be/src/udf/python/python_env.h
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 | | #pragma once |
19 | | |
20 | | #include <gen_cpp/BackendService_types.h> |
21 | | |
22 | | #include <filesystem> |
23 | | #include <utility> |
24 | | |
25 | | #include "common/status.h" |
26 | | |
27 | | namespace doris { |
28 | | |
29 | | namespace fs = std::filesystem; |
30 | | |
31 | | enum class PythonEnvType { CONDA, VENV }; |
32 | | |
33 | | struct PythonVersion { |
34 | | std::string full_version; // e.g. "3.9.16" |
35 | | std::string base_path; // e.g. "/root/anaconda3/envs/python3.9" |
36 | | std::string executable_path; // e.g. "{base_path}/bin/python3" |
37 | | |
38 | 8 | PythonVersion() = default; |
39 | | |
40 | | explicit PythonVersion(std::string full_version, std::string base_path, |
41 | | std::string executable_path) |
42 | 54 | : full_version(std::move(full_version)), |
43 | 54 | base_path(std::move(base_path)), |
44 | 54 | executable_path(std::move(executable_path)) {} |
45 | | |
46 | 17 | bool operator==(const PythonVersion& other) const { |
47 | 17 | return full_version == other.full_version && base_path == other.base_path && |
48 | 17 | executable_path == other.executable_path; |
49 | 17 | } |
50 | | |
51 | 1 | const std::string& get_base_path() const { return base_path; } |
52 | | |
53 | 58 | const std::string& get_executable_path() const { return executable_path; } |
54 | | |
55 | 28 | bool is_valid() const { |
56 | 28 | return !full_version.empty() && !base_path.empty() && !executable_path.empty() && |
57 | 28 | fs::exists(base_path) && fs::exists(executable_path); |
58 | 28 | } |
59 | | |
60 | 26 | std::string to_string() const { |
61 | 26 | return fmt::format("[full_version: {}, base_path: {}, executable_path: {}]", full_version, |
62 | 26 | base_path, executable_path); |
63 | 26 | } |
64 | | }; |
65 | | |
66 | | struct PythonEnvironment { |
67 | | std::string env_name; // e.g. "base" or "myenv" |
68 | | PythonVersion python_version; |
69 | | |
70 | | PythonEnvironment(const std::string& name, const PythonVersion& python_version); |
71 | | |
72 | | std::string to_string() const; |
73 | | |
74 | | bool is_valid() const; |
75 | | |
76 | | static Status scan_from_conda_root_path(const fs::path& conda_root_path, |
77 | | std::vector<PythonEnvironment>* environments); |
78 | | |
79 | | static Status scan_from_venv_root_path(const fs::path& venv_root_path, |
80 | | const std::vector<std::string>& interpreter_paths, |
81 | | std::vector<PythonEnvironment>* environments); |
82 | | }; |
83 | | |
84 | | class PythonEnvScanner { |
85 | | public: |
86 | 13 | PythonEnvScanner(const fs::path& env_root_path) : _env_root_path(env_root_path) {} |
87 | | |
88 | 13 | virtual ~PythonEnvScanner() = default; |
89 | | |
90 | | virtual Status scan() = 0; |
91 | | |
92 | | Status get_versions(std::vector<PythonVersion>* versions) const; |
93 | | |
94 | | Status get_version(const std::string& runtime_version, PythonVersion* version) const; |
95 | | |
96 | 0 | const std::vector<PythonEnvironment>& get_envs() const { return _envs; } |
97 | | |
98 | 2 | std::string root_path() const { return _env_root_path.string(); } |
99 | | |
100 | | virtual PythonEnvType env_type() const = 0; |
101 | | |
102 | | virtual std::string to_string() const = 0; |
103 | | |
104 | | protected: |
105 | | fs::path _env_root_path; |
106 | | std::vector<PythonEnvironment> _envs; |
107 | | }; |
108 | | |
109 | | class CondaEnvScanner : public PythonEnvScanner { |
110 | | public: |
111 | 9 | CondaEnvScanner(const fs::path& python_root_path) : PythonEnvScanner(python_root_path) {} |
112 | | |
113 | 9 | ~CondaEnvScanner() override = default; |
114 | | |
115 | | Status scan() override; |
116 | | |
117 | | std::string to_string() const override; |
118 | | |
119 | 1 | PythonEnvType env_type() const override { return PythonEnvType::CONDA; } |
120 | | }; |
121 | | |
122 | | class VenvEnvScanner : public PythonEnvScanner { |
123 | | public: |
124 | | VenvEnvScanner(const fs::path& python_root_path, |
125 | | const std::vector<std::string>& interpreter_paths) |
126 | 4 | : PythonEnvScanner(python_root_path), _interpreter_paths(interpreter_paths) {} |
127 | | |
128 | 4 | ~VenvEnvScanner() override = default; |
129 | | |
130 | | Status scan() override; |
131 | | |
132 | | std::string to_string() const override; |
133 | | |
134 | 1 | PythonEnvType env_type() const override { return PythonEnvType::VENV; } |
135 | | |
136 | | private: |
137 | | std::vector<std::string> _interpreter_paths; |
138 | | }; |
139 | | |
140 | | class PythonVersionManager { |
141 | | public: |
142 | 0 | static PythonVersionManager& instance() { |
143 | 0 | static PythonVersionManager instance; |
144 | 0 | return instance; |
145 | 0 | } |
146 | | |
147 | | Status init(PythonEnvType env_type, const fs::path& python_root_path, |
148 | | const std::string& python_venv_interpreter_paths); |
149 | | |
150 | 0 | Status get_version(const std::string& runtime_version, PythonVersion* version) const { |
151 | 0 | return _env_scanner->get_version(runtime_version, version); |
152 | 0 | } |
153 | | |
154 | 0 | const std::vector<PythonEnvironment>& get_envs() const { return _env_scanner->get_envs(); } |
155 | | |
156 | 0 | PythonEnvType env_type() const { return _env_scanner->env_type(); } |
157 | | |
158 | 0 | std::string to_string() const { return _env_scanner->to_string(); } |
159 | | |
160 | | std::vector<TPythonEnvInfo> env_infos_to_thrift() const; |
161 | | |
162 | | std::vector<TPythonPackageInfo> package_infos_to_thrift( |
163 | | const std::vector<std::pair<std::string, std::string>>& packages) const; |
164 | | |
165 | | private: |
166 | | std::unique_ptr<PythonEnvScanner> _env_scanner; |
167 | | }; |
168 | | |
169 | | // List installed pip packages for a given Python version. |
170 | | // Returns pairs of (package_name, version). |
171 | | Status list_installed_packages(const PythonVersion& version, |
172 | | std::vector<std::pair<std::string, std::string>>* packages); |
173 | | |
174 | | } // namespace doris |
175 | | |
176 | | namespace std { |
177 | | template <> |
178 | | struct hash<doris::PythonVersion> { |
179 | 50 | size_t operator()(const doris::PythonVersion& v) const noexcept { |
180 | 50 | return hash<string> {}(v.full_version); |
181 | 50 | } |
182 | | }; |
183 | | } // namespace std |