Coverage Report

Created: 2026-08-06 18:32

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
be/src/format/table/paimon_jni_reader.cpp
Line
Count
Source
1
// Licensed to the Apache Software Foundation (ASF) under one
2
// or more contributor license agreements.  See the NOTICE file
3
// distributed with this work for additional information
4
// regarding copyright ownership.  The ASF licenses this file
5
// to you under the Apache License, Version 2.0 (the
6
// "License"); you may not use this file except in compliance
7
// with the License.  You may obtain a copy of the License at
8
//
9
//   http://www.apache.org/licenses/LICENSE-2.0
10
//
11
// Unless required by applicable law or agreed to in writing,
12
// software distributed under the License is distributed on an
13
// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
14
// KIND, either express or implied.  See the License for the
15
// specific language governing permissions and limitations
16
// under the License.
17
18
#include "paimon_jni_reader.h"
19
20
#include <map>
21
#include <string_view>
22
#include <vector>
23
24
#include "core/types.h"
25
#include "format/jni/jni_data_bridge.h"
26
#include "runtime/descriptors.h"
27
#include "runtime/exec_env.h"
28
#include "runtime/runtime_state.h"
29
#include "storage/options.h"
30
#include "util/string_util.h"
31
#include "util/uid_util.h"
32
33
namespace doris {
34
class RuntimeProfile;
35
class RuntimeState;
36
class Block;
37
} // namespace doris
38
39
namespace doris {
40
41
namespace {
42
constexpr std::string_view PAIMON_JNI_SCANNER_IO_TMP_DIR = "paimon_jni_scanner_io_tmp";
43
} // namespace
44
45
const std::string PaimonJniReader::PAIMON_OPTION_PREFIX = "paimon.";
46
const std::string PaimonJniReader::HADOOP_OPTION_PREFIX = "hadoop.";
47
const std::string PaimonJniReader::DORIS_ENABLE_JNI_IO_MANAGER = "jni.enable_jni_io_manager";
48
const std::string PaimonJniReader::DORIS_JNI_IO_MANAGER_TMP_DIR = "jni.io_manager.tmp_dir";
49
50
PaimonJniReader::PaimonJniReader(const std::vector<SlotDescriptor*>& file_slot_descs,
51
                                 RuntimeState* state, RuntimeProfile* profile,
52
                                 const TFileRangeDesc& range,
53
                                 const TFileScanRangeParams* range_params)
54
2
        : JniReader(
55
2
                  file_slot_descs, state, profile, "org/apache/doris/paimon/PaimonJniScanner",
56
2
                  [&]() {
57
2
                      std::vector<std::string> column_names;
58
2
                      std::vector<std::string> column_types;
59
2
                      std::vector<std::string> encoded_column_types;
60
2
                      for (const auto& desc : file_slot_descs) {
61
0
                          column_names.emplace_back(desc->col_name());
62
0
                          column_types.emplace_back(
63
0
                                  JniDataBridge::get_jni_type_with_different_string(desc->type()));
64
0
                          encoded_column_types.emplace_back(
65
0
                                  JniDataBridge::get_jni_type_with_encoded_struct_fields(
66
0
                                          desc->type()));
67
0
                      }
68
2
                      const auto& paimon_params = range.table_format_params.paimon_params;
69
2
                      std::map<String, String> params;
70
2
                      params["paimon_split"] = paimon_params.paimon_split;
71
2
                      if (range_params->__isset.paimon_predicate &&
72
2
                          !range_params->paimon_predicate.empty()) {
73
2
                          params["paimon_predicate"] = range_params->paimon_predicate;
74
2
                      } else if (paimon_params.__isset.paimon_predicate) {
75
0
                          params["paimon_predicate"] = paimon_params.paimon_predicate;
76
0
                      }
77
2
                      params["required_fields"] = join(column_names, ",");
78
2
                      params["columns_types"] = join(column_types, "#");
79
                      // V1 and V2 must publish the same safe schema protocol because session
80
                      // routing can select either producer for the same Paimon scanner.
