CODEWITHSUNDEEP
bashfileshelldesign-patternsawk
raj
raj

Reputation: 43

in awk facing issues while using variable as search pattern while searching in a file

date;
tocompdate=$(date --date="-90 days ago" +%d%m%Y)
echo $tocompdate;

awk -v pattern="$testdate" '{print "Some matter "nr[NR--]" matter "$testdate""};' tocompare.txt >> mail2sent.txt`

Here I am taking current date and adding 90 days to it and storing it in tocompdate. And comparing tocompdate with a file, where same date is present in a line and some other matter in other line before that line.

while search pattern is matched I need some text to be printed and "the line before matched line" to be printed and again matched line after few words.

For example:

File has below content

a1
28112019
b2
04032018
c3
04032018
d4
27072015
e5
27072015
f6
06012030

If I am comparing with 27072015, I want output as below

The d4 is on 27072015
The e5 is on 27072015

Upvotes: 2

Views: 37

Answers (3)

user4832408
user4832408

Reputation:

I propose a solution using only shell programming:

#!/bin/sh 

match="27072015"

while IFS= read -r i
do if [ "$i" = "$match" ]
   then printf "The %s is on %s\n" "$last" "$match"
   fi
   # test that we are not in a whitespace
   if [ -n "$i" ]
   then last="$i"
   fi
done < tocompare.txt

Note: This solution is only appropriate for very small and simple data sets without special characters. Additionally it's much slower than using awk.

Upvotes: 0

Jose Ricardo Bustos M.
Jose Ricardo Bustos M.

Reputation: 8164

Just for fun: Another solution using sed, grep and xargs

xargs -n 2 <file | grep -E '\b27072015$' | sed -r 's/^(\S+)\s(\S+)$/The \1 is on \2/g'

you get,

The d4 is on 27072015
The e5 is on 27072015

Upvotes: 0

Ed Morton
Ed Morton

Reputation: 203284

$ awk -v tgt="27072015" '$0==tgt{printf "The %s is on %s\n", p, $0} {p=$0}' file
The d4 is on 27072015
The e5 is on 27072015

Upvotes: 3

Related Questions