Remove long prefix only from a specific field in all the lines of a file?

问题

I have a file containing the following lines (3 fields delimited by space):

component1 /dev/user/test 12344
component2 master abcefa123
component3 trunk 72812
component4 /branch/user/integration bc89fa
component5 trunk 989091 
component6 integration/test bc7829ac
component7 /branch/dev/user/various eded34512

I need to manipulate the field 2 to cut its long prefix (same as you do in bash with ${string##*}) and to get the following result:

component1 test 12344
component2 master abcefa123
component3 trunk 72812
component4 integration bc89fa
component5 trunk 989091 
component6 test bc7829ac
component7 various eded34512

I have no idea how to do it.

回答1:

1st solution: Could you please try following, written and tested with shown samples in GNU awk.

awk '{num=split($2,arr,"/");$2=arr[num]} 1' Input_file

2nd solution: OR with shown samples only try setting field separators as space or /.

awk -F'[ /]' '{print $1,$(NF-1),$NF}' Input_file

3rd solution(using sed): Using sed, you could try like:

sed 's/\([^ ]*\).*\/\(.*\)/\1 \2/' Input_file

Explanation(1st solution): Adding detailed explanation for above.

awk '                   ##Starting awk program from here.
{
  num=split($2,arr,"/") ##Splitting 2nd field into array arr with / as field separator.
                        ##num is number of total elements of array arr.
  $2=arr[num]           ##Assigning last element of arr with index of num into 2nd field.
}
1                       ##Mentioning 1 will print the current line.
' Input_file            ##mentioning Input_file name here.

回答2:

I would use AWK for that following way, let content of file.txt be:

component1 /dev/user/test 12344
component2 master abcefa123
component3 trunk 72812
component4 /branch/user/integration bc89fa
component5 trunk 989091 
component6 integration/test bc7829ac
component7 /branch/dev/user/various eded34512

then

awk '{sub(/^.*\//, "", $2);print}' file.txt

outputs:

component1 test 12344
component2 master abcefa123
component3 trunk 72812
component4 integration bc89fa
component5 trunk 989091 
component6 test bc7829ac
component7 various eded34512

Explanation I simply replace everything from begin up to last / (which needs to be escaped hence \) in column of interest with empty string, then print it.

(tested in GNU Awk 5.0.1)

回答3:

A solution with awk:

awk '{ split($2,s,"/"); $2=s[length(s)]; print }' inputfile

The split($2,s,"/") will split the second varable to an array

$2=s[length(s)]; will assign the second variable with the last value of the array

print will print the comlete line.

回答4:

awk '{ split($2,map,"/");$2=map[length(map)] }1' file

Using awk. split the second space delimited field into an array called map using / as the delimiter. Substitute $2 with the last element of the map array. Print the lines with shorthand 1.

回答5:

Using sed:

sed -rn 's/(^.*)([[:space:]])(.*\/)?(.*)([[:space:]])(.*$)/\1 \4 \6/p' file

Split each line into sections based on regular expressions and substitute the line for the relevant sections, printing the result.

回答6:

Using sed:

sed -E 's/^([^ ]* )([^/]*\/)*/\1/' infile

回答7:

And also this awk using the loop while:

awk '{while ( n=split($2,a,/\//) ) {$2=a[n];print;next}}' file
component1 test 12344
component2 master abcefa123
component3 trunk 72812
component4 integration bc89fa
component5 trunk 989091
component6 test bc7829ac
component7 various eded34512

来源：https://stackoverflow.com/questions/64942896/remove-long-prefix-only-from-a-specific-field-in-all-the-lines-of-a-file