81
2
                      params["required_fields_base64"] =
82
2
                              JniDataBridge::encode_schema_values(column_names);
83
2
                      params["columns_types_base64"] =
84
2
                              JniDataBridge::encode_schema_values(encoded_column_types);
85
2
                      params["time_zone"] = state->timezone();
86
2
                      if (range_params->__isset.serialized_table) {
87
2
                          params["serialized_table"] = range_params->serialized_table;
88
2
                      }
89
2
                      params["serialized_table_cache_key"] =
90
2
                              range_params->__isset.serialized_table_cache_key &&
91
2
                                              !range_params->serialized_table_cache_key.empty()
92
2
                                      ? range_params->serialized_table_cache_key
93
2
                                      : generate_uuid_string();
94
2
                      if (range_params->__isset.paimon_options &&
95
2
                          !range_params->paimon_options.empty()) {
96
0
                          for (const auto& kv : range_params->paimon_options) {
97
0
                              params[PAIMON_OPTION_PREFIX + kv.first] = kv.second;
98
0
                          }
99
2
                      } else if (paimon_params.__isset.paimon_options) {
100
0
                          for (const auto& kv : paimon_params.paimon_options) {
101
0
                              params[PAIMON_OPTION_PREFIX + kv.first] = kv.second;
102
0
                          }
103
0
                      }
104
2
                      const std::string enable_io_manager_key =
105
2
                              PAIMON_OPTION_PREFIX + DORIS_ENABLE_JNI_IO_MANAGER;
106
2
                      const std::string io_manager_tmp_dir_key =
107
2
                              PAIMON_OPTION_PREFIX + DORIS_JNI_IO_MANAGER_TMP_DIR;
108
2
                      auto enable_io_manager_it = params.find(enable_io_manager_key);
109
2
                      if (enable_io_manager_it != params.end() &&
110
2
                          iequal(enable_io_manager_it->second, "true") &&
111
2
                          params.find(io_manager_tmp_dir_key) == params.end()) {
112
0
                          std::vector<std::string> tmp_dirs;
113
0
                          for (const auto& store_path : state->exec_env()->store_paths()) {
114
0
                              tmp_dirs.push_back(store_path.path + "/" +
115
0
                                                 std::string(PAIMON_JNI_SCANNER_IO_TMP_DIR));
116
0
                          }
117
0
                          DORIS_CHECK(!tmp_dirs.empty());
118
                          // Paimon's IOManager creates and later removes its own paimon-* child
119
                          // directory under these Doris storage-root scoped parent directories.
120
0
                          params[io_manager_tmp_dir_key] = join(tmp_dirs, ":");
121
0
                      }
122
2
                      if (range_params->__isset.properties && !range_params->properties.empty()) {
123
0
                          for (const auto& kv : range_params->properties) {
124
0
                              params[HADOOP_OPTION_PREFIX + kv.first] = kv.second;
125
0
                          }
126
2
                      } else if (paimon_params.__isset.hadoop_conf) {
127
0
                          for (const auto& kv : paimon_params.hadoop_conf) {
128
0
                              params[HADOOP_OPTION_PREFIX + kv.first] = kv.second;
129
0
                          }
130
0
                      }
131
2
                      return params;
132
2
                  }(),
133
2
                  [&]() {
134
2
                      std::vector<std::string> names;
135
2
                      for (const auto& desc : file_slot_descs) {
136
0
                          names.emplace_back(desc->col_name());
137
0
                      }
138
2
                      return names;
139
2
                  }(),
140
2
                  range.__isset.self_split_weight ? range.self_split_weight : -1) {
141
2
    if (range.table_format_params.__isset.table_level_row_count) {
142
2
        _remaining_table_level_row_count = range.table_format_params.table_level_row_count;
143
2
    } else {
144
0
        _remaining_table_level_row_count = -1;
145
0
    }
146
2
}
147
148
0
Status PaimonJniReader::_do_get_next_block(Block* block, size_t* read_rows, bool* eof) {
149
0
    if (_push_down_agg_type == TPushAggOp::type::COUNT && _remaining_table_level_row_count >= 0) {
150
0
        auto rows = std::min(_remaining_table_level_row_count,
151
0
                             (int64_t)_state->query_options().batch_size);
152
0
        _remaining_table_level_row_count -= rows;
153
0
        auto mutable_columns_guard = block->mutate_columns_scoped();
154
0
        auto& mutate_columns = mutable_columns_guard.mutable_columns();
155
0
        for (auto& col : mutate_columns) {
156
0
            col->resize(rows);
157
0
        }
158
0
        *read_rows = rows;
159
0
        if (_remaining_table_level_row_count == 0) {
160
0
            *eof = true;
161
0
        }
162
163
0
        return Status::OK();
164
0
    }
165
0
    return JniReader::_do_get_next_block(block, read_rows, eof);
166
0
}
167
168
0
Status PaimonJniReader::init_reader() {
169
0
    return open(_state, _profile);
170
0
}
171
} // namespace doris