/root/doris/be/src/util/load_util.cpp
Line | Count | Source (jump to first uncovered line) |
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 "util/load_util.h" |
19 | | |
20 | | #include <string> |
21 | | |
22 | | #include "util/string_util.h" |
23 | | |
24 | | namespace doris { |
25 | | void LoadUtil::parse_format(const std::string& format_str, const std::string& compress_type_str, |
26 | | TFileFormatType::type* format_type, |
27 | 20 | TFileCompressType::type* compress_type) { |
28 | 20 | if (format_str.empty()) { |
29 | 1 | parse_format("CSV", compress_type_str, format_type, compress_type); |
30 | 1 | return; |
31 | 1 | } |
32 | 19 | *compress_type = TFileCompressType::PLAIN; |
33 | 19 | *format_type = TFileFormatType::FORMAT_UNKNOWN; |
34 | 19 | if (iequal(format_str, "CSV")) { |
35 | 7 | if (compress_type_str.empty()) { |
36 | 1 | *format_type = TFileFormatType::FORMAT_CSV_PLAIN; |
37 | 6 | } else if (iequal(compress_type_str, "GZ")) { |
38 | 1 | *format_type = TFileFormatType::FORMAT_CSV_GZ; |
39 | 1 | *compress_type = TFileCompressType::GZ; |
40 | 5 | } else if (iequal(compress_type_str, "LZO")) { |
41 | 1 | *format_type = TFileFormatType::FORMAT_CSV_LZO; |
42 | 1 | *compress_type = TFileCompressType::LZO; |
43 | 4 | } else if (iequal(compress_type_str, "BZ2")) { |
44 | 1 | *format_type = TFileFormatType::FORMAT_CSV_BZ2; |
45 | 1 | *compress_type = TFileCompressType::BZ2; |
46 | 3 | } else if (iequal(compress_type_str, "LZ4")) { |
47 | 1 | *format_type = TFileFormatType::FORMAT_CSV_LZ4FRAME; |
48 | 1 | *compress_type = TFileCompressType::LZ4FRAME; |
49 | 2 | } else if (iequal(compress_type_str, "LZ4_BLOCK")) { |
50 | 0 | *format_type = TFileFormatType::FORMAT_CSV_LZ4BLOCK; |
51 | 0 | *compress_type = TFileCompressType::LZ4BLOCK; |
52 | 2 | } else if (iequal(compress_type_str, "LZOP")) { |
53 | 1 | *format_type = TFileFormatType::FORMAT_CSV_LZOP; |
54 | 1 | *compress_type = TFileCompressType::LZO; |
55 | 1 | } else if (iequal(compress_type_str, "SNAPPY_BLOCK")) { |
56 | 0 | *format_type = TFileFormatType::FORMAT_CSV_SNAPPYBLOCK; |
57 | 0 | *compress_type = TFileCompressType::SNAPPYBLOCK; |
58 | 1 | } else if (iequal(compress_type_str, "DEFLATE")) { |
59 | 1 | *format_type = TFileFormatType::FORMAT_CSV_DEFLATE; |
60 | 1 | *compress_type = TFileCompressType::DEFLATE; |
61 | 1 | } |
62 | 12 | } else if (iequal(format_str, "JSON")) { |
63 | 10 | if (compress_type_str.empty()) { |
64 | 1 | *format_type = TFileFormatType::FORMAT_JSON; |
65 | 9 | } else if (iequal(compress_type_str, "GZ")) { |
66 | 1 | *format_type = TFileFormatType::FORMAT_JSON; |
67 | 1 | *compress_type = TFileCompressType::GZ; |
68 | 8 | } else if (iequal(compress_type_str, "LZO")) { |
69 | 0 | *format_type = TFileFormatType::FORMAT_JSON; |
70 | 0 | *compress_type = TFileCompressType::LZO; |
71 | 8 | } else if (iequal(compress_type_str, "BZ2")) { |
72 | 1 | *format_type = TFileFormatType::FORMAT_JSON; |
73 | 1 | *compress_type = TFileCompressType::BZ2; |
74 | 7 | } else if (iequal(compress_type_str, "LZ4")) { |
75 | 1 | *format_type = TFileFormatType::FORMAT_JSON; |
76 | 1 | *compress_type = TFileCompressType::LZ4FRAME; |
77 | 6 | } else if (iequal(compress_type_str, "LZ4_BLOCK")) { |
78 | 1 | *format_type = TFileFormatType::FORMAT_JSON; |
79 | 1 | *compress_type = TFileCompressType::LZ4BLOCK; |
80 | 5 | } else if (iequal(compress_type_str, "LZOP")) { |
81 | 2 | *format_type = TFileFormatType::FORMAT_JSON; |
82 | 2 | *compress_type = TFileCompressType::LZO; |
83 | 3 | } else if (iequal(compress_type_str, "SNAPPY_BLOCK")) { |
84 | 1 | *format_type = TFileFormatType::FORMAT_JSON; |
85 | 1 | *compress_type = TFileCompressType::SNAPPYBLOCK; |
86 | 2 | } else if (iequal(compress_type_str, "DEFLATE")) { |
87 | 1 | *format_type = TFileFormatType::FORMAT_JSON; |
88 | 1 | *compress_type = TFileCompressType::DEFLATE; |
89 | 1 | } |
90 | 10 | } else if (iequal(format_str, "PARQUET")) { |
91 | 1 | *format_type = TFileFormatType::FORMAT_PARQUET; |
92 | 1 | } else if (iequal(format_str, "ORC")) { |
93 | 1 | *format_type = TFileFormatType::FORMAT_ORC; |
94 | 1 | } else if (iequal(format_str, "WAL")) { |
95 | 0 | *format_type = TFileFormatType::FORMAT_WAL; |
96 | 0 | } else if (iequal(format_str, "ARROW")) { |
97 | 0 | *format_type = TFileFormatType::FORMAT_ARROW; |
98 | 0 | } |
99 | 19 | return; |
100 | 20 | } |
101 | | |
102 | 14 | bool LoadUtil::is_format_support_streaming(TFileFormatType::type format) { |
103 | 14 | switch (format) { |
104 | 1 | case TFileFormatType::FORMAT_CSV_PLAIN: |
105 | 2 | case TFileFormatType::FORMAT_CSV_BZ2: |
106 | 3 | case TFileFormatType::FORMAT_CSV_DEFLATE: |
107 | 4 | case TFileFormatType::FORMAT_CSV_GZ: |
108 | 5 | case TFileFormatType::FORMAT_CSV_LZ4FRAME: |
109 | 5 | case TFileFormatType::FORMAT_CSV_LZ4BLOCK: |
110 | 6 | case TFileFormatType::FORMAT_CSV_LZO: |
111 | 7 | case TFileFormatType::FORMAT_CSV_LZOP: |
112 | 8 | case TFileFormatType::FORMAT_JSON: |
113 | 8 | case TFileFormatType::FORMAT_WAL: |
114 | 8 | case TFileFormatType::FORMAT_ARROW: |
115 | 8 | return true; |
116 | 6 | default: |
117 | 6 | return false; |
118 | 14 | } |
119 | 0 | return false; |
120 | 14 | } |
121 | | } // namespace doris |