Coverage Report

Created: 2026-10-14 20:14

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
be/src/service/http/http_client.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 <curl/curl.h>
21
#include <curl/system.h>
22
#include <stdint.h>
23
24
#include <cstdio>
25
#include <functional>
26
#include <string>
27
#include <unordered_set>
28
29
#include "common/status.h"
30
#include "service/http/http_headers.h"
31
#include "service/http/http_method.h"
32
33
namespace doris {
34
35
// Helper class to access HTTP resource
36
class HttpClient {
37
public:
38
    HttpClient();
39
    ~HttpClient();
40
41
    // you can call this function to execute HTTP request with retry,
42
    // if callback return OK, this function will end and return OK.
43
    // This function will return FAIL if three are more than retry_times
44
    // that callback return FAIL.
45
    static Status execute_with_retry(int retry_times, int sleep_time,
46
                                     const std::function<Status(HttpClient*)>& callback);
47
48
    // this function must call before other function,
49
    // you can call this multiple times to reuse this object
50
    Status init(const std::string& url, bool set_fail_on_error = true);
51
52
    // Release request-owned options and callbacks while retaining reusable connections.
53
    void reset();
54
55
    void set_method(HttpMethod method);
56
57
21
    void set_basic_auth(const std::string& user, const std::string& passwd) {
58
21
        curl_easy_setopt(_curl, CURLOPT_HTTPAUTH, CURLAUTH_BASIC);
59
21
        curl_easy_setopt(_curl, CURLOPT_USERNAME, user.c_str());
60
21
        curl_easy_setopt(_curl, CURLOPT_PASSWORD, passwd.c_str());
61
21
    }
62
63
    // Auth-Token: xxxx
64
2
    void set_auth_token(const std::string& token) {
65
2
        std::string scratch_str = HttpHeaders::AUTH_TOKEN + ": " + token;
66
2
        _header_list = curl_slist_append(_header_list, scratch_str.c_str());
67
2
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
68
2
    }
69
70
    // Authorization: xxxx
71
0
    void set_authorization(const std::string& auth) {
72
0
        std::string scratch_str = std::string(HttpHeaders::AUTHORIZATION) + ": " + auth;
73
0
        _header_list = curl_slist_append(_header_list, scratch_str.c_str());
74
0
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
75
0
    }
76
77
    // content_type such as "application/json"
78
301
    void set_content_type(const std::string content_type) {
79
301
        std::string scratch_str = "Content-Type: " + content_type;
80
301
        _header_list = curl_slist_append(_header_list, scratch_str.c_str());
81
301
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
82
301
    }
83
84
295
    void set_payload(const std::string& post_body) {
85
295
        curl_easy_setopt(_curl, CURLOPT_POSTFIELDSIZE, (long)post_body.length());
86
295
        curl_easy_setopt(_curl, CURLOPT_COPYPOSTFIELDS, post_body.c_str());
87
295
    }
88
89
    // Currently, only fake SSL configurations are supported
90
0
    void use_untrusted_ssl() {
91
0
        curl_easy_setopt(_curl, CURLOPT_SSL_VERIFYPEER, 0L);
92
0
        curl_easy_setopt(_curl, CURLOPT_SSL_VERIFYHOST, 0L);
93
0
    }
94
95
    void set_speed_limit();
96
97
    // key: value
98
265
    void set_header(const std::string& key, const std::string& value) {
99
265
        std::string header = key + ": " + value;
100
265
        _header_list = curl_slist_append(_header_list, header.c_str());
101
265
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
102
265
    }
103
104
0
    std::string get_response_content_type() {
105
0
        char* ct = nullptr;
106
0
        auto code = curl_easy_getinfo(_curl, CURLINFO_CONTENT_TYPE, &ct);
107
0
        if (code == CURLE_OK && ct != nullptr) {
108
0
            return ct;
109
0
        }
110
0
        return std::string();
111
0
    }
112
113
    // Set the long gohead parameter to 1L to continue send authentication (user+password)
114
    // credentials when following locations, even when hostname changed.
115
0
    void set_unrestricted_auth(int gohead) {
116
0
        curl_easy_setopt(_curl, CURLOPT_UNRESTRICTED_AUTH, gohead);
117
0
    }
118
119
333
    void set_timeout_ms(int64_t timeout_ms) {
120
333
        curl_easy_setopt(_curl, CURLOPT_TIMEOUT_MS, timeout_ms);
121
333
    }
122
123
    // Register a callback that libcurl polls while the request is in flight: often while
124
    // data is flowing, and about once per second when the connection is idle. Returning
125
    // true aborts the transfer right away, so `execute()` fails instead of blocking until
126
    // CURLOPT_TIMEOUT_MS expires.
