SaintBacchus commented on code in PR #10512: URL: https://github.com/apache/doris/pull/10512#discussion_r914553110
########## be/src/exec/arrow/parquet_row_group_reader.cpp: ########## @@ -0,0 +1,567 @@ +// Licensed to the Apache Software Foundation (ASF) under one +// or more contributor license agreements. See the NOTICE file +// distributed with this work for additional information +// regarding copyright ownership. The ASF licenses this file +// to you under the Apache License, Version 2.0 (the +// "License"); you may not use this file except in compliance +// with the License. You may obtain a copy of the License at +// +// http://www.apache.org/licenses/LICENSE-2.0 +// +// Unless required by applicable law or agreed to in writing, +// software distributed under the License is distributed on an +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +// KIND, either express or implied. See the License for the +// specific language governing permissions and limitations +// under the License. + +#include "exec/arrow/parquet_row_group_reader.h" + +#include <exprs/expr_context.h> +#include <exprs/in_predicate.h> +#include <parquet/encoding.h> + +#include <cstring> + +#define _PLAIN_DECODE(T, value, min_bytes, max_bytes, out_value, out_min, out_max) \ + const T out_min = reinterpret_cast<const T*>(min_bytes)[0]; \ + const T out_max = reinterpret_cast<const T*>(max_bytes)[0]; \ + T out_value = *((T*)value); + +#define _PLAIN_DECODE_SINGLE(T, value, bytes, conjunct_value, out) \ + const T out = reinterpret_cast<const T*>(bytes)[0]; \ + T conjunct_value = *((T*)value); + +#define _FILTER_GROUP_BY_EQ_PRED(conjunct_value, min, max) \ + if (conjunct_value < min || conjunct_value > max) { \ + return true; \ + } + +#define _FILTER_GROUP_BY_GT_PRED(conjunct_value, max) \ + if (max <= conjunct_value) { \ + return true; \ + } + +#define _FILTER_GROUP_BY_GE_PRED(conjunct_value, max) \ + if (max < conjunct_value) { \ + return true; \ + } + +#define _FILTER_GROUP_BY_LT_PRED(conjunct_value, min) \ + if (min >= conjunct_value) { \ + return true; \ + } + +#define _FILTER_GROUP_BY_LE_PRED(conjunct_value, min) \ + if (min > conjunct_value) { \ + return true; \ + } + +#define _FILTER_GROUP_BY_IN(T, in_pred_values, min_bytes, max_bytes) \ + std::vector<T> in_values; \ + for (auto val : in_pred_values) { \ + T value = reinterpret_cast<T*>(val)[0]; \ + in_values.emplace_back(value); \ + } \ + if (in_values.empty()) { \ + return false; \ + } \ + std::sort(in_values.begin(), in_values.end()); \ + T in_min = in_values.front(); \ + T in_max = in_values.back(); \ + const T group_min = reinterpret_cast<const T*>(min_bytes)[0]; \ + const T group_max = reinterpret_cast<const T*>(max_bytes)[0]; \ + if (in_max < group_min || in_min > group_max) { \ + return true; \ + } + +namespace doris { + +RowGroupReader::RowGroupReader(RuntimeProfile* profile, + const std::vector<ExprContext*>& conjunct_ctxs, + std::shared_ptr<parquet::FileMetaData>& file_metadata, + ParquetReaderWrap* parent) + : _conjunct_ctxs(conjunct_ctxs), + _file_metadata(file_metadata), + _profile(profile), + _parent(parent) {} + +RowGroupReader::~RowGroupReader() { + _slot_conjuncts.clear(); + _filter_group.clear(); +} + +Status RowGroupReader::init_filter_groups(const TupleDescriptor* tuple_desc, + const std::map<std::string, int>& map_column, + const std::vector<int>& include_column_ids) { + std::unordered_set<int> parquet_column_ids(include_column_ids.begin(), + include_column_ids.end()); + _init_conjuncts(tuple_desc, map_column, parquet_column_ids); + int total_group = _file_metadata->num_row_groups(); + _parent->statistics()->total_groups = total_group; + _parent->statistics()->total_rows = _file_metadata->num_rows(); + + int64_t filtered_num_row_groups = 0; + int64_t filtered_num_rows = 0; + int64_t filtered_total_byte_size = 0; + bool need_filter = false; + for (int row_group_id = 0; row_group_id < total_group; row_group_id++) { + auto row_group_meta = _file_metadata->RowGroup(row_group_id); + for (SlotId slot_id = 0; slot_id < tuple_desc->slots().size(); slot_id++) { + const std::string& col_name = tuple_desc->slots()[slot_id]->col_name(); + auto col_iter = map_column.find(col_name); + if (col_iter == map_column.end()) { + continue; + } + int parquet_col_id = col_iter->second; + if (parquet_column_ids.end() == parquet_column_ids.find(parquet_col_id)) { + // Column not exist in parquet file + continue; + } + auto slot_iter = _slot_conjuncts.find(slot_id); + if (slot_iter == _slot_conjuncts.end()) { + continue; + } + auto statistic = row_group_meta->ColumnChunk(parquet_col_id)->statistics(); + if (!statistic->HasMinMax()) { + continue; + } + // Min-max of statistic is plain-encoded value + const std::string& min = statistic->EncodeMin(); + const std::string& max = statistic->EncodeMax(); + + need_filter = _determine_filter_row_group(slot_iter->second, min, max); Review Comment: is this `need_filter` determined by the last row_group? -- 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: commits-unsubscr...@doris.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@doris.apache.org For additional commands, e-mail: commits-h...@doris.apache.org
