binsee
binsee
暂不能查看下一页，待解决
推送的语音消息xml结构中，有两个关键字段length和voicelength。前者是语音silk数据尺寸，后者是语音时间长度(ms)。 当语音数据大于30Kb时，不会在推送中拉取原始语音数据，需要单独调用getMsgVoice接口拉取原始语音数据。 但是这个接口最大拉取的数据尺寸为60Kb（或许是缓冲区限制成了60Kb），因此大于60Kb的语音数据，会丢失后半截数据，导致语音不全。 测试日志如下： ``` [2018-05-04T05:21:09.944] [INFO] app - 收到来自 xxx 的语言消息，包含语音数据：false，xml内容： [2018-05-04T05:21:10.256] [INFO] app - 获取消息原始语音结果：true, 获得语音base64尺寸：81920 [2018-05-04T05:21:10.257] [INFO] app - 语音数据语音秒数：51082 ms，xml尺寸：67477，拉取到数据尺寸：61440 [2018-05-04T05:21:10.257] [INFO] app - 语音数据base64尺寸：81920 [2018-05-04T05:21:10.777]...
并不是服务器缓存问题，是目前协议内部的一个小bug
Is there any progress on this please? @alexander-fenster
See #80 for updated results
@huan I updated the test. And made some modifications with reference to the nodejs code, it should work fine now. However, I still think 60 seconds is too long and...
Ok, I don't think there is enough testing in various scenarios now, and there is no need to set the default value too small. In use, I can set the...
I've spent a day researching this problem, but didn't find a solution. Therefore, I changed the test to run concurrently to reduce the overall time taken by this test. Looks...
I investigated the source code of nodejs and found that the problem should be that `socket.setTimeout` uses the built-in `setUnrefTimeout` instead of the usual `setTimeout`. Regardless of sion or jest,...
Have you found any good solution?