Coverage Report

Created: 2026-03-23 14:17

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
/root/doris/be/src/util/timezone_utils.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 "util/timezone_utils.h"
19
20
#include <cctz/civil_time.h>
21
#include <cctz/time_zone.h>
22
#include <fcntl.h>
23
#include <glog/logging.h>
24
#include <re2/re2.h>
25
#include <re2/stringpiece.h>
26
#include <sys/mman.h>
27
#include <sys/stat.h>
28
#include <sys/types.h>
29
#include <unistd.h>
30
31
#include <boost/algorithm/string.hpp>
32
#include <boost/algorithm/string/case_conv.hpp>
33
#include <cstdlib>
34
#include <filesystem>
35
#include <memory>
36
#include <string>
37
38
#include "common/exception.h"
39
#include "common/logging.h"
40
#include "common/status.h"
41
42
using boost::algorithm::to_lower_copy;
43
44
namespace fs = std::filesystem;
45
#include "common/compile_check_begin.h"
46
47
namespace doris {
48
49
using ZoneList = std::unordered_map<std::string, cctz::time_zone>;
50
51
RE2 time_zone_offset_format_reg(R"(^[+-]{1}\d{2}\:\d{2}$)"); // visiting is thread-safe
52
53
// for ut, make it never nullptr.
54
std::unique_ptr<ZoneList> lower_zone_cache_ = std::make_unique<ZoneList>();
55
56
const std::string TimezoneUtils::default_time_zone = "+08:00";
57
static const char* tzdir = "/usr/share/zoneinfo"; // default value, may change by TZDIR env var
58
59
6
void TimezoneUtils::clear_timezone_caches() {
60
6
    lower_zone_cache_->clear();
61
6
}
62
2
size_t TimezoneUtils::cache_size() {
63
2
    return lower_zone_cache_->size();
64
2
}
65
66
142k
static bool parse_save_name_tz(const std::string& tz_name) {
67
142k
    cctz::time_zone tz;
68
142k
    PROPAGATE_FALSE(cctz::load_time_zone(tz_name, &tz));
69
141k
    lower_zone_cache_->emplace(to_lower_copy(tz_name), tz);
70
141k
    return true;
71
142k
}
72
73
238
void TimezoneUtils::load_timezones_to_cache() {
74
238
    std::string base_str;
75
    // try get from system
76
238
    char* tzdir_env = std::getenv("TZDIR");
77
238
    if (tzdir_env && *tzdir_env) {
78
0
        tzdir = tzdir_env;
79
0
    }
80
81
238
    base_str = tzdir;
82
238
    base_str += '/';
83
84
238
    const auto root_path = fs::path {base_str};
85
238
    if (!exists(root_path)) {
86
0
        throw Exception(Status::FatalError("Cannot find system tzfile. Doris exiting!"));
87
0
    }
88
89
238
    std::set<std::string> ignore_paths = {"posix", "right"}; // duplications. ignore them.
