CODEWITHSUNDEEP
regexgrep
Matt Blake
Matt Blake

Reputation: 1

How to do a proper search/replace using GREP Regular Expressions in a text editor?

I am trying to run some regular expressions(grep) on a text file of about 4K lines. The main portion that I need replaced looks like this: 

1,"An Internet-Ready Resume",1,2,"","

And I need it to look like this:

<item>
<title>An Internet-Ready Resume</title>
<category>1</category>
<author>2</author>
<content>

So far, this is what I was trying to no avail:

[0-9]{1}\,\"*\"\,[0-9]\,[0-9]\,\"\"\,\"

Upvotes: 0

Views: 152

Answers (2)

John Kugelman
John Kugelman

Reputation: 361927

The error is the \"*\" part. When you use the * operator you need to tell it what is to be repeated. As written it is going to repeat the previous quote character. Instead of that you should tell it to repeat any character (.), thus: \".*\"

A secondary comment is that you have a lot of unnecessary backslashes. In fact, none of them are necessary as far as I can tell. Without them your regex looks like:

[0-9],".*",[0-9],[0-9],"","

Upvotes: 0

Miguel Ventura
Miguel Ventura

Reputation: 10468

You should start with doing a little reading on regular expressions. There are tons of useful resources online. Then you would see that:

  • you needn't escape everything (such as commas or quotes)
  • the asterisk * doesn't mean anything, but zero or more times
  • the any character is the . character. .* means any character any number of times (or anything)
  • if you need to make substitutions where you need atoms of what you're searching, you have to set those atoms by using (<atom content>) where <atom content> is a bit of a regexp.

A tip to start: instead of \"*\" try ".*"; Check the reference.

Also note that the part regarding the replacement will depend on the text editor/tool you're using. Usually a regexp such as (a)(b) (where a,b are regexp atoms) being replaced by x\1y\2z would produce xaybz.

Upvotes: 1

Related Questions