Question

我有以下数据：

{
 foo,
 bar,
 qux,
 zuu,
 sdf,
 sdfasdf
}
{
 abc,
 efg,
 hij,
 klm,
 nop,
 qrs
}

我想要做的是将它们转换为一行和逗号分隔的块：

 foo,bar,qux,zuu,sdf,sdfasdf
 abc,efg,hij,klm,nop,qrs

最好的unix shell编程方法是什么？

我正在尝试处理JSON数据，如下所示：

  {
   "Items": [
    {
        "A": {
            "S": "http://"
        }, 
        "B": {
            "N": "1"
        }, 
        "C": {
            "N": "0"
        }, 
        "D": {
            "N": "0"
        }, 
        "E": {
            "S": "2110221641826713463977"
        }, 
        "F": {
            "N": "2.5"
        }, 
        "G": {
            "S": " "
        }, 
        "H": {
            "B": "IA"
        }, 
        "I": {
            "N": "0"
        }, 
        "J": {
            "N": "0"
        }
    }, 
    {
        "A": {
            "S": "http://"
        }, 
        "B": {
            "N": "0"
        }, 
        "C": {
            "N": "1"
        }, 
        "D": {
            "N": "2"
        }, 
        "E": {
            "S": "67"
        }, 
        "F": {
            "N": "0"
        }, 
        "G": {
            "S": " "
        }, 
        "H": {
            "B": "Agh"
        }, 
        "I": {
            "N": "9"
        }, 
        "J": {
            "N": "1"
        }
    },     
  ]
 }

我正在使用JQ解析如下：

  jq -r '.Items[]'  | jq -r '{(.A.S): [.A.S, .B.B, .C.N, .D.S, .E.N, .F.N, .G.N, .H.N, .I.S]}'

这时解析JSON但保留上面的数据。

Answer 1

使用gnu awk可以在一个命令中完成：

gawk -v RS='\n?[{}]\n?' 'NF{gsub(/ *\n */, ""); print}' file
 foo,bar,qux,zuu,sdf,sdfasdf
 abc,efg,hij,klm,nop,qrs

Answer 2

这是你在找什么？

$ jq -c '.Items[]|[.[][]]' test.json
["http://","1","0","0","2110221641826713463977","2.5"," ","IA","0","0"]
["http://","0","1","2","67","0"," ","Agh","9","1"]

Answer 3

使用tr和sed：

tr -d '\n' < file | sed -r 's@\{ @@g; s@, @,@g; s@}@\n@g'

输出：

foo,bar,qux,zuu,sdf,sdfasdf
abc,efg,hij,klm,nop,qrs

此处还有一个仅使用sed的表单：

sed -n '$!{ H; b; }; g; s|\n||g; s|{ ||g; s|, |,|g; s|}|\n|g; p' file

Answer 4

这是awk

awk '!/^[{}]/ {printf "%s"(/,$/?"":RS),$1}' file
foo,bar,qux,zuu,sdf,sdfasdf
abc,efg,hij,klm,nop,qrs

Answer 5

您也可以使用此sed方法

sed ':loop ;N; /\n\}$/{P;d;t} ;/^{/{D} ;s/ *\(.*\)\n /\1/g ;t loop' filename

Answer 6

$ awk -v RS=} '{$1=$1; gsub(/[{ ]/,"")} $0' file
foo,bar,qux,zuu,sdf,sdfasdf
abc,efg,hij,klm,nop,qrs

有条件地删除shell脚本中的新行

6 个答案: