我从一个zip文件中提取了一串字符串：

866275 2016-02-06 11:25 Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg

我想要第一个字段（大小）和路径。

在一个在线正则表达式测试工具中测试时，这给出了我想要的结果：

(^[0-9]+)" "[0-9]{4}-[0-9]{2}-[0-9]{2}" "[0-9]{2}:[0-9]{2}" "(.+)

但当我把它放到脚本里时却没有输出

if [[ $x =~ (^[0-9]+)" "[0-9]{4}-[0-9]{2}-[0-9]{2}" "[0-9]{2}:[0-9]{2}" "(.+) ]]; then 
 echo bash0 Captured : ${BASH_REMATCH[0]}
 echo bash1 Captured number: ${BASH_REMATCH[1]}
 echo bash2 Captured NAME: ${BASH_REMATCH[2]}
fi

带有第一项：

if [[ $x =~ (^[0-9]+) ]]; then

它给出数字，但把其余任何部分加上后就没有输出

我需要改动什么才能让它工作？

解决方案

我把你原来的脚本放在 test.sh 里，在 GNU bash, version 5.2.21(1)-release (x86_64-pc-cygwin) 上按预期工作：

x='866275 2016-02-06 11:25 Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg'
if [[ $x =~ (^[0-9]+)" "[0-9]{4}-[0-9]{2}-[0-9]{2}" "[0-9]{2}:[0-9]{2}" "(.+) ]]; then
 echo bash0 Captured : ${BASH_REMATCH[0]}
 echo bash1 Captured number: ${BASH_REMATCH[1]}
 echo bash2 Captured NAME: ${BASH_REMATCH[2]}
fi

$ bash test.sh
bash0 Captured : 866275 2016-02-06 11:25 Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg
bash1 Captured number: 866275
bash2 Captured NAME: Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg

为便于阅读，我建议把正则表达式放到一个变量里：

x='866275 2016-02-06 11:25 Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg'
regex='(^[0-9]+) [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2} (.+)'
if [[ $x =~ $regex ]]; then
 echo bash0 Captured : ${BASH_REMATCH[0]}
 echo bash1 Captured number: ${BASH_REMATCH[1]}
 echo bash2 Captured NAME: ${BASH_REMATCH[2]}
fi

备选方案

看起来问题是在于 " " 在 bash 中。

看起来它把第一个 " 视为 (^[0-9]+) 的结束，第二个 " 视为更长字符串的开头。

我找到了两种方法：

1. 使用 \ 来转义空格，而不是 " "

x='866275 2016-02-06 11:25 Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg'

if [[ $x =~ (^[0-9]+)\ [0-9]{4}-[0-9]{2}-[0-9]{2}\ [0-9]{2}:[0-9]{2}\ (.+) ]]; then 
 echo bash0 Captured : ${BASH_REMATCH[0]}
 echo bash1 Captured number: ${BASH_REMATCH[1]}
 echo bash2 Captured NAME: ${BASH_REMATCH[2]}
fi

1. 把所有内容放在 ( ) 里，并使用普通空格而不是 " "
   但这需要在 BASH_REMATCH 中改变索引

x='866275 2016-02-06 11:25 Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg'

if [[ $x =~ ((^[0-9]+) [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2} (.+)) ]]; then 
 echo bash0 Captured : ${BASH_REMATCH[0]}
 echo bash1 Captured number: ${BASH_REMATCH[2]}
 echo bash2 Captured NAME: ${BASH_REMATCH[3]}
fi

结果：

bash0 Captured : 866275 2016-02-06 11:25 Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg
bash1 Captured number: 866275
bash2 Captured NAME: Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg

在以下环境测试：GNU bash 5.2.21，Linux Mint 22（基于Ubuntu 24.04）

用于测试的完整代码：

#!/bin/bash 

x='866275 2016-02-06 11:25 Documents/familystuff/Household/Agri Stuff/Sheep/sheep Photos/2014 ewe lambs/074Cleo 6Feb16a.jpg'

echo '--- 1 ---'

if [[ $x =~ ((^[0-9]+) [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2} (.+)) ]]; then 
 echo bash0 Captured : ${BASH_REMATCH[0]}
 echo bash1 Captured : ${BASH_REMATCH[1]}
 echo bash2 Captured number: ${BASH_REMATCH[2]}
 echo bash3 Captured NAME: ${BASH_REMATCH[3]}
fi

echo '--- 2 ---'

if [[ $x =~ (^[0-9]+)\ [0-9]{4}-[0-9]{2}-[0-9]{2}\ [0-9]{2}:[0-9]{2}\ (.+) ]]; then 
 echo bash0 Captured : ${BASH_REMATCH[0]}
 echo bash1 Captured number: ${BASH_REMATCH[1]}
 echo bash2 Captured NAME: ${BASH_REMATCH[2]}
fi

我在其他Stack Overflow站点也发现了类似的问题

• Stack Overflow： regex - 如何在Bash中用正则表达式匹配空格？
• Unix & Linux： regular expression - Bash regex whitespace before match

也许不完全是同一个问题，但也是寻找帮助的好去处

• Server Fault： 如何在Bash变量中匹配空格
• Super User： bash - Conditional spaces in a regular expression?