delete records from a file

Shell Programming and Scripting
1

I have a big file with "|" delimiter. I want to delete all the records that have 'abc' in the 2nd field. How can i do that? I am not abe to open it in VI that is why i need to do it from outside. Please suggest

2 
awk -F'|' '$2 !~ /^abc$/' input_file.txt > output_file.txt
3

You can also use sed:

sed -e "/^[^\|]*\|abc\|/d" input_file.txt > output_file.txt
4

Thanks Klashxx

Can you explain what this means.

awk -F'|' '$2 !~ /^abc$/' input_file.txt > output_file.txt

does !~ means equal to? i thought its not equal to?

5

I think this means 'contains'

the '^abs' means is there the chars 'abc' at the beginning of $2 which is the field delimited by the character given in the '-F' flag to awk

6

Yes it's almost right. The operator "!~" means "does not match regexp".
The given regular expression matches exactly the word "abc" in the second field, so all the lines that does match this criteria are discarded.

7

But it won't match for a field value 'comabcpvt'.For that just use /abc/ in the regex.

With regards
Dileep Pattayath