https://blog.youkuaiyun.com/InvictusCoderzh-cnhttps://blog.youkuaiyun.com/https://blog.youkuaiyun.com/InvictusCoder/article/details/169503049https://blog.youkuaiyun.com/InvictusCoder/article/details/169503049InvictusCoderFri, 30 Oct 2026 05:49:01 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169502957https://blog.youkuaiyun.com/InvictusCoder/article/details/169502957InvictusCoderFri, 30 Oct 2026 05:48:29 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169502869https://blog.youkuaiyun.com/InvictusCoder/article/details/169502869InvictusCoderFri, 30 Oct 2026 05:47:57 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169502772https://blog.youkuaiyun.com/InvictusCoder/article/details/169502772InvictusCoderFri, 30 Oct 2026 05:47:25 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169502661https://blog.youkuaiyun.com/InvictusCoder/article/details/169502661InvictusCoderFri, 30 Oct 2026 05:46:45 +0800 b的 lambda。看起来只是换个比较符号，但这里藏着 C++ 排序最常见的未定义行为：比较器不满足严格弱序严格弱序要求比较关系满足四条：非自反（comp(x, x)必须为假）、非对称（comp(x, y)为真则comp(y, x)为假）、传递（comp(x, y)且comp(y, z)则comp(x, z)）、等价关系传递。写成a <= b或a >= b就违反了非自反，std::sort。]]>https://blog.youkuaiyun.com/InvictusCoder/article/details/169502560https://blog.youkuaiyun.com/InvictusCoder/article/details/169502560InvictusCoderFri, 30 Oct 2026 05:46:09 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169502462https://blog.youkuaiyun.com/InvictusCoder/article/details/169502462InvictusCoderFri, 30 Oct 2026 05:45:37 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169502375https://blog.youkuaiyun.com/InvictusCoder/article/details/169502375InvictusCoderFri, 30 Oct 2026 05:45:02 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169502270https://blog.youkuaiyun.com/InvictusCoder/article/details/169502270InvictusCoderFri, 30 Oct 2026 05:44:28 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169502174https://blog.youkuaiyun.com/InvictusCoder/article/details/169502174InvictusCoderFri, 30 Oct 2026 05:43:54 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169502088https://blog.youkuaiyun.com/InvictusCoder/article/details/169502088InvictusCoderFri, 30 Oct 2026 05:43:19 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169501981https://blog.youkuaiyun.com/InvictusCoder/article/details/169501981InvictusCoderFri, 30 Oct 2026 05:42:44 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169501891https://blog.youkuaiyun.com/InvictusCoder/article/details/169501891InvictusCoderFri, 30 Oct 2026 05:42:12 +0800（C 语言里是它最大的特点是「类型无关」：通过void*接收任意数组的首地址，用元素字节数告诉它每个元素多大，再通过一个函数指针回调比较逻辑。C++ 标准库把它们一并纳入，所以在 C++ 里写std::qsort是合法的。标题里的「冒泡模拟实现」不是指用冒泡排序去替代qsort的内部算法，而是指按qsort的接口约定，用冒泡排序实现一个同签名的函数——同样的void* base。]]>https://blog.youkuaiyun.com/InvictusCoder/article/details/169501800https://blog.youkuaiyun.com/InvictusCoder/article/details/169501800InvictusCoderFri, 30 Oct 2026 05:41:39 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169501627https://blog.youkuaiyun.com/InvictusCoder/article/details/169501627InvictusCoderFri, 30 Oct 2026 05:40:36 +0800这些头文件里该有什么。C++ 代码当然可以直接调用它们——头文件里的声明本身就带extern "C"，所以不存在名字修饰（name mangling）不匹配的问题。POSIX 不是「Linux API」。Linux 基本符合 POSIX，又加了一大堆自己的扩展（]]>https://blog.youkuaiyun.com/InvictusCoder/article/details/169501534https://blog.youkuaiyun.com/InvictusCoder/article/details/169501534InvictusCoderFri, 30 Oct 2026 05:40:04 +0800中。它最常见的用途是让函数返回两个结果、作为关联容器的元素类型、以及在排序时把"键 + 值"当成一个整体比较。常见误解是"pair是一个小类，随便用用就行"。std::map的元素类型是，键是 const 的，不能通过first修改；会对参数做类型退化，传引用不会保留引用；结构化绑定不加会得到一份拷贝。本文按创建、访问、比较、典型用途的顺序讲pair，并说明它和std::tuple的关系。是两个公开成员firstsecond。]]>https://blog.youkuaiyun.com/InvictusCoder/article/details/169501436https://blog.youkuaiyun.com/InvictusCoder/article/details/169501436InvictusCoderFri, 30 Oct 2026 05:39:33 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169500702https://blog.youkuaiyun.com/InvictusCoder/article/details/169500702InvictusCoderFri, 30 Oct 2026 05:35:11 +0800https://blog.youkuaiyun.com/InvictusCoder/article/details/169500544https://blog.youkuaiyun.com/InvictusCoder/article/details/169500544InvictusCoderFri, 30 Oct 2026 05:34:19 +0800> n;为什么s是空的？为什么读文件循环多做了一次？为什么输入一个字母后，后面所有cin >>全失败？为什么读二进制文件总是差几个字节？流状态位（stream state flags）。本文按"标准 IO → 文件 IO → 字符串 IO"的顺序讲清楚。内容。]]>https://blog.youkuaiyun.com/InvictusCoder/article/details/169374172https://blog.youkuaiyun.com/InvictusCoder/article/details/169374172InvictusCoderThu, 29 Oct 2026 02:45:13 +0800