Coverage Report

Created: 2026-10-15 15:43

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
struct HttpTransferMetrics {
36
    int64_t new_connections = 0;
37
    int64_t dns_time_ns = 0;
38
    int64_t connect_time_ns = 0;
39
    int64_t tls_time_ns = 0;
40
    int64_t first_byte_time_ns = 0;
41
    int64_t total_time_ns = 0;
42
};
43
44
// Helper class to access HTTP resource
45
class HttpClient {
46
public:
47
    HttpClient();
48
    ~HttpClient();
49
50
    // you can call this function to execute HTTP request with retry,
51
    // if callback return OK, this function will end and return OK.
52
    // This function will return FAIL if three are more than retry_times
53
    // that callback return FAIL.
54
    static Status execute_with_retry(int retry_times, int sleep_time,
55
                                     const std::function<Status(HttpClient*)>& callback);
56
57
    // this function must call before other function,
58
    // you can call this multiple times to reuse this object
59
    Status init(const std::string& url, bool set_fail_on_error = true);
60
61
    // Release request-owned options and callbacks while retaining reusable connections.
62
    void reset();
63
64
    void set_method(HttpMethod method);
65
66
21
    void set_basic_auth(const std::string& user, const std::string& passwd) {
67
21
        curl_easy_setopt(_curl, CURLOPT_HTTPAUTH, CURLAUTH_BASIC);
68
21
        curl_easy_setopt(_curl, CURLOPT_USERNAME, user.c_str());
69
21
        curl_easy_setopt(_curl, CURLOPT_PASSWORD, passwd.c_str());
70
21
    }
71
72
    // Auth-Token: xxxx
73
505
    void set_auth_token(const std::string& token) {
74
505
        std::string scratch_str = HttpHeaders::AUTH_TOKEN + ": " + token;
75
505
        _header_list = curl_slist_append(_header_list, scratch_str.c_str());
76
505
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
77
505
    }
78
79
    // Authorization: xxxx
80
20
    void set_authorization(const std::string& auth) {
81
20
        std::string scratch_str = std::string(HttpHeaders::AUTHORIZATION) + ": " + auth;
82
20
        _header_list = curl_slist_append(_header_list, scratch_str.c_str());
83
20
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
84
20
    }
85
86
    // content_type such as "application/json"
87
353
    void set_content_type(const std::string content_type) {
88
353
        std::string scratch_str = "Content-Type: " + content_type;
89
353
        _header_list = curl_slist_append(_header_list, scratch_str.c_str());
90
353
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
91
353
    }
92
93
347
    void set_payload(const std::string& post_body) {
94
347
        curl_easy_setopt(_curl, CURLOPT_POSTFIELDSIZE, (long)post_body.length());
95
347
        curl_easy_setopt(_curl, CURLOPT_COPYPOSTFIELDS, post_body.c_str());
96
347
    }
97
98
    // Currently, only fake SSL configurations are supported
99
0
    void use_untrusted_ssl() {
100
0
        curl_easy_setopt(_curl, CURLOPT_SSL_VERIFYPEER, 0L);
101
0
        curl_easy_setopt(_curl, CURLOPT_SSL_VERIFYHOST, 0L);
102
0
    }
103
104
    void set_speed_limit();
105
106
    // key: value
107
456
    void set_header(const std::string& key, const std::string& value) {
108
456
        std::string header = key + ": " + value;
109
456
        _header_list = curl_slist_append(_header_list, header.c_str());
110
456
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
111
456
    }
112
113
0
    std::string get_response_content_type() {
114
0
        char* ct = nullptr;
115
0
        auto code = curl_easy_getinfo(_curl, CURLINFO_CONTENT_TYPE, &ct);
116
0
        if (code == CURLE_OK && ct != nullptr) {
117
0
            return ct;
118
0
        }
119
0
        return std::string();
120
0
    }
121
122
    // Set the long gohead parameter to 1L to continue send authentication (user+password)
123
    // credentials when following locations, even when hostname changed.
124
0
    void set_unrestricted_auth(int gohead) {
125
0
        curl_easy_setopt(_curl, CURLOPT_UNRESTRICTED_AUTH, gohead);
126
0
    }
127
128
388
    void set_timeout_ms(int64_t timeout_ms) {
129
388
        curl_easy_setopt(_curl, CURLOPT_TIMEOUT_MS, timeout_ms);
130
388
    }
131
132
    // Register a callback that libcurl polls while the request is in flight: often while
133
    // data is flowing, and about once per second when the connection is idle. Returning
134
    // true aborts the transfer right away, so `execute()` fails instead of blocking until
135
    // CURLOPT_TIMEOUT_MS expires.
