正则表达式：匹配多个空格不起作用

Question

I have this content in file:

  core        = 7.x
  base theme  = adaptivetheme
  release     = 7.x-3.x
  engine      = phptemplate

I'm tring to get everything after: base theme = to the end of line, so I would like to get adaptivetheme .

Based on https://unix.stackexchange.com/a/24151/37309 I'm trying to do:

sed -n 's/base\stheme\s+=\s//p' file
sed -n 's/base\stheme += //p' file

But it doesn't work. However this does work:

sed -n 's/base\stheme  = //p' file

So what I'm doing wrong? I would like to match it regardless of number of spaces that there can be.

Answer 1

Use the -r -flag for extended regex-support:

sed -r 's/base\stheme\s+=\s//'

Answer 2

You can use

sed -n 's/base theme \+= //p' file

This says "one or more spaces" (that's the meaning of the escaped \\+ ).

By itself, sed uses only the basic Regex (BRE). If you want to use extended regex some rules change - like what characters need to be escaped or not. You can invoke the extended mode with the -r flag. Then the + no longer needs to be escaped.

These dialects of regex are very confusing. I keep having to look it up myself, too...