CODEWITHSUNDEEP
regexlinuxbashsed
MelonSmasher
MelonSmasher

Reputation: 670

Sed remove time stamp from file with regex

I'm trying to use sed to delete a pattern from an html file. The time stamp consist of a 1-2 digit number a four letter word and then the word ago

example:

25 mins ago

or:

1 hour ago

and so on. I've tried using sed like this:

sed -i "s/([0-9]{1,2}) [a-z]* ago//g"

Sed does nothing, i'm not sure if my regex is wrong or if I am not escaping characters the right way.

Edit: I fixed that expression by removing an extra space, thanks choroba. Now sed removes mos of the text from the file. The expression needs to be less greedy? should also mention that ever time stamp is surrounded by > < example:

>1 hour ago<

Edit: This is what worked for me. Thanks ravoori.

sed -i 's/[0-9]\{,2\} [[:alpha:]]\{4,5\} ago//g'

Any help is appreciated!

Upvotes: 1

Views: 2393

Answers (2)

iruvar
iruvar

Reputation: 23364

Try the below. You need to escape the quantifier metacharacters { and } with sed

echo "1 hour ago" | sed  's/[0-9]\{,2\} [[:alpha:]]\{4\} ago//g'

Upvotes: 3

iamauser
iamauser

Reputation: 11479

 echo "1 hour ago" | sed -e 's/.*ago$//g'

or

 sed -e 's/.*ago$//g' <filename>

This should remove any line that ends with ago in filename

You may not want to use this if you have other lines than just the timestamp that ends with ago. You didn't specify.

Upvotes: 0

Related Questions