CODEWITHSUNDEEP
awknumbersextractmaxlength
tok19
tok19

Reputation: 41

awk numbers extraction length

I am unable to catch numbers from file1 from column $3, where should be met condition, that just only numbers in length of exact 8 numbers before underscore or after underscore are extracted.

This below doesn't work. Only 2nd is a bit close to solution:

awk '{gsub(/[^0-9]{8}/,"",$3) && $22 = substr($3,1,8)}1' file1
awk '{gsub(/[^[:digit:]]/,"",$3) && $3 = substr($3,1,8)}1' file1

file1:

44446464654 field 2020135_7777_5x_definition
49989798797 mach 202013_7777_a5_19800511
498797877 pat math_200418_memo05874_12345678
44444 def math_200418_memo05874789_12345678
5659 explaination 28008874_memo
5659 explaination _28008874_memo

Needed output:

49989798797 mach 19800511
498797877 pat 12345678
44444 def 12345678
5659 explaination 28008874
5659 explaination 28008874

Upvotes: 1

Views: 67

Answers (1)

Cyrus
Cyrus

Reputation: 88989

With GNU awk. Split last field ($NF) with _ in an array and use a for loop with its elements. In every loop check content of current element with a regex.

awk '{split($NF,array,"_"); for(i in array) { if(array[i]~/^[0-9]{8}$/) { print $1,$2,array[i]; next } } }' file

Output:

49989798797 mach 19800511
498797877 pat 12345678
44444 def 12345678
5659 explaination 28008874
5659 explaination 28008874

See: 8 Powerful Awk Built-in Variables – FS, OFS, RS, ORS, NR, NF, FILENAME, FNR

Upvotes: 2

Related Questions