@ -1143,5 +1143,143 @@ TEST(TestColumnWriter, WriteDataPageV2Header) {
}
}
+class TestSizeEstimated : public ::testing::Test {
+ public:
+ void SetUp() {
+ sink_ = CreateOutputStream();
+ node_ = std::static_pointer_cast<GroupNode>(
+ GroupNode::Make("schema", Repetition::REQUIRED,
+ {
+ schema::Int32("required", Repetition::REQUIRED),
+ }));
+ std::vector<schema::NodePtr> fields;
+ schema_descriptor_ = std::make_unique<SchemaDescriptor>();
+ schema_descriptor_->Init(node_);
+ }
+
+ std::shared_ptr<parquet::Int32Writer> BuildWriter(Compression::type
compression,
+ bool buffered,
+ bool enable_dictionary =
false) {
+ auto builder = WriterProperties::Builder();
+ builder.disable_dictionary()
+ ->disable_dictionary()
+ ->compression(compression)
+ ->data_pagesize(100 * sizeof(int));
+ if (enable_dictionary) {
+ builder.enable_dictionary();
+ } else {
+ builder.disable_dictionary();
+ }
+ writer_properties_ = builder.build();
+ metadata_ = ColumnChunkMetaDataBuilder::Make(writer_properties_,
+
schema_descriptor_->Column(0));
+
+ std::unique_ptr<PageWriter> pager = PageWriter::Open(
+ sink_, compression, Codec::UseDefaultCompressionLevel(),
metadata_.get(),
+ /* row_group_ordinal */ -1, /* column_chunk_ordinal*/ -1,
+ ::arrow::default_memory_pool(), /* buffered_row_group */ buffered,
+ /* header_encryptor */ NULLPTR, /* data_encryptor */ NULLPTR,
+ /* enable_checksum */ false);
+ return std::static_pointer_cast<parquet::Int32Writer>(
+ ColumnWriter::Make(metadata_.get(), std::move(pager),
writer_properties_.get()));
+ }
+
+ std::shared_ptr<::arrow::io::BufferOutputStream> sink_;
+ std::shared_ptr<GroupNode> node_;
+ std::unique_ptr<SchemaDescriptor> schema_descriptor_;
+
+ std::shared_ptr<WriterProperties> writer_properties_;
+ std::unique_ptr<ColumnChunkMetaDataBuilder> metadata_;
+};
+
+TEST_F(TestSizeEstimated, NonBuffered) {
+ auto required_writer = this->BuildWriter(Compression::UNCOMPRESSED, false);
+
+ // Write half page
+ for (int32_t i = 0; i < 50; i++) {
+ required_writer->WriteBatch(1, nullptr, nullptr, &i);
+ }
+ // Page flushed, check size
+ EXPECT_EQ(0, required_writer->total_bytes_written());
+ EXPECT_EQ(0, required_writer->total_compressed_bytes()); // unbuffered
+ EXPECT_EQ(0, required_writer->total_compressed_bytes_written());
+ for (int32_t i = 0; i < 50; i++) {
+ required_writer->WriteBatch(1, nullptr, nullptr, &i);
+ }
+ // Page flushed, check size
+ EXPECT_LT(400, required_writer->total_bytes_written());
+ EXPECT_EQ(0, required_writer->total_compressed_bytes());
+ EXPECT_LT(400, required_writer->total_compressed_bytes_written());
+}
+
+TEST_F(TestSizeEstimated, Buffered) {
+ auto required_writer = this->BuildWriter(Compression::UNCOMPRESSED, true);
+
+ // Write half page
+ for (int32_t i = 0; i < 50; i++) {
+ required_writer->WriteBatch(1, nullptr, nullptr, &i);
+ }
+ // Page flushed, check size
+ EXPECT_EQ(0, required_writer->total_bytes_written());
+ EXPECT_EQ(0, required_writer->total_compressed_bytes()); // buffered
+ EXPECT_EQ(0, required_writer->total_compressed_bytes_written());
+ for (int32_t i = 0; i < 50; i++) {
+ required_writer->WriteBatch(1, nullptr, nullptr, &i);
+ }
+ // Page flushed, check size
+ EXPECT_LT(400, required_writer->total_bytes_written());
+ EXPECT_EQ(0, required_writer->total_compressed_bytes());
+ EXPECT_LT(400, required_writer->total_compressed_bytes_written());
+}
+
+TEST_F(TestSizeEstimated, NonBufferedDictionary) {
+ auto required_writer = this->BuildWriter(Compression::UNCOMPRESSED, false,
true);
+
+ // Write half page
+ // for dict, keep value equal
+ int32_t dict_value = 1;
+ for (int32_t i = 0; i < 50; i++) {
+ required_writer->WriteBatch(1, nullptr, nullptr, &dict_value);
+ }
+ // Page flushed, check size
+ EXPECT_EQ(0, required_writer->total_bytes_written());
+ EXPECT_EQ(0, required_writer->total_compressed_bytes());
+ EXPECT_EQ(0, required_writer->total_compressed_bytes_written());
+ // write a huge batch to trigger page flush
+ for (int32_t i = 0; i < 50000; i++) {
Review Comment:
In parquet-format v2, Dictionary has 2 pages:
1. (Dict index page) PLAIN
2. (Data Page) RLE_DICTIONARY
`RLE_DICTIONARY` uses `RLE/Bit-packing` as encoding. It can well compress
the page when all values are equal. Arrow C++ Parquet default use `size` as
Page Split condition. So, I use a large value here to trigger page flush.
--
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]