90
91
148k
    for (fs::recursive_directory_iterator it {base_str}; it != end(it); it++) {
92
147k
        const auto& dir_entry = *it;
93
147k
        if (dir_entry.is_regular_file() ||
94
147k
            (dir_entry.is_symlink() && is_regular_file(read_symlink(dir_entry)))) {
95
142k
            auto tz_name = dir_entry.path().string().substr(base_str.length());
96
142k
            if (!parse_save_name_tz(tz_name)) {
97
1.19k
                LOG(WARNING) << "Meet illegal tzdata file: " << tz_name << ". skipped";
98
1.19k
            }
99
142k
        } else if (dir_entry.is_directory() && ignore_paths.contains(dir_entry.path().filename())) {
100
476
            it.disable_recursion_pending();
101
476
        }
102
147k
    }
103
    // some special cases. Z = Zulu. CST = Asia/Shanghai
104
238
    if (auto it = lower_zone_cache_->find("zulu"); it != lower_zone_cache_->end()) {
105
238
        lower_zone_cache_->emplace("z", it->second);
106
238
    }
107
238
    if (auto it = lower_zone_cache_->find("asia/shanghai"); it != lower_zone_cache_->end()) {
108
238
        lower_zone_cache_->emplace("cst", it->second);
109
238
    }
110
111
238
    lower_zone_cache_->erase("lmt"); // local mean time for every timezone
112
113
238
    load_offsets_to_cache();
114
238
    LOG(INFO) << "Preloaded" << lower_zone_cache_->size() << " timezones.";
115
238
}
116
117
20.6k
static std::string to_hour_string(int arg) {
118
20.6k
    if (arg < 0 && arg > -10) { // -9 to -1
119
6.88k
        return std::string {"-0"} + std::to_string(std::abs(arg));
120
13.7k
    } else if (arg >= 0 && arg < 10) { //0 to 9
121
7.65k
        return std::string {"0"} + std::to_string(arg);
122
7.65k
    }
123
6.12k
    return std::to_string(arg);
124
20.6k
}
125
126
255
void TimezoneUtils::load_offsets_to_cache() {
127
255
    static constexpr int supported_minutes[] = {0, 30, 45};
128
7.14k
    for (int hour = -12; hour <= +14; hour++) {
129
20.6k
        for (int minute : supported_minutes) {
130
20.6k
            char min_str[3];
131
20.6k
            snprintf(min_str, sizeof(min_str), "%02d", minute);
132
20.6k
            std::string offset_str = (hour >= 0 ? "+" : "") + to_hour_string(hour) + ':' + min_str;
133
20.6k
            cctz::time_zone result;
134
20.6k
            parse_tz_offset_string(offset_str, result);
135
20.6k
            lower_zone_cache_->emplace(offset_str, result);
136
20.6k
        }
137
6.88k
    }
138
    // -00 for hour is also valid
139
255
    std::string offset_str = "-00:00";
140
255
    cctz::time_zone result;
141
255
    parse_tz_offset_string(offset_str, result);
142
255
    lower_zone_cache_->emplace(offset_str, result);
143
255
    offset_str = "-00:30";
144
255
    parse_tz_offset_string(offset_str, result);
145
255
    lower_zone_cache_->emplace(offset_str, result);
146
255
    offset_str = "-00:45";
147
255
    parse_tz_offset_string(offset_str, result);
148
255
    lower_zone_cache_->emplace(offset_str, result);
149
255
}
150
151
180k
bool TimezoneUtils::find_cctz_time_zone(const std::string& timezone, cctz::time_zone& ctz) {
152
180k
    if (auto it = lower_zone_cache_->find(to_lower_copy(timezone)); it != lower_zone_cache_->end())
153
132k
            [[likely]] {
154
132k
        ctz = it->second;
155
132k
        return true;
156
132k
    }
157
48.7k
    return false;
158
180k
}
159
160
21.4k
bool TimezoneUtils::parse_tz_offset_string(const std::string& timezone, cctz::time_zone& ctz) {
161
    // like +08:00, which not in timezone_names_map_
162
21.4k
    re2::StringPiece value;
163
21.4k
    if (time_zone_offset_format_reg.Match(timezone, 0, timezone.size(), RE2::UNANCHORED, &value, 1))
164
21.4k
            [[likely]] {
165
21.4k
        bool positive = value[0] != '-';
166
167
        //Regular expression guarantees hour and minute must be int
168
21.4k
        int hour = std::stoi(value.substr(1, 2).as_string());
169
21.4k
        int minute = std::stoi(value.substr(4, 2).as_string());
170
171
        // timezone offsets around the world extended from -12:00 to +14:00
172
21.4k
        if (!positive && hour > 12) {
173
1
            return false;
174
21.4k
        } else if (positive && hour > 14) {
175
1
            return false;
176
1
        }
177
21.4k
        int offset = hour * 60 * 60 + minute * 60;
178
21.4k
        offset *= positive ? 1 : -1;
179
21.4k
        ctz = cctz::fixed_time_zone(cctz::seconds(offset));
180
21.4k
        return true;
181
21.4k
    }
182
1
    return false;
183
21.4k
}
184
185
#include "common/compile_check_end.h"
186
} // namespace doris