Redis教程(十五)：C语言连接操作代码实例

#include <stdio.h>#include <stdlib.h>#include <stddef.h>#include <stdarg.h>#include <string.h>#include <assert.h>#include <hiredis.h>void doTest(){  int timeout = 10000;  struct timeval tv;  tv.tv_sec = timeout / 1000;  tv.tv_usec = timeout * 1000;  //以带有超时的方式链接Redis服务器，同时获取与Redis连接的上下文对象。  //该对象将用于其后所有与Redis操作的函数。  redisContext* c = redisConnectWithTimeout("192.168.149.137",6379,tv);  if (c->err) {    redisFree(c);    return;  }  const char* command1 = "set stest1 value1";  redisReply* r = (redisReply*)redisCommand(c,command1);  //需要注意的是，如果返回的对象是NULL，则表示客户端和服务器之间出现严重错误，必须重新链接。  //这里只是举例说明，简便起见，后面的命令就不再做这样的判断了。  if (NULL == r) {    redisFree(c);    return;  }  //不同的Redis命令返回的数据类型不同，在获取之前需要先判断它的实际类型。  //至于各种命令的返回值信息，可以参考Redis的官方文档，或者查看该系列博客的前几篇  //有关Redis各种数据类型的博客。:)  //字符串类型的set命令的返回值的类型是REDIS_REPLY_STATUS，然后只有当返回信息是"OK"  //时，才表示该命令执行成功。后面的例子以此类推，就不再过多赘述了。  if (!(r->type == REDIS_REPLY_STATUS && strcasecmp(r->str,"OK") == 0)) {    printf("Failed to execute command[%s]./n",command1);    freeReplyObject(r);    redisFree(c);    return;  }  //由于后面重复使用该变量，所以需要提前释放，否则内存泄漏。  freeReplyObject(r);  printf("Succeed to execute command[%s]./n",command1);  const char* command2 = "strlen stest1";  r = (redisReply*)redisCommand(c,command2);  if (r->type != REDIS_REPLY_INTEGER) {    printf("Failed to execute command[%s]./n",command2);    freeReplyObject(r);    redisFree(c);    return;  }  int length = r->integer;  freeReplyObject(r);  printf("The length of 'stest1' is %d./n",length);  printf("Succeed to execute command[%s]./n",command2);  const char* command3 = "get stest1";  r = (redisReply*)redisCommand(c,command3);  if (r->type != REDIS_REPLY_STRING) {    printf("Failed to execute command[%s]./n",command3);    freeReplyObject(r);    redisFree(c);    return;  }  printf("The value of 'stest1' is %s./n",r->str);  freeReplyObject(r);  printf("Succeed to execute command[%s]./n",command3);  const char* command4 = "get stest2";  r = (redisReply*)redisCommand(c,command4);  //这里需要先说明一下，由于stest2键并不存在，因此Redis会返回空结果，这里只是为了演示。  if (r->type != REDIS_REPLY_NIL) {    printf("Failed to execute command[%s]./n",command4);    freeReplyObject(r);    redisFree(c);    return;  }  freeReplyObject(r);  printf("Succeed to execute command[%s]./n",command4);  const char* command5 = "mget stest1 stest2";  r = (redisReply*)redisCommand(c,command5);  //不论stest2存在与否，Redis都会给出结果，只是第二个值为nil。  //由于有多个值返回，因为返回应答的类型是数组类型。  if (r->type != REDIS_REPLY_ARRAY) {    printf("Failed to execute command[%s]./n",command5);    freeReplyObject(r);    redisFree(c);    //r->elements表示子元素的数量，不管请求的key是否存在，该值都等于请求是键的数量。    assert(2 == r->elements);    return;  }  for (int i = 0; i < r->elements; ++i) {    redisReply* childReply = r->element[i];    //之前已经介绍过，get命令返回的数据类型是string。    //对于不存在key的返回值，其类型为REDIS_REPLY_NIL。    if (childReply->type == REDIS_REPLY_STRING)      printf("The value is %s./n",childReply->str);  }  //对于每一个子应答，无需使用者单独释放，只需释放最外部的redisReply即可。  freeReplyObject(r);  printf("Succeed to execute command[%s]./n",command5);  printf("Begin to test pipeline./n");  //该命令只是将待发送的命令写入到上下文对象的输出缓冲区中，直到调用后面的  //redisGetReply命令才会批量将缓冲区中的命令写出到Redis服务器。这样可以  //有效的减少客户端与服务器之间的同步等候时间，以及网络IO引起的延迟。  //至于管线的具体性能优势，可以考虑该系列博客中的管线主题。  if (REDIS_OK != redisAppendCommand(c,command1)    || REDIS_OK != redisAppendCommand(c,command2)    || REDIS_OK != redisAppendCommand(c,command3)    || REDIS_OK != redisAppendCommand(c,command4)    || REDIS_OK != redisAppendCommand(c,command5)) {    redisFree(c);    return;  }  redisReply* reply = NULL;  //对pipeline返回结果的处理方式，和前面代码的处理方式完全一直，这里就不再重复给出了。  if (REDIS_OK != redisGetReply(c,(void**)&reply)) {    printf("Failed to execute command[%s] with Pipeline./n",command1);    freeReplyObject(reply);    redisFree(c);  }  freeReplyObject(reply);  printf("Succeed to execute command[%s] with Pipeline./n",command1);  if (REDIS_OK != redisGetReply(c,(void**)&reply)) {    printf("Failed to execute command[%s] with Pipeline./n",command2);    freeReplyObject(reply);    redisFree(c);  }  freeReplyObject(reply);  printf("Succeed to execute command[%s] with Pipeline./n",command2);  if (REDIS_OK != redisGetReply(c,(void**)&reply)) {    printf("Failed to execute command[%s] with Pipeline./n",command3);    freeReplyObject(reply);    redisFree(c);  }  freeReplyObject(reply);  printf("Succeed to execute command[%s] with Pipeline./n",command3);  if (REDIS_OK != redisGetReply(c,(void**)&reply)) {    printf("Failed to execute command[%s] with Pipeline./n",command4);    freeReplyObject(reply);    redisFree(c);  }  freeReplyObject(reply);  printf("Succeed to execute command[%s] with Pipeline./n",command4);  if (REDIS_OK != redisGetReply(c,(void**)&reply)) {    printf("Failed to execute command[%s] with Pipeline./n",command5);    freeReplyObject(reply);    redisFree(c);  }  freeReplyObject(reply);  printf("Succeed to execute command[%s] with Pipeline./n",command5);  //由于所有通过pipeline提交的命令结果均已为返回，如果此时继续调用redisGetReply，  //将会导致该函数阻塞并挂起当前线程，直到有新的通过管线提交的命令结果返回。  //最后不要忘记在退出前释放当前连接的上下文对象。  redisFree(c);  return;}int main() {  doTest();  return 0;}//输出结果如下：//Succeed to execute command[set stest1 value1].//The length of 'stest1' is 6.//Succeed to execute command[strlen stest1].//The value of 'stest1' is value1.//Succeed to execute command[get stest1].//Succeed to execute command[get stest2].//The value is value1.//Succeed to execute command[mget stest1 stest2].//Begin to test pipeline.//Succeed to execute command[set stest1 value1] with Pipeline.//Succeed to execute command[strlen stest1] with Pipeline.//Succeed to execute command[get stest1] with Pipeline.//Succeed to execute command[get stest2] with Pipeline.//Succeed to execute command[mget stest1 stest2] with Pipeline.