136
    //
137
    // This is for callers that must be able to give up on a request which may never be
138
    // answered, e.g. a background worker that is being stopped while the http service
139
    // serving its request is going away. Must be called after init(), which resets all
140
    // curl options. Passing an empty callback clears a previously registered one.
141
    void set_abort_callback(std::function<bool()> callback);
142
143
    // used to get content length
144
    // return -1 as error
145
6
    Status get_content_length(uint64_t* length) const {
146
6
        curl_off_t cl;
147
6
        auto code = curl_easy_getinfo(_curl, CURLINFO_CONTENT_LENGTH_DOWNLOAD_T, &cl);
148
6
        if (!code) {
149
6
            if (cl < 0) {
150
0
                return Status::InternalError(
151
0
                        fmt::format("failed to get content length, it should be a positive value, "
152
0
                                    "actual is : {}",
153
0
                                    cl));
154
0
            }
155
6
            *length = (uint64_t)cl;
156
6
            return Status::OK();
157
6
        }
158
0
        return Status::InternalError("failed to get content length. err code: {}", code);
159
6
    }
160
161
    // Get the value of the header CONTENT-MD5. The output is empty if no such header exists.
162
    Status get_content_md5(std::string* md5) const;
163
164
278
    CURLcode get_last_curl_code() const { return _last_curl_code; }
165
166
    // Metrics for the latest transfer, including failed requests. Read before reset/init.
167
    // DNS/first-byte/total are cumulative curl timings; connect/TLS are phase durations.
168
    HttpTransferMetrics get_transfer_metrics() const;
169
170
    // Read the final response header; clear value when the header is absent.
171
    Status get_response_header(const char* name, std::string* value) const;
172
173
332
    long get_http_status() const {
174
332
        long code;
175
332
        curl_easy_getinfo(_curl, CURLINFO_RESPONSE_CODE, &code);
176
332
        return code;
177
332
    }
178
179
    // execute a head method
180
2
    Status head() {
181
2
        set_method(HEAD);
182
2
        return execute();
183
2
    }
184
185
    // helper function to download a file, you can call this function to download
186
    // a file to local_path
187
    Status download(const std::string& local_path);
188
    Status download_multi_files(const std::string& local_dir,
189
                                const std::unordered_set<std::string>& expected_files);
190
191
    Status execute_post_request(const std::string& payload, std::string* response);
192
193
    Status execute_put_request(const std::string& payload, std::string* response);
194
195
    Status execute_delete_request(const std::string& payload, std::string* response);
196
197
    // execute a simple method, and its response is saved in response argument
198
    Status execute(std::string* response);
199
200
    // execute remote call action
201
    Status execute(const std::function<bool(const void* data, size_t length)>& callback = {});
202
203
    // execute remote call action with retry, like execute_with_retry but keep the http client instance
204
    Status execute(int retry_times, int sleep_time,
205
                   const std::function<Status(HttpClient*)>& callback);
206
207
    size_t on_response_data(const void* data, size_t length);
208
209
    // The file name of the variant column with the inverted index contains %
210
    // such as: 020000000000003f624c4c322c568271060f9b5b274a4a95_0_10133@properties%2Emessage.idx
211
    //  {rowset_id}_{seg_num}_{index_id}_{variant_column_name}{%2E}{extracted_column_name}.idx
212
    // We need to handle %, otherwise it will cause an HTTP 404 error.
213
    // Because the percent ("%") character serves as the indicator for percent-encoded octets,
214
    // it must be percent-encoded as "%25" for that octet to be used as data within a URI.
215
    // https://datatracker.ietf.org/doc/html/rfc3986
216
    Status _escape_url(const std::string& url, std::string* escaped_url);
217
218
2
    void set_range(size_t offset, size_t length) {
219
2
        std::string range_header = "Range: bytes=" + std::to_string(offset) + "-" +
220
2
                                   std::to_string(offset + length - 1);
221
2
        _header_list = curl_slist_append(_header_list, range_header.c_str());
222
        curl_easy_setopt(_curl, CURLOPT_HTTPHEADER, _header_list);
223
2
    }
224
225
private:
226
    static std::string _get_ca_certificate_path(const std::string& configured_paths);
227
    Status _set_ca_certificate_from_config();
228
    const char* _to_errmsg(CURLcode code) const;
229
    const char* _get_url() const;
230
231
private:
232
    CURL* _curl = nullptr;
233
    CURLcode _last_curl_code = CURLE_OK;
234
    using HttpCallback = std::function<bool(const void* data, size_t length)>;
235
    const HttpCallback* _callback = nullptr;
236
    std::function<bool()> _abort_callback;
237
    char _error_buf[CURL_ERROR_SIZE];
238
    curl_slist* _header_list = nullptr;
239
    HttpMethod _method = GET;
240
};
241
242
} // namespace doris