使用 sed 命令提取子字符串

Question

I have one file which includes this string in it:

2020-12-21 10:46:49.165 INFO: [41] browser.leaveCallAndQuitBrowser() [2020-12-21T10:46:36+0000] [FINE] DevTools WebSocket Event: Runtime.consoleAPICalled 74D7A734C0BD6EEFA60271821A6A2F55 {
   "args": [ {
      "type": "string",
      "value": "2020-12-21T10:46:36.633Z"
   }, {
      "type": "string",
      "value": "[modules/xmpp/xmpp.js]"
   }, {
      "type": "string",
      "value": "\u003CA.connectionHandler>: "
   }, {
      "type": "string",
      "value": "My Jabber ID: recorder@recorder.mydomain/hjE0dMPL"
   } ],
...

I am trying to extract this string from the file recorder@recorder.mydomain.com/vT1gTnAz with sed command but I couldn't get it done. Can anyone with sed and regex experience help or guide me to do it?

Currently, I am doing it with two commands: I am first getting "My Jabber ID: recorder@recorder.mydomain.com/hjE0dMPL" and then replacing My Jabber ID: with an empty string.

grep -EiEio '\bMy Jabber ID: (recorder@[A-Z0-9.-]+\.[A-Z]{2,4}.*)\b' browser.0.txt | sed 's/^My Jabber ID: //g'

It would be more elegant to do it in one command though.

Answer 1

You should use a proper json parser like jq for this but if you cannot use jq for some reason, you can use sed:

sed -rn 's/(^.*My Jabber ID: )(.*)(".*$)/\2/p' file

Enable regular expressions with -r and then split the line into three sections using regular expressions, substituting the line for just the second section and printing.

Answer 2

You can solve this directly with grep. Suppose your content is in test.txt file:

 cat test.txt |grep -Po '"value": "My Jabber ID: \K[^"]*'

will return

recorder@recorder.room-test5.11sight.com/hjE0dMPL