sed无意中抹去了行的剩余部分

Question

试图在文件中找到各种类似的事件。

<address DOMAIN='0x000' bus='PCOIP_USB0' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='PCOIP_USB1' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='PCOIP_USB7' slot='0x00' function '0x0' />
<address type='bmu' domain='0x000' bus='PCOIP_USB9' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='PCOIP_USB4' slot='0x00' function '0x0' />

我试图简单地定位以PCOIP_开头的任何实例，并将该字符串替换为0x86

结果如下所示：

<address DOMAIN='0x000' bus='0x86' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='0x86' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='0x86' slot='0x00' function '0x0' />
<address type='bmu' domain='0x000' bus='0x86' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='0x86' slot='0x00' function '0x0' />

我目前正在尝试的处理替换必要字符串的命令，但它也删除了整行的其余部分：

sed s/PCOIP_\(.*\)/0x86/g' file

命令(原名为)使文件看起来像：

<address DOMAIN='0x000' bus='0x86
<address DOMAIN='0x000' bus='0x86
<address DOMAIN='0x000' bus='0x86
<address type='bmu' domain='0x000' bus='0x86
<address DOMAIN='0x000' bus='0x86

Answer 1

您可以使用

sed "s/PCOIP_[^']*/0x86/g" file

在这里，PCOIP_[^']*匹配PCOIP_，然后匹配除'字符以外的任何零或多个字符。整个匹配被替换为0x86子字符串。

请参阅在线演示：

#!/bin/bash
s="<address DOMAIN='0x000' bus='PCOIP_USB0' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='PCOIP_USB1' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='PCOIP_USB7' slot='0x00' function '0x0' />
<address type='bmu' domain='0x000' bus='PCOIP_USB9' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='PCOIP_USB4' slot='0x00' function '0x0' />"
sed "s/PCOIP_[^']*/0x86/g" <<< "$s"

输出：

<address DOMAIN='0x000' bus='0x86' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='0x86' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='0x86' slot='0x00' function '0x0' />
<address type='bmu' domain='0x000' bus='0x86' slot='0x00' function '0x0' />
<address DOMAIN='0x000' bus='0x86' slot='0x00' function '0x0' />