ios嵌套的异步并行任务场景
场景1:
从数据源读到N个值,然后需要遍历这N个值,分别发起http请求。处理完成之后,调用一个最终汇总的方法
这个场景如果用js的async框架,很容易就能实现:
async.series([task1, task2, task3], function(err){ // 汇总代码 }); function task1(callback){ // http请求 callback(null); } function task2(callback){ // http请求 callback(null); } function task3(callback){ // http请求 callback(null); }
但是在ios里,就比较麻烦,最后是用dispatch_group实现的:
dispatch_group_t group = dispatch_group_create(); while([rs next]){ dispatch_group_enter(group); [syncService refreshMembersWithEnterpriseId:enterpriseId LatestSyncTime:latestSyncDate Block:^(BOOL flag){ dispatch_group_leave(group); }]; } dispatch_group_notify(group, dispatch_get_main_queue(), ^{ // 汇总代码 });
场景2:
后来需求变化了,场景更复杂了点。针对每个值,都要发起多个http请求(场景1只需要请求1次)。所有请求都结束以后,再调用汇总方法
所以一开始我写成类似这样:
dispatch_group_t group = dispatch_group_create(); while([rs next]){ dispatch_group_enter(group); dispatch_group_t group_inner = dispatch_group_create(); dispatch_group_enter(group_inner); [syncService refreshMembersWithEnterpriseId:enterpriseId LatestSyncTime:latestSyncDate Block:^(BOOL flag){ dispatch_group_leave(group_inner); }]; dispatch_group_enter(group_inner); [syncService refreshReportsWithEnterpriseId:enterpriseId LatestSyncTime:latestSyncDate Block:^(BOOL flag){ dispatch_group_leave(group_inner); }]; dispatch_group_notify(group_inner, dispatch_get_global_queue(0, 0), ^{ dispatch_group_leave(group); }); } dispatch_group_notify(group, dispatch_get_main_queue(), ^{ // 汇总代码 });
写完之后发现,其实在这个场景下,不需要嵌套group,共用一个group也没问题
dispatch_group_t group = dispatch_group_create(); while([rs next]){ dispatch_group_enter(group); [syncService refreshMembersWithEnterpriseId:enterpriseId LatestSyncTime:contactLatestSync Block:^(BOOL flag){ dispatch_group_leave(group); }]; dispatch_group_enter(group); [syncService refreshReportsWithEnterpriseId:enterpriseId LatestSyncTime:reportLatestSync Block:^(BOOL flag){ dispatch_group_leave(group); }]; } dispatch_group_notify(group, dispatch_get_main_queue(), ^{ // 汇总代码 });
因为这里并不需要针对每个组做汇总处理,如果每个组还有自己的汇总逻辑需要执行的话,就需要嵌套group了
郑重声明:本站内容如果来自互联网及其他传播媒体,其版权均属原媒体及文章作者所有。转载目的在于传递更多信息及用于网络分享,并不代表本站赞同其观点和对其真实性负责,也不构成任何其他建议。