linux awk搜索并删除多行

Question

我有一些json文件，其中有一些逗号出现在大括号

之前

 "good": "line,"
 "something": "blah",
}

然而，情况并非如此，

 "also_good": "line",
 "other": "blah2"
}

我在创建一个awk命令时遇到问题，该命令只会在新行上的大括号之前删除逗号。

Answer 1

这awk可能会：

awk '/^[ \t]*}/ {sub(/,$/,"",s)} NR>1 {print s} {s=$0} END {print s}' file
 "good": "line,"
 "something": "blah"
}

如果下一行以,

开头，它只会删除行尾的}

Answer 2

使用GNU awk表示多字符RS和gensub（）以及字符类缩写（例如\s）：

awk -v RS='^$' '{$0=gensub(/,(\s*\n\s*})/,"\\1","g")}1' file

e.g。当在此输入文件上运行时（注意保留第二个块之间的空格和}}：

$ cat file
 "good": "line,"
 "something": "blah",
}

 "even_gooder": "line,"
 "something": "note the spaces",


    }

 "also_good": "line",
 "other": "blah2"
}

$ awk -v RS='^$' '{$0=gensub(/,(\s*\n\s*})/,"\\1","g")}1' file
 "good": "line,"
 "something": "blah"
}

 "even_gooder": "line,"
 "something": "note the spaces"


    }

 "also_good": "line",
 "other": "blah2"
}