How do i remove a specefic character from a file?

问题

I have this JSON file in this format;

[{
"item-id" : "0",
"item-name" : "Dwarf remains",
"item-examine" : "The body of a Dwarf savaged by Goblins.",
"shop-value" : "0.0",
"low-alch" : "0.0",
"high-alch" : "0.0",
"stab-bonus" : "0",
"slash-bonus" : "0",
"crush-bonus" : "0",
"magic-bonus" : "0",
"ranged-bonus" : "0",
"stab-defence" : "0",
"slash-defence" : "0",
"crush-defence" : "0",
"magic-defence" : "0",
"ranged-defence" : "0",
"strength-bonus" : "0",
"prayer-bonus" : "0",
},
{
"item-id" : "1",
"item-name" : "Toolkit",
"item-examine" : "Good for repairing a broken cannon.",
"shop-value" : "0.0",
"low-alch" : "0.0",
"high-alch" : "0.0",
"stab-bonus" : "0",
"slash-bonus" : "0",
"crush-bonus" : "0",
"magic-bonus" : "0",
"ranged-bonus" : "0",
"stab-defence" : "0",
"slash-defence" : "0",
"crush-defence" : "0",
"magic-defence" : "0",
"ranged-defence" : "0",
"strength-bonus" : "0",
"prayer-bonus" : "0",
}]

But with 25 000 entries

But in the file there is something wrong;

"prayer-bonus" : "0",

should be

"prayer-bonus" : "0"

How can i use regex or whatever to only replace the last bit?

Im using java.

By the way: the value is not always "0"

So how can i have it be a wildcard so it replaces all ?

回答1:

You can use this (written in java style, to use with replaceAll() method):

search: (?<!\\\\)(\"(?>[^\\\"]++|\\\\{2}|\\\\.)*+\"),(\\s*+})
replace: $1$2

The main interest of this pattern is that you avoid escaped double quotes that may be inside quotes.

Example:

String pattern = "(?<!\\\\)(\"(?>[^\\\"]++|\\\\{2}|\\\\.)*+\"),(\\s*+})";
String result = yourJson.replaceAll(pattern, "$1$2");

回答2:

use sed:

sed -i.bak 's/\("prayer-bonus" : "0"\),/\1/' file

To match any number:

sed -i.bak 's/\("prayer-bonus" : "[0-9]*"\),/\1/' file

回答3:

It seems like you may be trying to patch a problem caused by improperly formed JSON blobs. You could take the other responder's answer of using sed to fix these symptoms but the problem lies in how the JSON is being created.