Boost PropertyTree provides a tree structure to store key/value pairs. Tree structures means that a trunk exists with numerous branches that have numerous twigs. A file system is a good example of a tree structure. File systems have a root directory with subdirectories that themselves can have subdirectories and so on.

1. accessing data in boost::property_tree::ptree

#include <boost/property_tree/ptree.hpp>
#include <iostream> using boost::property_tree::ptree; int main() {
ptree pt;
pt.put("C:.Windows.System", "20 files"); ptree& c = pt.get_child("C:");
ptree& windows = c.get_child("Windows");
ptree& system = windows.get_child("System");
std::cout << system.get_value<std::string>() << std::endl;
return ;
}

put() expects two parameters because boost::property_tree::ptree is a tree structure that saves key/value pairs. The tree doesn't just consist of branches and twigs, a value must be assigned to each branch and twig.

The first parameter passed to put() is more interesting. It is a path to a directory. However, it doesn't use the backslash, which is the common path separator on Windows. It uses the dot.

To access a subbranch, you call get_child(), which returns a reference to an object of the same type get_child() was called on.

2. accessing data in basic_ptree<std::string, int>

#include <boost/property_tree/ptree.hpp>
#include <utility>
#include <iostream> int main() {
typedef boost::property_tree::basic_ptree<std::string, int> ptree;
ptree pt;
pt.put(ptree::path_type{"C:\\Windows\\System", '\\'}, );
pt.put(ptree::path_type{"C:\\Windows\\Cursors", '\\'}, ); ptree& windows = pt.get_child(ptree::path_type{"C:\\Windows"}, '\\');
int files = ;
for (const std::pair<std::string, ptree>& p : windows) {
files += p.second.get_value<int>();
}
std::cout << files << std::endl;
return ;
}

By default, Boost.PropertyTree uses a dot as the separator for keys. If you need to use another character, such as backslash, as the separator, you don't pass the key as a string to put(). Instand you wrap it in an object of type boost::property_tree::ptree::path_type. The constructor of this class, which depends on boost::property_tree::ptree, takes the key as its first parameter and the separator as its second parameter.

3. accessing data with a translator

#include <boost/property_tree/ptree.hpp>
#include <boost/optional.hpp>
#include <cstdlib> struct string_to_int_translator {
typedef std::string internal_type;
typedef int external_type; boost::optional<int> get_value(const std::string& s) {
char* c;
long l = std::strtol(s.c_str(), &c, );
return boost::make_optional(c != s.c_str(), static_cast<int>());
}
}; int main() {
typedef boost::property_tree::iptree ptree;
ptree pt;
pt.put(ptree::path_type{"C:\\Windows\\System", '\\'}, "20 fifles");
pt.put(ptree::path_type{"C:\\Windows\\"}, "50 files"); string_to_int_translator tr;
int files = pt.get<int>(ptree::path_type{"C:\\windows\\system", '\\'}, tr) + pt.get<int>(ptree::path_type{"C:\\windows\\cursors", '\\'}, tr);
std::cout << files << std::endl;
return ;
}

boost::property_tree::iptree doesn't distinguish between lower and upper case. Just as put() can be used to store a value in a subbranch directly, a value from a subbranch can be read with get(). The key is defined the same way---using boost::property_tree::iptree::path_type.

Like get_value(), get() is a function template. You have to pass the type of the return value as a template parameter. Boost.PropertyTree does an automatic type conversion.

string_to_int_translator converts a value of type std::string to int. The translator is passed as an additional parameter to get(). Because the translator is just used to read, it only defines one member function, get_value(). If you want to use the translator for writing, too, then you would need to define a member function put_value() and then pass the translator as an additional parameter to put().

4. various member functions of boost::property_tree::ptree

#include <boost/property_tree/ptree.hpp>
#include <utility>
#include <iostream> using boost::property_tree::ptree; int main() {
ptree pt;
pt.put("C:.Windows.System", "20 files"); boost::optional<std::string> c = pt.get_optional<std::string>("C:");
std::cout << std::boolalpha << c.is_initialized() << std::endl; pt.put_child("D:.Program Files", ptree{"50 files"});
pt.put_child("D:.Program Files", ptree{"60 files"}); ptree d = pt.get_child("D:");
for (const std::pair<std::string, ptree>& p : d) {
std::cout << p.second.get_value<std::string>() << std::endl;
} boost::optional<ptree&> e = pt.get_child_optional("E:");
std::cout << e.is_intialized() << std::endl;
return ;
}

You can call the member function get_optional() if you want to read the value of a key, but you aren't sure if the key exists. get_optional() returns the value in an object of type boost::optional. The object is empty if the key wasn't found. Otherwise, get_optional() works the same as get().

The difference between put_child() and add_child() is that put_child() accesses a key if that key already exists, while add_child() always inserts a new key into the tree. get_child_optional() is used like get_child(). get_child_optional() returns an object of type boost::optional.

5. serializing a boost::property_tree::ptree in the JSON format

#include <boost/property_tree/ptree.hpp>
#include <boost/property_tree/json_parser.hpp>
#include <iostream> using namespace boost::property_tree; int main() {
ptree pt;
pt.put("C:.Windows.System", "20 files");
pt.put("C:.Windows.Cursors", "50 files"); json_parser::write_json("file.json", pt); ptree pt2;
json_parser::read_json("file.json", pt2); std::cout << std::boolalpha << (pt == pt2) << std::endl;
return ;
}

write_json() and read_json() make it possible to save and load a boost::property_tree::ptree serialized in the JSON format. That way you can support configuration files in the JSON format.

Except json format, Boost.PropertyTree supports additional data formats. you use boost::property_tree::ini_parser::write_ini() adn boost::property_tree::ini_parser::read_ini() to supprot ini-files; boost::property_tree::xml_parser::write_xml() and boost::property_tree::xml_parser::read_xml to support xml format.

boost propertyTree的更多相关文章

  1. 使用Boost.PropertyTree处理XML、JSON和INI数据

    Boost.PropertyTree 应该是 Boost 1.41.0 开始正式加入 Boost 版本的.目前 ( 2010/02/28 ) 能下到的最新版本是 1.42.0. 主要作用/应用场合 B ...

  2. Boost.PropertyTree读取ini文件(Linux环境)

    昨天因为需要读取配置文件略略伤神.网上很多例子但是我用都会报错,很多人把Boost.PropertyTree的函数写很麻烦的包所以报错我也不知道什么问题,所以今天整理下. 头上附上官网链接:Boost ...

  3. Boost.JSON Boost的JSON解析库(1.75首发)

    目录 目录 Boost的1.75版本新库 JSON库简介 JSON的简单使用 编码 最通用的方法 使用std::initializer_list json对象的输出 两种对比 解码 简单的解码 增加错 ...

  4. Awesome C/C++

    Awesome C/C++ A curated list of awesome C/C++ frameworks, libraries, resources, and shiny things. In ...

  5. awesome cpp

    https://github.com/fffaraz/awesome-cpp Awesome C/C++ A curated list of awesome C/C++ frameworks, lib ...

  6. C/C++ 开源库及示例代码

    C/C++ 开源库及示例代码 Table of Contents 说明 1 综合性的库 2 数据结构 & 算法 2.1 容器 2.1.1 标准容器 2.1.2 Lockfree 的容器 2.1 ...

  7. TinyXML2的快速实践

    最近遇到个需要在C++中处理XML文件的需求,虽然对此方面并不是很熟,但好在有GitHub上的awesome-cpp项目的帮助,还是收获了足够的相关知识. 类库 常用的或被推荐的XML类库有以下数个选 ...

  8. [转]awsome c++

    原文链接 Awesome C++ A curated list of awesome C++ (or C) frameworks, libraries, resources, and shiny th ...

  9. 基于BOOST 实现并发服务器框架

    一:设计思路 本服务器框架使用 UDP 传输协议,程序柱线程等待客户端数据,并将数组存取队列缓冲区.另外可开启多个工作线程,工作线程可以依据具体项目实现不同的功能 ,例如可以将队列缓冲区中的数据逐个取 ...

随机推荐

  1. 第九届ECNU Coder F.蚂蚁(栈)

    题目链接:http://acm.ecnu.edu.cn/contest/16/problem/F/ 题目: F. 蚂蚁 Time limit per test: 0.5 seconds Time li ...

  2. 用maven给SpringBoot项目打包

    注意要点: 1.注意某个moule有依赖需要在对应的pom.xml里填写有关的信息,如: <dependencies> <dependency> <artifactId& ...

  3. 这才是Tomcat内存配置的正确姿势

    1.背景 虽然阅读了各大牛的博客或文章,但并没有找到特别全面的关于JVM内存分配方法的文章,很多都是复制黏贴 为了严谨,本文特别备注只介绍基于HotSpot VM虚拟机,并且基于JDK1.7的内存分配 ...

  4. 点击" ./start_navicat"安装出现界面便面为乱码

    环境:ubuntu16.4 下载Navicat:navicat112_mariadb_cs_x64.tar.gz 点击" ./start_navicat"安装出现界面便面为乱码 解 ...

  5. PHP不使用内置函数intval(),实现字符串转整数

    平时我们用PHP时,将字符串转化为整型时,一般都是使用 intval() 内置函数,那么如果我们自己写,怎么写一个呢? 此时我们可以利用 ASCII 码计算得整数的特性,因为每个字符都对应一个 ASC ...

  6. EhCache缓存框架的使用

    EhCache 是一个纯Java的进程内缓存框架,具有快速.精干等特点,是Hibernate中默认的CacheProvider. 我们使用EhCache缓存框架主要是为了判断重复Url,每次爬取一个网 ...

  7. Jsoup获取DOM元素

    (1)doc.getElementsByTag(String tagName); (2)doc.getElementById(String id); (3)doc.getElementsByClass ...

  8. Ubuntu 18.04 截图工具-flameshot(安装及使用)

    安装flameshot:https://github.com/lupoDharkael/flameshot 安装命令: sudo apt-get install flameshot 设置>设备& ...

  9. jenkins展示html测试报告(不使用html publisher)

     最终效果: 点击『测试报告』,查看测试报告: 一开始的思路是: jenkins上展示html报告[转载] jenkins使用HTML Publisher Plugin插件 拉取报告样式缺失问题解决 ...

  10. (转载)Manacher'sAlgorithm: O(n)时间求字符串的最长回文子串

    以下内容转载自:传送门 源于这两篇文章: http://blog.csdn.net/ggggiqnypgjg/article/details/6645824http://zhuhongcheng.wo ...