127
    //
128
    // This is for callers that must be able to give up on a request which may never be
129
    // answered, e.g. a background worker that is being stopped while the http service
130
    // serving its request is going away. Must be called after init(), which resets all
131
    // curl options. Passing an empty callback clears a previously registered one.
132
    void set_abort_callback(std::function<bool()> callback);
133
134
    // used to get content length
135
    // return -1 as error
136
6
    Status get_content_length(uint64_t* length) const {
137
6
        curl_off_t cl;
138
6
        auto code = curl_easy_getinfo(_curl, CURLINFO_CONTENT_LENGTH_DOWNLOAD_T, &cl);
139
6
        if (!code) {
140
6
            if (cl < 0) {
141
0
                return Status::InternalError(
142
0
                        fmt::format("failed to get content length, it should be a positive value, "
143
0
                                    "actual is : {}",
144
0
                                    cl));
145
0
            }
146
6
            *length = (uint64_t)cl;
147
6
            return Status::OK();
148
6
        }
149
0
        return Status::InternalError("failed to get content length. err code: {}", code);
150
6
    }
151
152
    // Get the value of the header CONTENT-MD5. The output is empty if no such header exists.
153
    Status get_content_md5(std::string* md5) const;
154
155
247
    CURLcode get_last_curl_code() const { return _last_curl_code; }
156
157
    // Read the final response header; clear value when the header is absent.
158
    Status get_response_header(const char* name, std::string* value) const;
159
160
279
    long get_http_status() const {
161
279
        long code;
162
279
        curl_easy_getinfo(_curl, CURLINFO_RESPONSE_CODE, &code);
163
279
        return code;
164
279
    }
165
166
    // execute a head method
167
2
    Status head() {
168
2
        set_method(HEAD);
169
2
        return execute();
170
2
    }
171
172
    // helper function to download a file, you can call this function to download
173
    // a file to local_path
174
    Status download(const std::string& local_path);
175
    Status download_multi_files(const std::string& local_dir,
176
                                const std::unordered_set<std::string>& expected_files);
177
178
    Status execute_post_request(const std::string& payload, std::string* response);
179
180
    Status execute_put_request(const std::string& payload, std::string* response);
181
182
    Status execute_delete_request(const std::string& payload, std::string* response);
183
184
    // execute a simple method, and its response is saved in response argument
185
    Status execute(std::string* response);
186
187
    // execute remote call action
188
    Status execute(const std::function<bool(const void* data, size_t length)>& callback = {});
189
190
    // execute remote call action with retry, like execute_with_retry but keep the http client instance
191
    Status execute(int retry_times, int sleep_time,
192
                   const std::function<Status(HttpClient*)>& callback);
193
194
    size_t on_response_data(const void* data, size_t length);
195
196
    // The file name of the variant column with the inverted index contains %
197
    // such as: 020000000000003f624c4c322c568271060f9b5b274a4a95_0_10133@properties%2Emessage.idx
198
    //  {rowset_id}_{seg_num}_{index_id}_{variant_column_name}{%2E}{extracted_column_name}.idx
199
    // We need to handle %, otherwise it will cause an HTTP 404 error.
200
    // Because the percent ("%") character serves as the indicator for percent-encoded octets,
201
    // it must be percent-encoded as "%25" for that octet to be used as data within a URI.
202
    // https://datatracker.ietf.org/doc/html/rfc3986
203
    Status _escape_url(const std::string& url, std::string* escaped_url);
204
205
2
    void set_range(size_t offset, size_t length) {
206
2
        std::string range_header = "Range: bytes=" + std::to_string(offset) + "-" +
207
2
                                   std::to_string(offset + length - 1);
208
2
        _header_list = curl_slist_append(_header_list, range_header.c_str());
209
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
210
2
    }
211
212
private:
213
    static std::string _get_ca_certificate_path(const std::string& configured_paths);
214
    Status _set_ca_certificate_from_config();
215
    const char* _to_errmsg(CURLcode code) const;
216
    const char* _get_url() const;
217
218
private:
219
    CURL* _curl = nullptr;
220
    CURLcode _last_curl_code = CURLE_OK;
221
    using HttpCallback = std::function<bool(const void* data, size_t length)>;
222
    const HttpCallback* _callback = nullptr;
223
    std::function<bool()> _abort_callback;
224
    char _error_buf[CURL_ERROR_SIZE];
225
    curl_slist* _header_list = nullptr;
226
    HttpMethod _method = GET;
227
};
228
229
} // namespace doris