使用 bash 中的 sed 或 awk 修改文件中两个匹配字符串之间的行

Question

In bash how do I modify lines in yaml between two matching strings using awk or sed?

Like say I want lines between bbb and ddd modified,

aaa:
  bbb:
  - c1
  - c2
  - c3
  ddd: 111

should become,

aaa:
  bbb: ["c1","c2","c3"]
  ddd: 111

Answer 1

Something like

sed -e '/bbb:/,/ddd:/c\  bbb: ["c1","c2","c3"]\n  ddd: 111' myfile

Answer 2

Disclaimer: Regex is not suitable for scalable structured data solutions. Please use a dedicated YAML parser if at all possible.

awk -v key=bbb '
    NF == 1 && $1 == key ":" {flag=$0; next}
    flag != "" {
        if(sub(/^[\t ]*- /, "")) {
            array[NR] = $0; next
        } else {
            for(i in array) out = out "," "\"" array[i] "\""
            print flag, "[" substr(out,2) "]"
            split("", array) # POSIX delete array
            flag=out="" # POSIX delete variables
        }
    } {print}
'

this is an extremely specific, brittle solution and should by no means ever be applied to general YAML parsing

Answer 3

This might work for you (GNU sed):

sed -E '/^\s*bbb:/{:a;N;/^\s*ddd:/M!ba;s/^\s*-\s*//mg;s/\n/","/g;s/",(.*),"/ [\1]\n/}' file

Gather up lines between bbb: and ddd: .

Remove extra spaces and - characters at the beginnings or lines.

Replace newlines by "," .

Replace first ", by [ and last ," by ]\n .