Merge pull request #47492 from ClickHouse/revert-45878

Revert #45878 and add a test

src/Core/Settings.h

@@ -803,7 +803,6 @@ class IColumn;
     M(Bool, input_format_tsv_detect_header, true, "Automatically detect header with names and types in TSV format", 0) \
     M(Bool, input_format_custom_detect_header, true, "Automatically detect header with names and types in CustomSeparated format", 0) \
     M(Bool, input_format_parquet_skip_columns_with_unsupported_types_in_schema_inference, false, "Skip columns with unsupported types while schema inference for format Parquet", 0) \
-    M(UInt64, input_format_parquet_max_block_size, 8192, "Max block size for parquet reader.", 0) \
     M(Bool, input_format_protobuf_skip_fields_with_unsupported_types_in_schema_inference, false, "Skip fields with unsupported types while schema inference for format Protobuf", 0) \
     M(Bool, input_format_capn_proto_skip_fields_with_unsupported_types_in_schema_inference, false, "Skip columns with unsupported types while schema inference for format CapnProto", 0) \
     M(Bool, input_format_orc_skip_columns_with_unsupported_types_in_schema_inference, false, "Skip columns with unsupported types while schema inference for format ORC", 0) \

src/Formats/FormatFactory.cpp

@@ -117,7 +117,6 @@ FormatSettings getFormatSettings(ContextPtr context, const Settings & settings)
     format_settings.parquet.skip_columns_with_unsupported_types_in_schema_inference = settings.input_format_parquet_skip_columns_with_unsupported_types_in_schema_inference;
     format_settings.parquet.output_string_as_string = settings.output_format_parquet_string_as_string;
     format_settings.parquet.output_fixed_string_as_fixed_byte_array = settings.output_format_parquet_fixed_string_as_fixed_byte_array;
-    format_settings.parquet.max_block_size = settings.input_format_parquet_max_block_size;
     format_settings.parquet.output_compression_method = settings.output_format_parquet_compression_method;
     format_settings.pretty.charset = settings.output_format_pretty_grid_charset.toString() == "ASCII" ? FormatSettings::Pretty::Charset::ASCII : FormatSettings::Pretty::Charset::UTF8;
     format_settings.pretty.color = settings.output_format_pretty_color;

src/Formats/FormatSettings.h

@@ -211,7 +211,6 @@ struct FormatSettings
         std::unordered_set<int> skip_row_groups = {};
         bool output_string_as_string = false;
         bool output_fixed_string_as_fixed_byte_array = true;
-        UInt64 max_block_size = 8192;
         ParquetVersion output_version;
         ParquetCompression output_compression_method = ParquetCompression::SNAPPY;
     } parquet;

src/Processors/Formats/Impl/ParquetBlockInputFormat.cpp

@@ -45,44 +45,38 @@ Chunk ParquetBlockInputFormat::generate()
     block_missing_values.clear();
 
     if (!file_reader)
-    {
         prepareReader();
-        file_reader->set_batch_size(format_settings.parquet.max_block_size);
-        std::vector<int> row_group_indices;
-        for (int i = 0; i < row_group_total; ++i)
-        {
-            if (!skip_row_groups.contains(i))
-                row_group_indices.emplace_back(i);
-        }
-        auto read_status = file_reader->GetRecordBatchReader(row_group_indices, column_indices, &current_record_batch_reader);
-        if (!read_status.ok())
-            throw DB::ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading Parquet data: {}", read_status.ToString());
-    }
 
     if (is_stopped)
         return {};
 
-    auto batch = current_record_batch_reader->Next();
-    if (!batch.ok())
-    {
+    while (row_group_current < row_group_total && skip_row_groups.contains(row_group_current))
+        ++row_group_current;
+
+    if (row_group_current >= row_group_total)
+        return res;
+
+    std::shared_ptr<arrow::Table> table;
+
+    std::unique_ptr<::arrow::RecordBatchReader> rbr;
+    std::vector<int> row_group_indices { row_group_current };
+    arrow::Status get_batch_reader_status = file_reader->GetRecordBatchReader(row_group_indices, column_indices, &rbr);
+
+    if (!get_batch_reader_status.ok())
         throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading Parquet data: {}",
-                               batch.status().ToString());
-    }
-    if (*batch)
-    {
-        auto tmp_table = arrow::Table::FromRecordBatches({*batch});
+                               get_batch_reader_status.ToString());
+
+    arrow::Status read_status = rbr->ReadAll(&table);
+
+    if (!read_status.ok())
+        throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Error while reading Parquet data: {}", read_status.ToString());
+
+    ++row_group_current;
+
     /// If defaults_for_omitted_fields is true, calculate the default values from default expression for omitted fields.
     /// Otherwise fill the missing columns with zero values of its type.
     BlockMissingValues * block_missing_values_ptr = format_settings.defaults_for_omitted_fields ? &block_missing_values : nullptr;
-        arrow_column_to_ch_column->arrowTableToCHChunk(res, *tmp_table, (*tmp_table)->num_rows(), block_missing_values_ptr);
-    }
-    else
-    {
-        current_record_batch_reader.reset();
-        file_reader.reset();
-        return {};
-    }
-
+    arrow_column_to_ch_column->arrowTableToCHChunk(res, table, table->num_rows(), block_missing_values_ptr);
     return res;
 }
 

tests/queries/0_stateless/02588_parquet_bug.reference

@@ -0,0 +1,3 @@
+cta	224.0.90.10	1670964058771367936	64066044	NYSE	cqs_pillar	quote	\N	\N	\N	82.92	1	R	82.97	2	R
+1670964058771367936
+1670946478544048640	ARCA	cqs_pillar

tests/queries/0_stateless/02588_parquet_bug.sh

@@ -0,0 +1,11 @@
+#!/usr/bin/env bash
+# Tags: no-fasttest
+
+CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
+# shellcheck source=../shell_config.sh
+. "$CURDIR"/../shell_config.sh
+
+$CLICKHOUSE_LOCAL -q "select * from file('$CURDIR/data_parquet/02588_data.parquet') where exchange_ts = 1670964058771367936"
+$CLICKHOUSE_LOCAL -q "select exchange_ts from file('$CURDIR/data_parquet/02588_data.parquet') where exchange_ts = 1670964058771367936"
+$CLICKHOUSE_LOCAL -q "select exchange_ts, market, product from file('$CURDIR/data_parquet/02588_data.parquet') where exchange_ts = 1670946478544048640"
+