Git URL - Pull out substring via Shell (awk & sed)?

Data Mastery 0
Viewed 54

I have got the following URL:

https://xcg5847@git.rz.bankenit.de/scm/smat/sma-mes-test.git

I need to pull out smat-mes-test and smat:

git config --local remote.origin.url|sed -n 's#.*/\([^.]*\)\.git#\1#p'

sma-mes-test

This works. But I also need the project name, which is smat

I am not really familiar to complex regex and sed, I was able to find the other command in another post here. Does anyone know how I am able to extract the smat value here?

awk sed
6 Answers

With your shown samples please try following awk code. Simple explanation would be, setting field separator(s) as / and .git for all the lines and in main program printing 3rd last and 3nd last elements from the line.

your_git_command | awk -F'/|\\.git' '{print $(NF-2),$(NF-1)}'
RavinderSingh130

Your sed is pretty close. You can just extend it to capture 2 values and print them:

git config --local remote.origin.url |
sed -E 's~.*/([^/]+)/([^.]+)\.git$~\1 \2~'

smat sma-mes-test

If you want to populate shell variable using these 2 values then use this read command in bash:

read v1 v2 < <(git config --local remote.origin.url |
sed -E 's~.*/([^/]+)/([^.]+)\.git$~\1 \2~')

# check variable values
declare -p v1 v2
declare -- v1="smat"
declare -- v2="sma-mes-test"
anubhava0

Using sed

$ sed -E 's#.*/([^/]*)/#\1 #' input_file
smat sma-mes-test.git
HatLess0

I would harness GNU AWK for this task following way, let file.txt content be

https://xcg5847@git.rz.bankenit.de/scm/smat/sma-mes-test.git

then

awk 'BEGIN{FS="/"}{sub(/\.git$/,"",$NF);print $(NF-1),$NF}' file.txt

gives output

smat sma-mes-test

Explanation: I instruct GNU AWK that field separator is slash character, then I replace .git (observe that . is escaped to mean literal dot) adjacent to end ($) in last field ($NF), then I print 2nd from end field ($(NF-1)) and last field ($NF), which are sheared by space, which is default output field separator, if you wish to use other character for that purpose set OFS (output field separator) in BEGIN. If you want to know more about NF then read 8 Powerful Awk Built-in Variables – FS, OFS, RS, ORS, NR, NF, FILENAME, FNR

(tested in gawk 4.2.1)

Daweo0

Why not sed 's!.*/\(.*/.*\)!\1!'?

Enlico0 
string=$(config --local remote.origin.url | tail -c -21)

var1=$(echo "${string}" | cut -d'/' -f1)
var2=$(echo "${string}" | cut -d'/' -f2 | sed s'@\.git@@')

If you have multiple urls with variable lengths, this will not work, but if you only have the one, it will.

var1=smat
var2=sma-mes-test.git

If I did have something variable, personally I would replace all of the forward slashes with carriage returns, throw them into a file, and then export the last and second last lines with ed, which would give me the two last segments of the url.

Regular expressions literally give me a migraine headache, but as long as I can get everything on its' own line, I can quite easily bypass the need for them entirely.

petrus40
Related