Coverage Report

Created: 2026-07-04 02:13

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
be/src/format/parquet/decoder.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 "format/parquet/decoder.h"
19
20
#include <cctz/time_zone.h>
21
#include <gen_cpp/parquet_types.h>
22
23
#include "format/parquet/bool_plain_decoder.h"
24
#include "format/parquet/bool_rle_decoder.h"
25
#include "format/parquet/byte_array_dict_decoder.h"
26
#include "format/parquet/byte_array_plain_decoder.h"
27
#include "format/parquet/byte_stream_split_decoder.h"
28
#include "format/parquet/delta_bit_pack_decoder.h"
29
#include "format/parquet/fix_length_dict_decoder.hpp"
30
#include "format/parquet/fix_length_plain_decoder.h"
31
32
namespace doris {
33
Status Decoder::get_decoder(tparquet::Type::type type, tparquet::Encoding::type encoding,
34
8.51k
                            std::unique_ptr<Decoder>& decoder) {
35
8.51k
    switch (encoding) {
36
1.28k
    case tparquet::Encoding::PLAIN:
37
1.28k
        switch (type) {
38
298
        case tparquet::Type::BOOLEAN:
39
298
            decoder.reset(new BoolPlainDecoder());
40
298
            break;
41
323
        case tparquet::Type::BYTE_ARRAY:
42
323
            decoder.reset(new ByteArrayPlainDecoder());
43
323
            break;
44
412
        case tparquet::Type::INT32:
45
490
        case tparquet::Type::INT64:
46
502
        case tparquet::Type::INT96:
47
554
        case tparquet::Type::FLOAT:
48
609
        case tparquet::Type::DOUBLE:
49
666
        case tparquet::Type::FIXED_LEN_BYTE_ARRAY:
50
666
            decoder.reset(new FixLengthPlainDecoder());
51
666
            break;
52
0
        default:
53
0
            return Status::InternalError("Unsupported type {}(encoding={}) in parquet decoder",
54
0
                                         tparquet::to_string(type), tparquet::to_string(encoding));
55
1.28k
        }
56
1.28k
        break;
57
7.02k
    case tparquet::Encoding::RLE_DICTIONARY:
58
7.02k
        switch (type) {
59
0
        case tparquet::Type::BOOLEAN:
60
0
            return Status::InternalError("Bool type can't has dictionary page");
61
2.16k
        case tparquet::Type::BYTE_ARRAY:
62
2.16k
            decoder.reset(new ByteArrayDictDecoder());
63
2.16k
            break;
64
2.68k
        case tparquet::Type::INT32:
65
2.68k
            decoder.reset(new FixLengthDictDecoder<tparquet::Type::INT32>());
66
2.68k
            break;
67
851
        case tparquet::Type::INT64:
68
851
            decoder.reset(new FixLengthDictDecoder<tparquet::Type::INT64>());
69
851
            break;
70
76
        case tparquet::Type::INT96:
71
76
            decoder.reset(new FixLengthDictDecoder<tparquet::Type::INT96>());
72
76
            break;
73
145
        case tparquet::Type::FLOAT:
74
145
            decoder.reset(new FixLengthDictDecoder<tparquet::Type::FLOAT>());
75
145
            break;
76
896
        case tparquet::Type::DOUBLE:
77
896
            decoder.reset(new FixLengthDictDecoder<tparquet::Type::DOUBLE>());
78
896
            break;
79
208
        case tparquet::Type::FIXED_LEN_BYTE_ARRAY:
80
208
            decoder.reset(new FixLengthDictDecoder<tparquet::Type::FIXED_LEN_BYTE_ARRAY>());
81
208
            break;
82
0
        default:
83
0
            return Status::InternalError("Unsupported type {}(encoding={}) in parquet decoder",
84
0
                                         tparquet::to_string(type), tparquet::to_string(encoding));
85
7.02k
        }
86
7.02k
        break;
87
7.02k
    case tparquet::Encoding::RLE:
88
0
        switch (type) {
89
0
        case tparquet::Type::BOOLEAN:
90
0
            decoder.reset(new BoolRLEDecoder());
91
0
            break;
92
0
        default:
93
0
            return Status::InternalError("Unsupported type {}(encoding={}) in parquet decoder",
94
0
                                         tparquet::to_string(type), tparquet::to_string(encoding));
95
0
        }
96
0
        break;
97
195
    case tparquet::Encoding::DELTA_BINARY_PACKED:
98
        // Supports only INT32 and INT64.
99
195
        switch (type) {
100
0
        case tparquet::Type::INT32:
101
0
            decoder.reset(new DeltaBitPackDecoder<int32_t>());
102
0
            break;
103
195
        case tparquet::Type::INT64:
104
195
            decoder.reset(new DeltaBitPackDecoder<int64_t>());
105
195
            break;
106
0
        default:
107
0
            return Status::InternalError("DELTA_BINARY_PACKED only supports INT32 and INT64");
108
195
        }
109
195
        break;
110
195
    case tparquet::Encoding::DELTA_BYTE_ARRAY:
111
0
        switch (type) {
112
0
        case tparquet::Type::BYTE_ARRAY:
113
0
        case tparquet::Type::FIXED_LEN_BYTE_ARRAY:
114
0
            decoder.reset(new DeltaByteArrayDecoder());
115
0
            break;
116
0
        default:
117
0
            return Status::InternalError(
118
0
                    "DELTA_BYTE_ARRAY only supports BYTE_ARRAY, FIXED_LEN_BYTE_ARRAY.");
119
0
        }
120
0
        break;
121
0
    case tparquet::Encoding::DELTA_LENGTH_BYTE_ARRAY:
122
0
        switch (type) {
123
0
        case tparquet::Type::BYTE_ARRAY:
124
0
            decoder.reset(new DeltaLengthByteArrayDecoder());
125
0
            break;
126
0
        default:
127
0
            return Status::InternalError("DELTA_LENGTH_BYTE_ARRAY only supports BYTE_ARRAY.");
128
0
        }
129
0
        break;
130
0
    case tparquet::Encoding::BYTE_STREAM_SPLIT:
131
0
        switch (type) {
132
0
        case tparquet::Type::INT32:
133
0
        case tparquet::Type::INT64:
134
0
        case tparquet::Type::INT96:
135
0
        case tparquet::Type::FLOAT:
136
0
        case tparquet::Type::DOUBLE:
137
0
        case tparquet::Type::FIXED_LEN_BYTE_ARRAY:
138
0
            decoder.reset(new ByteStreamSplitDecoder());
139
0
            break;
140
0
        default:
141
0
            return Status::InternalError("Unsupported type {}(encoding={}) in parquet decoder",
142
0
                                         tparquet::to_string(type), tparquet::to_string(encoding));
143
0
        }
144
0
        break;
145
0
    default:
146
0
        return Status::InternalError("Unsupported encoding {}(type={}) in parquet decoder",
147
0
                                     tparquet::to_string(encoding), tparquet::to_string(type));
148
8.51k
    }
149
8.50k
    return Status::OK();
150
8.51k
}
151
152
} // namespace doris