How to extract an arbitrary tag from a text file and measure the content length in Linux

Question

I have a text file, the content is made of custom tags :

<description>text</description><shortdescription>text for short description</shortdescription>

I want to extract in Linux command line the content of , which is "text for short description" and output its length if it is possible. Thank you.

Provide the xml node structure from root level? from root node to node having `shortdescription`? — Inian, Dec 19 '16 at 14:13
Possible duplicate of [How to execute XPath one-liners from shell?](http://stackoverflow.com/questions/15461737/how-to-execute-xpath-one-liners-from-shell) — Quentin, Dec 19 '16 at 14:13
@YvonHuynh — Why is this question marked as [tag:xml] then? Why are you using an XML-like format instead of actual XML? — Quentin, Dec 19 '16 at 14:19
@YvonHuynh: So you want to extract text within `` tag and its length? — Inian, Dec 19 '16 at 14:20
@Inian yes, the reason i want to do with shell is because i have potentially 3000 of these files and do not want to do this with a program, I modified the title — Yvon Huynh, Dec 19 '16 at 14:21
Can you output `grep --version` and see if it is `GNU grep`? — Inian, Dec 19 '16 at 14:22

Inian · Accepted Answer · 2016-12-19T14:32:41.087

Assuming your file is a flat file and NOT an XML you can find the string within the tags using GNU grep which supports perl style regEx match with a the -P flag. The flags -o and -m1 are for returning only the matching part and return 1 instance of the match only.

grep -oPm1 "(?<=<shortdescription>)[^<]+" input-file
text for short description

and for length of the string, store it in a variable the output of above command

stringLength="$(grep -oPm1 "(?<=<shortdescription>)[^<]+" file)"
echo "${#stringLength}"
26

How to extract an arbitrary tag from a text file and measure the content length in Linux

1 Answers1