bash - 循环 sed 的结果

Question

我无法遍历存储在变量中的 sed 返回的行。

目前看起来有点像这样

lines=$(sed -rne '/'"$timestamplastupload"'/,/'"$timestampnow"'/ p' /var/log/test.log)

for line in "$lines"; do

        echo "This line is: $line"

done

但到目前为止这还行不通，循环运行一次，而 $line 变量的内容似乎是整个 $lines 变量，因此只打印一次而不是逐行循环，前缀为“This line是:“

我缺少什么能够逐条循环遍历 sed 输出的每一行。 (最终我会逐行将匹配的上传到 API)所以我需要单独的每一行，以便我可以根据需要处理它们。

根据要求，sed 的一些输出已针对安全性进行了处理

Wed Feb 28 22:33:11 2018 us=452112 xxx/1.2.3.4:55487 [xxx] Inactivity timeout (--ping-restart), restarting
Wed Feb 28 22:33:11 2018 us=452112 xxx/1.2.3.4:55487 [xxx] Inactivity timeout (--ping-restart), restarting
Wed Feb 28 22:33:11 2018 us=452112 xxx/1.2.3.4:55487 [xxx] Inactivity timeout (--ping-restart), restarting
Wed Feb 28 22:33:11 2018 us=452180 xxx/1.2.3.4:55487 SIGUSR1[soft,ping-restart] received, client-instance restarting
Wed Feb 28 22:33:11 2018 us=452180 xxx/1.2.3.4:55487 SIGUSR1[soft,ping-restart] received, client-instance restarting
Wed Feb 28 22:33:11 2018 us=452180 xxx/1.2.3.4:55487 SIGUSR1[soft,ping-restart] received, client-instance restarting
Wed Feb 28 22:33:11 2018 us=452180 xxx/1.2.3.4:55487 SIGUSR1[soft,ping-restart] received, client-instance restarting

Answer 1

不要使用for 来阅读。改为使用进程替换:

while read -r line; do
  # your logic
done < <(sed -rne '/'"$timestamplastupload"'/,/'"$timestampnow"'/ p' /var/log/test.log)

参见:

• BashFAQ/001 - How can I read a file (data stream, variable) line-by-line (and/or field-by-field)?